This service tests the validity of an RSS 2.0 feed, checking to see that it follows the rules of the RSS specification. For advice from the RSS Advisory Board on how to implement RSS and handle issues such as enclosures and HTML encoding, read the RSS Best Practices Profile. This checker is also a validator of Atom and RSS 1.0 feeds.
Use this tester regularly to ensure that your RSS feed continues to work well in the wide audience of RSS readers, podcast clients and other software that supports the format.
This is a valid RSS feed.
This feed is valid, but interoperability with the widest range of feed readers could be improved by implementing the following recommendations.
line 70, column 0: (2 occurrences) [help]
}
line 148, column 0: (66 occurrences) [help]
</description>
line 189, column 0: (7 occurrences) [help]
}
line 189, column 0: (6 occurrences) [help]
}
line 190, column 0: (12 occurrences) [help]
</code></pre>
<p>recently i came across <a href="https://www.uiua.org/?ref=s ...
<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>2024 Advent Of Code in Perl - (define sixhat (λ (dave) (display 'ideas)))</title>
<description>2024 Advent Of Code in Perl - (define sixhat (λ (dave) (display 'ideas)))</description>
<link>https://www.sixhat.net/</link>
<atom:link href="https://www.sixhat.net/rss.xml" rel="self" type="application/rss+xml" />
<language>en</language>
<copyright>Copyright 2023, David Rodrigues</copyright>
<pubDate>Mon, 09 Dec 2024 16:19:24 +0000</pubDate>
<item>
<title>On Rags - Retrieval Augmented Generation</title>
<link>https://www.sixhat.net/2024/06-on-rags-retrieval-augmented-generation.html</link>
<description>
<h1 id="on-rags---retrieval-augmented-generation"><a class="header" href="#on-rags---retrieval-augmented-generation">On Rags - Retrieval Augmented Generation</a></h1>
<p>A few links about Retrieval Augmented Generation:</p>
<p>— <a href="https://arxiv.org/abs/2005.11401?ref=sixhat.net">Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks</a></p>
<p>— <a href="https://r2r-docs.sciphi.ai/introduction?ref=sixhat.net">R2R - Open source RAG engine</a></p>
</description>
<pubDate>Wed, 26 Jun 2024 15:13:58 +0000</pubDate>
<guid>https://www.sixhat.net/2024/06-on-rags-retrieval-augmented-generation.html</guid>
</item>
<item>
<title>Reading on Darwin Machines</title>
<link>https://www.sixhat.net/2024/07-darwin-machines-brain-intelligence.html</link>
<description>
<h2 id="darwin-machines-must-read"><a class="header" href="#darwin-machines-must-read">Darwin Machines [must read]</a></h2>
<p>Cale Smith writes about <a href="https://vedgie.net/writing/darwin_machines.md?ref=sixhat.net">Darwin Machines</a> and their potential to harbour explanations for intelligence. </p>
<blockquote>
<p>If modern ML explains intelligence with layers, the theory of a Darwin Machine explains evolution with columns.</p>
</blockquote>
</description>
<pubDate>Wed, 17 Jul 2024 09:05:08 +0000</pubDate>
<guid>https://www.sixhat.net/2024/07-darwin-machines-brain-intelligence.html</guid>
</item>
<item>
<title>Eureka Labs, AI for Education</title>
<link>https://www.sixhat.net/2024/07-eureka-labs-ai-education.html</link>
<description>
<h2 id="ex-open-ai-researcher-launches-ai-platform-for-education"><a class="header" href="#ex-open-ai-researcher-launches-ai-platform-for-education">ex-Open AI researcher launches AI platform for education</a></h2>
<p>In the world of education ex-Open AI Andrej Karpathy is launching a new company--Eureka Labs--aimed at providing <em>AI Native</em> education. </p>
<p>Mainly the project aims to create AI assistants that will guide students through a human created curriculum. </p>
<blockquote>
<p>How can we approach an ideal experience for learning something new? For example, in the case of physics one could imagine working through very high quality course materials together with Feynman, who is there to guide you every step of the way. Unfortunately, subject matter experts who are deeply passionate, great at teaching, infinitely patient and fluent in all of the world's languages are also very scarce and cannot personally tutor all 8 billion of us on demand. </p>
</blockquote>
<p>via <a href="https://arstechnica.com/information-technology/2024/07/former-openai-researchers-new-company-will-teach-you-how-to-build-an-llm/?ref=sixhat.net">arstechnica</a></p>
</description>
<pubDate>Wed, 17 Jul 2024 09:04:07 +0000</pubDate>
<guid>https://www.sixhat.net/2024/07-eureka-labs-ai-education.html</guid>
</item>
<item>
<title>Adding something between the lines</title>
<link>https://www.sixhat.net/2024/06-something-between-the-lines.html</link>
<description>
<h3 id="adding-something-between-the-lines"><a class="header" href="#adding-something-between-the-lines">Adding something between the lines</a></h3>
<p>I've been fascinated with Japanese <a href="https://en.wikipedia.org/wiki/Furigana?ref=sixhat.net">furigana</a>. It is a writing system that runs parallel to other kanjis, whose function is mainly to clarify their pronunciation but, in some cases gives sentences another reading dimension that is not existent in other languages.
I was wondering why western languages don't have anything similar.
The closest think would be annotations in between the written lines, but nowadays with computers that is even more difficult—with typewriters you could half-step the line feed and write those notes—and so this kind of in between the writing seems restricted to pen and paper. Or is it? Although there's no modern text editor that allows for it, <span class="furigana">that I'm aware of,</span>in CSS we can simulate it with the following class: </p>
<pre><code class="language-css">.furigana {
position: absolute;
font-size: small;
transform: translate(-1rem -0.6rem);
}
</code></pre>
<style>
.furigana {
position: absolute;
font-size: small;
transform: translate(-1rem, -0.6rem);
}
</style>
</description>
<pubDate>Wed, 12 Jun 2024 10:52:37 +0000</pubDate>
<guid>https://www.sixhat.net/2024/06-something-between-the-lines.html</guid>
</item>
<item>
<title>Assange is free</title>
<link>https://www.sixhat.net/2024/06-25-assange-is-free.html</link>
<description>
<h2 id="julian-assange-is-free"><a class="header" href="#julian-assange-is-free">Julian Assange is free</a></h2>
<p>After five years (although is first arrest was in 14 years ago) incarcerated, it seems that this stalemate is no more. Wikileaks released information that Julian Assange was freed and is heading to Australia. He will accept a guilty verdict for espionage and will return home. </p>
<p>-- <a href="https://www.theguardian.com/media/article/2024/jun/25/julian-assange-plea-deal-with-us-free-to-return-australia?ref=sixhat.net">The Guardian</a></p>
</description>
<pubDate>Tue, 25 Jun 2024 06:41:08 +0000</pubDate>
<guid>https://www.sixhat.net/2024/06-25-assange-is-free.html</guid>
</item>
<item>
<title>New macs are rubbish</title>
<link>https://www.sixhat.net/2024/02-new-macs-are-rubish.html</link>
<description>
<h2 id="new-macs-are-rubish"><a class="header" href="#new-macs-are-rubish">New macs are rubish</a></h2>
<p>I just want to say something about new macs. They are rubish. I never had freezes and reboots with old macs.</p>
<p>I've been running a M2 air for more than one year and it has frozen and rebooted on me several times. If I push it too hard it panics and reboots.</p>
<p>Early on it was Wifi connection that wasn't stable and forced me to disable "Continuity" with my iPad, to try to keep it running longer than 5 minutes.</p>
<p>Now with AI trying to run the smallest Ollama models is a challenge. It might just poop on you and reboot. Twice in the last hour. ME NOT HAPPY.</p>
<p>Sorry, just wanted to get this out of my chest</p>
<p>PS - asking the wife for her Dell so I can run the models.</p>
</description>
<pubDate>Tue, 20 Feb 2024 10:33:48 +0000</pubDate>
<guid>https://www.sixhat.net/2024/02-new-macs-are-rubish.html</guid>
</item>
<item>
<title>My spelling config in vimrc</title>
<link>https://www.sixhat.net/2024/11-my-spelling-config-in-vimrc.html</link>
<description>
<h1 id="my-spelling-config-in-vimrc"><a class="header" href="#my-spelling-config-in-vimrc">My spelling config in .vimrc</a></h1>
<p>I jump between English and Portuguese in my writings and it is nice to have a spellchecker working to catch typos. In vim, I just have the following three lines in <strong>.vimrc</strong> mapping F10 (disable), F11 (Portuguese) and F12 (English) for the local buffer. I don't use a global setting with <code>:set</code> because coding doesn't need the spelling turned on. My <em>Leader</em> key is <em>Space</em>.</p>
<pre><code>nmap &lt;Leader&gt;&lt;F12&gt; &lt;Esc&gt;:setlocal spell spelllang=en_gb&lt;CR&gt;
nmap &lt;Leader&gt;&lt;F11&gt; &lt;Esc&gt;:setlocal spell spelllang=pt_pt&lt;CR&gt;
nmap &lt;Leader&gt;&lt;F10&gt; &lt;Esc&gt;:setlocal nospell&lt;CR&gt;
</code></pre>
</description>
<pubDate>Tue, 19 Nov 2024 12:40:51 +0000</pubDate>
<guid>https://www.sixhat.net/2024/11-my-spelling-config-in-vimrc.html</guid>
</item>
<item>
<title>added a rss feed</title>
<link>https://www.sixhat.net/2023/added-a-rss-feed.html</link>
<description>
<h2 id="added-an-rss-feed-to-sixhatnet"><a class="header" href="#added-an-rss-feed-to-sixhatnet">Added an RSS feed to sixhat.net</a></h2>
<p>I'm a big fan of serving content via RSS feed.
But <a href="https://rust-lang.github.io/mdBook/">MdBook</a> doesn't provide RSS.
It is mainly targeting the production of books.</p>
<p>I decided to script my blog's RSS and with a bit of <a href="https://www.python.org/">python</a> plus a bit of <a href="https://beautiful-soup-4.readthedocs.io/en/latest/">BeautifulSoup</a> I now have a <a href="rss.xml">RSS feed</a>.
It is basically a RSS 2.0 spec and <a href="https://www.rssboard.org/rss-validator/check.cgi?url=https%3A%2F%2Fwww.sixhat.net%2Frss.xml">passes validation</a>.</p>
<p>Some things might brake as the process is still not fully tested and robust.
It doesn't support enclosures for now (no podcasts or rich media in the pipe).
It doesn't have advanced features, has basic html in the content (escaped, not that CTDATA stuff).
It is simple, as <a href="https://en.wikipedia.org/wiki/RSS">RSS</a> should be (Real Simple Syndication).</p>
<p>Ah, if you need an <a href="https://github.com/ckampfe/russ">RSS reader</a>, try RUSS.
I'm loving it.</p>
</description>
<pubDate>Tue, 19 Dec 2023 19:13:41 +0000</pubDate>
<guid>https://www.sixhat.net/2023/added-a-rss-feed.html</guid>
</item>
<item>
<title>arduino tip120 demo code for classes</title>
<link>https://www.sixhat.net/2023/arduino-transistor-test-code.html</link>
<description>
<h2 id="arduino-npn-transistor-test-code-and-demo"><a class="header" href="#arduino-npn-transistor-test-code-and-demo">Arduino NPN Transistor test code and demo</a></h2>
<p>Online demo : <a href="https://www.tinkercad.com/things/fg3gBCshoEu-tip120-demo">https://www.tinkercad.com/things/fg3gBCshoEu-tip120-demo</a></p>
<p><img alt="Arduino demo NPN Transistor" src="https://csg.tinkercad.com/things/fg3gBCshoEu/t725.png?rev=1699959868545000000&amp;s=&amp;v=1&amp;type=circuits"/></p>
<ul>
<li>Arduino</li>
<li>LED strip or 12V Lamp</li>
<li>Power supply 12V</li>
<li>NPN transistor eg: TIP120 ref: <a href="https://www.onsemi.com/pdf/datasheet/tip120-d.pdf">https://www.onsemi.com/pdf/datasheet/tip120-d.pdf</a></li>
<li>Resistor (10kOhm)</li>
</ul>
<pre><code class="language-cpp">void setup() {
pinMode(6, OUTPUT);
}
void loop() {
for (int i = 0; i &lt; 256; i++) {
analogWrite(6, i);
delay(10);
}
for (int i = 0; i &lt; 256; i++) {
analogWrite(6, 255 - i);
delay(10);
}
delay(500);
for (int i = 0; i &lt; 10; i++) {
digitalWrite(6, HIGH);
delay(100);
digitalWrite(6, LOW);
delay(200);
}
delay(500);digitalWrite(6, HIGH);
delay(100);
}
</code></pre>
<html><link href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/default.min.css" rel="stylesheet"/>
<script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/highlight.min.js"></script>
<script>hljs.highlightAll();</script></html>
</description>
<pubDate>Tue, 19 Dec 2023 11:16:19 +0000</pubDate>
<guid>https://www.sixhat.net/2023/arduino-transistor-test-code.html</guid>
</item>
<item>
<title>veo, i veo google</title>
<link>https://www.sixhat.net/2024/05-and-google-has-video-too.html</link>
<description>
<h1 id="and-google-has-video-too"><a class="header" href="#and-google-has-video-too">And Google has video too.</a></h1>
<p>One day after GPT4o (that I still didn't manage to try, c'mon OpenAI) Google came with a bunch of its own models:</p>
<p>— Veo is their Text 2 Video AI model, and.... not available outside the US. Many aren't. I could use a VPN, but why bother. </p>
<p>— And Google is also building on top of Gemini with lighter products, <a href="https://www.theverge.com/2024/5/14/24155382/google-gemini-ai-chrome-nano-io?ref=sixhat.net">Nano</a>? Flash? </p>
<p>— And integrating AI directly in their android gmail. Can we still have anything non-AI?</p>
<p>— Search? Pagerank? Something of a footnote. AI will be the future, <a href="https://www.theverge.com/2024/5/14/24155321/google-search-ai-results-page-gemini-overview?ref=sixhat.net">says big G!</a>.</p>
<p>From this round of announcements it is becoming clear that we are entering an age of diminishing returns on the top models and companies are starting to search for side products to capitalize on. </p>
</description>
<pubDate>Tue, 14 May 2024 20:43:14 +0000</pubDate>
<guid>https://www.sixhat.net/2024/05-and-google-has-video-too.html</guid>
</item>
<item>
<title>Beamer slides with markdown and pandoc</title>
<link>https://www.sixhat.net/2024/11-beamer-slides-with-markdown-and-pandoc.html</link>
<description>
<h1 id="beamer-slides-with-markdown-and-pandoc"><a class="header" href="#beamer-slides-with-markdown-and-pandoc">Beamer slides with markdown and pandoc</a></h1>
<p>When it comes to producing quality prints nothing compares to Latex and I use it extensively throughout my working needs. In recent years I've moved from the full Latex syntax to the <strong>quick and dirty</strong> markdown for my source files. At least for my initial drafts. Then I convert either to TEX or directly to the final PDF format with pandoc<sup class="footnote-reference"><a href="#pandoc">1</a></sup>.</p>
<p>Pandoc is a swiss-army knife for the conversion between different document formats. It is capable of handling more formats than those one can imagine (or need). Also, it has very sensible defaults in terms of templates. It gives you full control over them if you need so. There are a ton of advanced features. Usually converting something is just a matter of issuing the command.</p>
<pre><code class="language-bash">pandoc -o output.html intupt.txt
</code></pre>
<p>For producing slides I have used marp<sup class="footnote-reference"><a href="#marp">2</a></sup> for most of my markdown to pdf conversions in the past, but its dependency on node makes it a bit brittle and forces me into repair mode every now and then.</p>
<p>Therefore, I have a pandoc pipeline that produces great slides and usually doesn't require much maintenance. Also, pandoc allows for extra functionality that you don't get with marp, like automatic tocs, sectioning and bibliographies.</p>
<p>I usually use something like this command to produce my slides. </p>
<pre><code class="language-bash">pandoc -t beamer -f markdown -o output.pdf --pdf-engine xelatex --highlight-style tango input.md
</code></pre>
<p>Note that I usually use xelatex (or sometimes luatex) mainly to support UTF8 in input files directly.</p>
<p>I usually have this command in a script file that makes it quicker to run on any markdown file I need.</p>
<pre><code class="language-bash">#!/bin/bash
set -e
for fich; do :; done
set -x
pandoc -t beamer -f markdown+implicit_figures -o "$fich".pdf --pdf-engine xelatex --highlight-style tango $*
open "$fich".pdf
</code></pre>
<p>Note that this uses a loop to get the last element of a input string into the variable <code>$fich</code> and then passes the list of inputs to the pandoc command with <code>$*</code>. This allows me to pass additional pandoc settings for a particular rendering, for example changing paper size, or styling.</p>
<p>Finally one last thing usually want when editing my slides is to rerun this script everytime I press save. For this I use entr<sup class="footnote-reference"><a href="#entr">3</a></sup> with the following command (md2beamer.sh is the script name):</p>
<pre><code class="language-bash">ls input.md | entr -r md2beamer.sh input.md
</code></pre>
<p>That's it. Simple MD to Beamer slides without complications. Sytling and colorthemes (among other things like titles, authors and institution) can be controlled by the options in the markdown yaml header.</p>
<h3 id="references"><a class="header" href="#references">References</a></h3>
<div class="footnote-definition" id="pandoc"><sup class="footnote-definition-label">1</sup>
<p>Pandoc <a href="https://pandoc.org/">https://pandoc.org/</a></p>
</div>
<div class="footnote-definition" id="marp"><sup class="footnote-definition-label">2</sup>
<p>Marp <a href="https://github.com/marp-team/marp-cli">https://github.com/marp-team/marp-cli</a></p>
</div>
<div class="footnote-definition" id="entr"><sup class="footnote-definition-label">3</sup>
<p>Entr <a href="https://github.com/eradman/entr">https://github.com/eradman/entr</a></p>
</div>
</description>
<pubDate>Tue, 12 Nov 2024 14:13:00 +0000</pubDate>
<guid>https://www.sixhat.net/2024/11-beamer-slides-with-markdown-and-pandoc.html</guid>
</item>
<item>
<title>Autopoietic Networks</title>
<link>https://www.sixhat.net/2024/11-autopoietic-networks.html</link>
<description>
<h1 id="autopoietic-networks"><a class="header" href="#autopoietic-networks">Autopoietic Networks</a></h1>
<ul>
<li><a href="https://gbragafibra.github.io/2024/10/08/autopoietic_nets.html">Autopoietic Networks</a>, to read with a little bit more focused when I have the time. Autopoiesis is a very interesting topic that leads to many research directions, but mainly concerns "how a system can build and maintain itself". There's been many models and I'm always fascinated by these toy models as they allow the exploration of the foundations of systems. </li>
</ul>
</description>
<pubDate>Thu, 21 Nov 2024 22:57:59 +0000</pubDate>
<guid>https://www.sixhat.net/2024/11-autopoietic-networks.html</guid>
</item>
<item>
<title>refactoring old code scares me</title>
<link>https://www.sixhat.net/2024/03-refactoring-old-code-scares-me.html</link>
<description>
<h2 id="refactoring-old-code-scares-me"><a class="header" href="#refactoring-old-code-scares-me">Refactoring old code scares me.</a></h2>
<p>I've two legacy projects that I need to tackle soon (meaning 2024).</p>
<p>One is a Python 2.7 based code.
It needs to move to python 3.x and along the way refactor some cumbersome aspects of the implementation.
This is elder of the two legacy code bases (2012). </p>
<p>The other is not much younger (2014), but is in Java and while I'll need this first, it is very scary to touch.
The code base isn't nice as in "engineering nice" and has parts that were coded by a 1970s C programmer.</p>
<p>It will really hurt my eyes to go trough this code.
Are there any automated tools that help one do this?
Can copilot help?
I'll learn the hard way.</p>
</description>
<pubDate>Thu, 14 Mar 2024 13:32:19 +0000</pubDate>
<guid>https://www.sixhat.net/2024/03-refactoring-old-code-scares-me.html</guid>
</item>
<item>
<title>tree</title>
<link>https://www.sixhat.net/bonsai.html</link>
<description>
<h1 id="bonsai"><a class="header" href="#bonsai">Bonsai</a></h1>
<h3 id="2024-julho"><a class="header" href="#2024-julho">2024 Julho</a></h3>
<p><img alt="" src="assets/bonsai-240711.jpeg"/></p>
<h3 id="2024-april"><a class="header" href="#2024-april">2024 April</a></h3>
<p><img alt="" src="assets/bonsai-240428.png"/></p>
<p><img alt="" src="assets/bonsai-240427.png"/></p>
<h3 id="2019-april"><a class="header" href="#2019-april">2019 April</a></h3>
<p><img alt="" src="assets/bonsai-190406.png"/></p>
<style>
img {
max-width: 100%;
max-height: 80svh;
margin: auto;
display: block;
}
</style>
</description>
<pubDate>Thu, 11 Jul 2024 12:34:29 +0000</pubDate>
<guid>https://www.sixhat.net/bonsai.html</guid>
</item>
<item>
<title>things I'm (re)learning as i play advent of code 2023</title>
<link>https://www.sixhat.net/2023/things-i-am-relearning-in-python-while-playing-aoc.html</link>
<description>
<h2 id="things-im-relearning-as-i-play-advent-of-code-2023"><a class="header" href="#things-im-relearning-as-i-play-advent-of-code-2023">Things I'm (re)learning as I play Advent of Code 2023.</a></h2>
<p>One of the best coding challenges out there is <a href="https://adventofcode.com/">Advent of Code</a>.
It is a really fun and engaging way to learn a new language or to revisit something that you've been away for so long.
In my case it was python.
Python was the language I wrote my PdD in, but at the time I was using version 2.7.
I barely touched python since.
This year I decided to do <strong>AoC</strong> with python 3.11 and wow, what a difference.
I love this language.
It is so pleasurable to work with.</p>
<p>I'm keeping this entry about the things I'm (re)learning<sup class="footnote-reference"><a href="#day">1</a></sup>.
Thing that either I didn't use before or were still absent in 2.7 days.
These things emerged as I progressed through the challenges and compare my code with others<sup class="footnote-reference"><a href="#github">2</a></sup>.</p>
<p>Around day 18—20 of AoC I <em>stopped</em> participating.
Challenges became more specific about using certain advanced algorithm or data structure and became harder.
As a consequence the time devoted to solving increased.
The issue is that I don't have that time, but in any case I'm still going through solutions.
The objective is to learn and recall the most of python possible.
And in the end it is all time management.</p>
<h3 id="tips-learnt"><a class="header" href="#tips-learnt">Tips learnt</a></h3>
<ul>
<li><code>enumerate</code> is a very useful command.</li>
<li><code>zip</code> is also very useful.</li>
<li>Prepare the input <em>data</em> by chaining everything with appropriate <strong>open</strong>..<strong>read</strong>..<strong>strip</strong>..<strong>split</strong>.
But sometimes <strong>numpy</strong> is easier with a simple <code>loadtxt</code>.</li>
<li><a href="https://jupyter.org/">Jupyter Labs</a> has come a long way.
I never liked it back then. Now either via the browser or in VSCode, interactive python is a breeze.</li>
<li>Many times <strong>list|dict|tuple</strong> comprehensions make the code very clean and easy to read.
I'm thinking twice about the <code>for</code> loops that I'm trying to implement.</li>
<li><a href="https://docs.python.org/3/library/typing.html">Type annotations</a> are very useful, even for small scripts, as they force you to decide on what you want to write.</li>
<li><code>*[]</code> and <code>[::1]</code></li>
<li>Many things are easy if I leverage <code>numpy, re, math</code> and don't try to reinvent the algorithms.</li>
<li><code>functools.cmp_to_key</code> works but I still find it a bit cumbersome to use.
There should be a cleaner way to sort by user-defined comparators.</li>
<li><code>numpy.where(condition)</code> in 2D matrices gives two arrays with the indicies of the entries.
Better use <code>numpy.argwhere(condition)</code> and <code>.tolist()</code>. The former case needs a lot of massaging to get a tuple of coordinates back<sup class="footnote-reference"><a href="#npargwhere">3</a></sup>.</li>
<li><code>scipy.spatial.distance</code> has methods for spatial distance calculations that are very useful, namely <code>pdist</code>, <code>cdist</code>, and <code>squareform</code></li>
<li><code>splitlines</code> can shorten the loading of the inputs... something in the line of <code>open(0).read().strip().splitlines()</code> to read from standard in.</li>
</ul>
<h3 id="things-to-explore-in-the-future"><a class="header" href="#things-to-explore-in-the-future">Things to explore in the future</a></h3>
<ul>
<li>The <a href="https://en.wikipedia.org/wiki/Shoelace_formula">Shoelace formula</a>, gets the area of a simple polygon from its coordinates.
I think I'll need to use this on a urban agent based model I'm running now.
As a follow up to this, the <a href="https://en.wikipedia.org/wiki/Pick%27s_theorem">Pick's theorem</a> is also important.</li>
<li><code>numba</code> This is exciting.
Never touched it before, so I'm eager to look into it when I have a bit more time.</li>
<li>Go deep with <code>numpy</code>.</li>
<li>Day 12 pointed in the direction of <strong>recursion</strong>.
I struggled with this exercise.
Recursion is something that everybody understands until you really have to apply it to a new problem.
Need to revisit this again. Not submitting any answers as I feel I didn't solve it.</li>
</ul>
<p><sup class="footnote-reference"><a href="#day">1</a></sup> And we are only at day 20 at the time of writing. I'll update this as <strong>AoC</strong> progresses.<br/>
<sup class="footnote-reference"><a href="#github">2</a></sup> My <a href="https://github.com/sixhat/adventofcode.com/"><strong>AoC</strong> repository</a> contains my daily solutions in case you want to check my progress.<br/>
<sup class="footnote-reference"><a href="#npargwhere">3</a></sup> <a href="https://numpy.org/doc/stable/reference/generated/numpy.argwhere.html">https://numpy.org/doc/stable/reference/generated/numpy.argwhere.html</a></p>
</description>
<pubDate>Sun, 31 Mar 2024 21:41:55 +0000</pubDate>
<guid>https://www.sixhat.net/2023/things-i-am-relearning-in-python-while-playing-aoc.html</guid>
</item>
<item>
<title>writing slides structure from the topics slide in marp</title>
<link>https://www.sixhat.net/2023/writing-slides-structure-from-the-topics-slide-in-marp.html</link>
<description>
<h2 id="writing-slides-structure-from-the-topics-slide-in-marp"><a class="header" href="#writing-slides-structure-from-the-topics-slide-in-marp">Writing slides structure from the Topics slide in marp</a></h2>
<p>I use marp to produce most of my slides and found it very interesting to layout the structure of the presentation in the <em>Agenda</em>/<em>Topics</em> slide. This slide acts as an organizer for the presentation.</p>
<p>I might write something like</p>
<pre><code>## Topics
- Talk about awk
- Explain why Julia rocks
- Show examples
</code></pre>
<p>But then I have to create the slides, and the following snippet of shell code allows me to get the code for the slides in my presentation.
It takes the <code>file.md</code> and parses the lines starting with '- ' and produces the outline like</p>
<pre><code>---
## Talk about awk
---
## Explain why Julia rocks
---
### Show examples
</code></pre>
<p>This is all accomplished with old faithful <code>awk</code>.</p>
<pre><code class="language-bash">function mdIndex {
awk '/^- / {$1="";print "\n---\n\n##"$0 } / +- /{$1="";print "\n---\n\n###"$0}' $1
}
</code></pre>
<p>Just put it in <code>.zprofile</code> and you can run it like this and get the results in the terminal,</p>
<pre><code class="language-bash">$ mdIndex file.md
</code></pre>
<p>or you can pipe it to <code>pbcopy</code> and then paste the output into your editor of choice.</p>
<pre><code class="language-bash">$ mdIndex file.md | pbcopy
</code></pre>
</description>
<pubDate>Sun, 21 Jan 2024 17:47:48 +0000</pubDate>
<guid>https://www.sixhat.net/2023/writing-slides-structure-from-the-topics-slide-in-marp.html</guid>
</item>
<item>
<title>interesting openstreet use for studying hospital accessibility</title>
<link>https://www.sixhat.net/2023/interesting-openstreet-use-for-studing-hospital-accessibility.html</link>
<description>
<h2 id="interesting-openstreet-use-for-studying-hospital-accessibility"><a class="header" href="#interesting-openstreet-use-for-studying-hospital-accessibility">Interesting Openstreet use for studying Hospital Accessibility</a></h2>
<p><a href="https://wcedmisten.fyi/post/visualizing-hospital-accessibility/">William Edmisten</a> made a great page about how he went on to study Hospital Accessibility based on data from OpenStreet. The read is about 15 min, but the author goes all the details needed to understand the process. Very impressive indeed. </p>
<p>Hm, with all the Hospitals closing in Portugal, should one do something inspired on his work?</p>
</description>
<pubDate>Sun, 14 Jan 2024 17:26:46 +0000</pubDate>
<guid>https://www.sixhat.net/2023/interesting-openstreet-use-for-studing-hospital-accessibility.html</guid>
</item>
<item>
<title>all 2023 posts</title>
<link>https://www.sixhat.net/2023/2023_index.html</link>
<description>
<h3 id="2023"><a class="header" href="#2023">2023</a></h3>
<h4 id="december-2023"><a class="header" href="#december-2023">December 2023</a></h4>
<ul>
<li><a href="./2023/added-a-rss-feed.html">Added a RSS feed</a></li>
<li><a href="./2023/code-snippet-for-highlightjs-in-markdown-pages.html">Code Snippet for Highlight.js in Markdown Pages</a></li>
<li><a href="css-is-this-the-only-css-youll-ever-need.html">Is this the only CSS you'll ever need?</a> - simple css for any web.</li>
<li><a href="things-I-am-relearning-in-python-while-playing-aoc.html">Things I'm (re)learning as I play Advent of Code 2023</a> - programming languages, coding challenge, python</li>
<li><a href="./2023/spellcheck-shell-sripts.html">Spell Checker for Shell Scripts</a> - bash productivity, terminal world, utilities</li>
<li><a href="a-simple-PS1-bash-prompt.html">A simple PS1 Bash Prompt</a> - bash productivity</li>
</ul>
<h4 id="november-2023"><a class="header" href="#november-2023">November 2023</a></h4>
<ul>
<li><a href="epsteins-levels-for-agent-based-models-performance.html">Axtel and Epstein's levels for agent based performance</a> - agent-based models</li>
<li><a href="arduino-transistor-test-code.html">Arduino TIP120 demo code for classes</a> - arduino</li>
<li><a href="blogs-without-server-side-rendering.html">Blogs without server side rendering</a> - internet</li>
<li><a href="./2023/data-science-handbook.html">Data Science Handbook</a> - research</li>
<li><a href="grimm-odd.html">Grimm's ODD standard protocol for describing individual-based and agent based models</a> - agent-based models</li>
<li><a href="models-of-creativity-.html">Models of Creativity</a> - November 10, 2023</li>
<li><a href="ants-are-amazing---what-about-organizations----reading.html">Ants are amazing - what about organizations - reading</a> - November 09, 2023</li>
<li><a href="creating-html-indices-with-tree.html">Creating.md indices with tree</a> - <code>computers</code></li>
</ul>
<h4 id="october-2023"><a class="header" href="#october-2023">October 2023</a></h4>
<ul>
<li><a href="m3-m3-pro-and-m3-max.html">M3, M3 Pro and M3 Max</a> - computers</li>
<li><a href="i-have-to-many-rss-feeds-in-my-reader.html">I have to many RSS feeds in my reader</a> - internet</li>
<li><a href="finder-explorer-nautilus-rox--spacedrive.html">Finder, Explorer, Nautilus, Rox, ... Spacedrive</a> - computers</li>
</ul>
<h4 id="september-2023"><a class="header" href="#september-2023">September 2023</a></h4>
<ul>
<li><a href="marginalia-in-the-modern-digital-world-is-it-possible.html">Marginalia in the modern digital world. Is it possible?</a> — September 28, 2023</li>
<li><a href="computer-related-stuff---how-these-machines-work.html">Computer related stuff---How these machines work</a> — September 28, 2023</li>
<li><a href="disable-macbook-air-autoboot-when-disconnecting-to-power-or-open-lid.html">Disable MacBook Air Autoboot when dis/connecting to power or open lid</a> — September 19, 2023</li>
<li><a href="red-led-at-13-blue-led-at-12---police-lights.html">RED LED at 13, BLUE LED at 12 - Police Lights</a> — September 02, 2023</li>
</ul>
<h4 id="july-2023"><a class="header" href="#july-2023">July 2023</a></h4>
<ul>
<li><a href="the-wei-web-environment-integrity-api-proposal-from-google-is-dangerous-and-should-not-go-forward-.html">The WEI (Web Environment Integrity) API proposal from Google is dangerous and should not go forward.</a> — July 26, 2023</li>
<li><a href="my-approach-to-managing-scratch-projects-with-bash-scripts.html">My approach to managing scratch projects with bash scripts</a> — July 25, 2023</li>
<li><a href="avoid-long-urls-extending-past-the-margins-of-text-in-latex-with-xurl.html">Avoid long urls extending past the margins of text in LaTeX with xurl</a> — July 21, 2023</li>
<li><a href="how-we-learn-and-how-to-organise-a-reading-inbox.html">How we learn and how to organise a Reading Inbox</a> — July 04, 2023</li>
<li><a href="im-back-to-social-networks-and-it-is-mastodon.html">I'm back to social networks, and it is Mastodon.</a> — July 04, 2023</li>
</ul>
<h4 id="june-2023"><a class="header" href="#june-2023">June 2023</a></h4>
<ul>
<li><a href="organising-stuff-is-hard-until-it-isnt.html">Organising stuff is hard until it isn't</a> — June 13, 2023</li>
</ul>
<h4 id="may-2023"><a class="header" href="#may-2023">May 2023</a></h4>
<ul>
<li><a href="writing-slides-structure-from-the-topics-slide-in-marp.html">Writing slides structure from the Topics slide in marp</a> — May 21, 2023</li>
<li><a href="a-simple-css-trick-por-dithered-images.html">A simple CSS trick 4 dithered images</a> — May 18, 2023</li>
<li><a href="readings-on-strange-programming-art-and-electronics.html">Readings on strange programming, art and electronics</a> — May 17, 2023</li>
</ul>
<h4 id="april-2023"><a class="header" href="#april-2023">April 2023</a></h4>
<ul>
<li><a href="gpt4-experiments---sparks-of-agi.html">GPT4 experiments - Sparks of AGI</a> — April 18, 2023</li>
<li><a href="interesting-openstreet-use-for-studing-hospital-accessibility.html">Interesting Openstreet use for studing Hospital Accessibility</a> — April 05, 2023</li>
<li><a href="a-simple-js-range-one-liner.html">A simple JS range one-liner</a> — April 03, 2023</li>
<li><a href="o-fim-das-trotinetes-de-aluguer.html">O fim das trotinetes de aluguer?</a> — April 03, 2023</li>
</ul>
<h4 id="march-2023"><a class="header" href="#march-2023">March 2023</a></h4>
<ul>
<li><a href="still-reading-about-ai-and-gpt-and-whats-next-in-this-space.html">Still reading about AI and GPT and what's next in this space</a> — March 23, 2023</li>
<li><a href="the-ai-races-for-march-22.html">The AI Races for March 22:</a> — March 22, 2023</li>
<li><a href="two-main-developments-in-the-ai-generators-world.html">Two main developments in the AI generators world</a> — March 21, 2023</li>
<li><a href="and-it-goes-dark.html">And it goes dark</a> — March 20, 2023</li>
<li><a href="fuzzy-logic-shell-alias.html">fuzzy logic shell alias</a> — March 20, 2023</li>
<li><a href="svelte-link-dump.html">Svelte link dump.</a> — March 15, 2023</li>
<li><a href="senhor-clemente-que-oportunidade-perdida-para-o-arrependimento.html">Senhor Clemente, que oportunidade perdida para o arrependimento.</a> — March 06, 2023</li>
</ul>
<h4 id="february-2023"><a class="header" href="#february-2023">February 2023</a></h4>
<ul>
<li><a href="toggling-light-bulb-problem.html">Toggling Light Bulb Problem</a> — February 28, 2023</li>
<li><a href="no-arrendamento-quem-se-lixa-e-quem-cumpre-e-ja-aluga.html">No arrendamento, quem se lixa é quem cumpre e já aluga</a> — February 27, 2023</li>
<li><a href="in-the-slow-movement-you-find-great-pearls-of-wisdom.html">In the slow movement you find great pearls of wisdom.</a> — February 16, 2023</li>
</ul>
<h4 id="january-2023"><a class="header" href="#january-2023">January 2023</a></h4>
<ul>
<li><a href="beja-e-alverca.html">Beja e Alverca</a> — January 28, 2023</li>
<li><a href="websites-que-funcionam-apenas-em-modo-texto.html">Websites que funcionam apenas em modo texto</a> — January 27, 2023</li>
<li><a href="small-is-beautiful.html">Small is beautiful</a> — January 27, 2023</li>
<li><a href="tools-for-modern-research.html">Tools for Modern Research</a> — January 24, 2023</li>
<li><a href="and-we-are-in-2023.html">And we are in 2023</a> — January 24, 2023</li>
</ul>
</description>
<pubDate>Sun, 14 Jan 2024 17:25:28 +0000</pubDate>
<guid>https://www.sixhat.net/2023/2023_index.html</guid>
</item>
<item>
<title>Editors, Editors</title>
<link>https://www.sixhat.net/2024/06-editors-editors.html</link>
<description>
<h2 id="editors-editors"><a class="header" href="#editors-editors">Editors, editors</a></h2>
<p>This is a simple memory consumption test I made earlier today. These are the results in my machine while opening a small markdown file. Don't take this in any other direction.</p>
<ul>
<li>ed - 1,281 KB</li>
<li>nano - 2,171 KB</li>
<li>vim - 4,305 KB</li>
<li>joe - 6,449 KB</li>
<li>hx - 8,066 KB</li>
<li>amp - 10 MB</li>
<li>neovim - 17 MB</li>
<li>zed - 179 MB</li>
<li>vscode - 60 MB + 8 Code Helpers at 60 MB each.</li>
</ul>
<p>(It is clear how modern crap is exponentially more hungry for resources, while old stuff still works fine for the same tasks.)</p>
<p>A favorite you ask? All of them. </p>
</description>
<pubDate>Sat, 29 Jun 2024 22:41:05 +0000</pubDate>
<guid>https://www.sixhat.net/2024/06-editors-editors.html</guid>
</item>
<item>
<title>Reading texts on paper beats reading texts on a computer screen</title>
<link>https://www.sixhat.net/2024/09-reading-text-on-paper-beats-reading-on-computer-screens.html</link>
<description>
<h2 id="reading-texts-on-paper-beats-reading-texts-on-a-computer-screen"><a class="header" href="#reading-texts-on-paper-beats-reading-texts-on-a-computer-screen">Reading texts on paper beats reading texts on a computer screen.</a></h2>
<blockquote>
<p>The results of this study indicate that reading linear narrative and expository texts on a computer screen leads to poorer reading comprehension than reading the same texts on paper. These results have several pedagogical implications.</p>
</blockquote>
<p>in <a href="http://clikmedia.ca/LMM/sites/default/files/pdf/mangen_2012_lecture_sur_ecran_lecture_papier_comprehension.pdf">Mangen, A., Walgermo, B. R., &amp; Brønnick, K. (2013). Reading linear texts on paper versus computer screen: Effects on reading comprehension. <em>International journal of educational research</em>, <em>58</em>, 61-68.</a></p>
</description>
<pubDate>Sat, 28 Sep 2024 09:49:15 +0000</pubDate>
<guid>https://www.sixhat.net/2024/09-reading-text-on-paper-beats-reading-on-computer-screens.html</guid>
</item>
<item>
<title>the wei (web environment integrity) api proposal from google is dangerous and should not go forward.</title>
<link>https://www.sixhat.net/2023/the-wei-web-environment-integrity-api-proposal-from-google-is-dangerous-and-should-not-go-forward-.html</link>
<description>
<h2 id="the-wei-web-environment-integrity-api-proposal-from-google-is-dangerous-and-should-not-go-forward"><a class="header" href="#the-wei-web-environment-integrity-api-proposal-from-google-is-dangerous-and-should-not-go-forward">the wei (web environment integrity) api proposal from google is dangerous and should not go forward.</a></h2>
<p>wei will allow the webpage provider to arbitrary block you if you're not using a certain browser, for example, via what is called an <em>attester</em> that will certify that you (the receiver, consumer of content) are trustworthy. </p>
<p>the idea is that wei will force you, the client, to consume the content in the way they envisioned. this means no adblockers, no text-based browsers, no plugins that change the content you consume in any way. no non-human reading of the website. this basically means you'll hand control over your computer environment or else you'll get no service. </p>
<p>in any case google is already pushing code to their development github of chrome meaning that they will push the proposal with, or without, the approval of the w3c (the consortium managing web standards). they will enforce it on top of having 65% market share and probably forcing every google service to go through chrome (this means gmail, ads, youtube, etc...). </p>
<p>if you want to read about this potential danger go here:</p>
<ul>
<li><a href="https://github.com/RupertBenWiser/Web-Environment-Integrity/blob/main/explainer.md#example-use-cases">https://github.com/RupertBenWiser/Web-Environment-Integrity/blob/main/explainer.md#example-use-cases</a></li>
<li><a href="https://vivaldi.com/blog/googles-new-dangerous-web-environment-integrity-spec/">https://vivaldi.com/blog/googles-new-dangerous-web-environment-integrity-spec/</a></li>
<li><a href="https://github.com/mozilla/standards-positions/issues/852">https://github.com/mozilla/standards-positions/issues/852</a></li>
</ul>
<p>in any case my suggestion is to <strong><a href="https://www.mozilla.org/en-GB/firefox/new/">move to firefox</a></strong> and help diminish chrome share. No browser should have such power and as Chrome is becoming the new IE you should give the alternatives a try.</p>
</description>
<pubDate>Sat, 20 Jan 2024 11:15:44 +0000</pubDate>
<guid>https://www.sixhat.net/2023/the-wei-web-environment-integrity-api-proposal-from-google-is-dangerous-and-should-not-go-forward-.html</guid>
</item>
<item>
<title>styles for clean bw slides in marp</title>
<link>https://www.sixhat.net/2024/styles-for-clean-bw-slides-in-marp.html</link>
<description>
<h2 id="styles-for-clean-bw-slides-in-marp"><a class="header" href="#styles-for-clean-bw-slides-in-marp">styles for clean bw slides in marp</a></h2>
<pre><code class="language-html">&lt;style&gt;
* {
color: black!important;
font-family: "ibm plex serif", serif;
background: white;
}
pre {
border: 2px solid;
background: white;
border-radius: 0.3rem;
}
blockquote {
border: 2px dashed;
margin: 0 3em;
padding: 0.3em 0.6em;
border-radius: 0.3rem;
}
h1, h2, h3 {
margin-top: 0;
}
section {
justify-content: flex-start;
}
code {
background: white;
}
ul {
list-style-type: '⁍ ';
}
section.title {
justify-content: center;
}
section.f-2 {font-size: 2em;}
section.f-3 {font-size: 3em;}
section.f-4 {font-size: 4em;}
section.f-5 {font-size: 5em;}
section.f-6 {font-size: 6em;}
section.f-8 {font-size: 8em;}
section.f-10 {font-size: 10em;}
section.f-14 {font-size: 14em;}
section.f-20 {font-size: 20em;}
section.center {text-align: center;}
&lt;/style&gt;
</code></pre>
</description>
<pubDate>Sat, 20 Jan 2024 11:01:15 +0000</pubDate>
<guid>https://www.sixhat.net/2024/styles-for-clean-bw-slides-in-marp.html</guid>
</item>
<item>
<title>a simple ps1 bash prompt</title>
<link>https://www.sixhat.net/2023/a-simple-ps1-bash-prompt.html</link>
<description>
<h2 id="a-simple-ps1-bash-prompt-and-one-for-zsh-too"><a class="header" href="#a-simple-ps1-bash-prompt-and-one-for-zsh-too">a simple ps1 bash prompt (and one for zsh too)</a></h2>
<pre><code class="language-shell">export PS1='\n($?, $(cat /proc/loadavg | colrm 5 | xargs), $(du -hs | colrm 8 | xargs), $(ls -1 | wc -l)) \u@\H:\w/\n\$ '
</code></pre>
<p>this outputs something like</p>
<pre><code class="language-shell">(0, 0.85, 6.8g, 13) david@sixhat.net:~/
$
</code></pre>
<p>where the individual values are (in order):</p>
<ul>
<li>the <strong>last command</strong> exit code <code>$?</code></li>
<li>the <strong>load</strong> on the cpu <code>$(cat /proc/loadavg | colrm 5 | xargs)</code></li>
<li>the <strong>size</strong> of current directory <code>$(du -hs | colrm 8 | xargs)</code></li>
<li>the <strong>number</strong> of files/dirs in current directory <code>$(ls -1 | wc -l)</code></li>
<li>the <strong>user@host:folder</strong> <code>\u@\h:\w/</code></li>
<li>the <strong>prompt type</strong> (normal $ |root #) <code>\$</code></li>
</ul>
<p>notice that it has no colours. although colours are a nice feature to have, they don't work in every terminal, but within your <code>.profile</code> you can do conditional prompts based on colour support by the terminal.</p>
<h4 id="what-about-zsh-prompts"><a class="header" href="#what-about-zsh-prompts">what about zsh prompts?</a></h4>
<p>well, zsh prompts are a little bit different to configure, but a similar one that i use in my macs is this:</p>
<pre><code class="language-shell">export PS1='
%T %(1d.%B[First Day of Month]%b .)%U(%?,%j)%u %n@%m %F{red}%0~%f
%F{green}%B%(!.%#.;)%b%f '
</code></pre>
<p>The expansions are different from bash, so you are better off reading the <a href="https://zsh.sourceforge.io/Doc/Release/Prompt-Expansion.html#Prompt-Expansion">manual about expansions</a>, but the above one can give you a nice start.</p>
</description>
<pubDate>Sat, 20 Jan 2024 10:59:32 +0000</pubDate>
<guid>https://www.sixhat.net/2023/a-simple-ps1-bash-prompt.html</guid>
</item>
<item>
<title>teaching physical computing notes</title>
<link>https://www.sixhat.net/2024/04-teching-physical-computing-notes.html</link>
<description>
<h2 id="teaching-physical-computing-notes"><a class="header" href="#teaching-physical-computing-notes">Teaching Physical Computing Notes</a></h2>
<p>Students mostly find it difficult to understand the electronics of the setups.</p>
<blockquote>
<p>The wiring, we can't understand the wiring.</p>
</blockquote>
<p>The other thing they usually complain about is the syntax of coding in C.
These are Design students that find mathematics and logic very hard, and have close to zero experience with programming.
Semicolons are a nightmare.
Matching brackets is a mystery.</p>
<p>How can one teach them in a way that eliminate these problems?</p>
<ul>
<li>About the C syntax problem, maybe using <a href="https://circuitpython.org/">CircuitPython</a> could solve this and also the solve the cycle code-compile-debug-repeat. But this would not work with Uno R3 based on ATmega chips.</li>
<li>Use the Firmata / Pyfirmata combo. This doesn't produce independent devices as they will be tethered to a computer.</li>
<li>On the wiring confusing that are their setups, the only easy way to teach them is to move to a groove modules based teaching. This would simplify the connections avoiding mistakes and some of their shortcomings.</li>
<li>Use XOD (visual programming ala Pd.)</li>
<li>Use Snap4Arduino (a Scratch type language)</li>
</ul>
<p>Do you have any experience with this? Hit me on <a href="https://datasci.social/@sixhat">discord if you do</a>.</p>
</description>
<pubDate>Sat, 20 Apr 2024 08:32:06 +0000</pubDate>
<guid>https://www.sixhat.net/2024/04-teching-physical-computing-notes.html</guid>
</item>
<item>
<title>HTMX resources</title>
<link>https://www.sixhat.net/2024/11-htmx-resources.html</link>
<description>
<h1 id="htmx-resources-and-ecosystem"><a class="header" href="#htmx-resources-and-ecosystem">HTMX resources and ecosystem</a></h1>
<ul>
<li><a href="https://htmx.org/">https://htmx.org/</a>, well... the main website for HTMX</li>
<li><a href="https://hypermedia.systems/">https://hypermedia.systems/</a>, The book on "<em>a simpler approach to building web applications on the web and beyond with HTMX and Hyperview</em>"</li>
<li><a href="https://hyperview.org/">https://hyperview.org/</a>, mobile apps? got it.</li>
<li><a href="https://david.guillot.me/en/posts/tech/following-up-mother-of-all-htmx-demos/">Following up "Mother of all htmx demos"</a></li>
<li><a href="https://hyperscript.org/">https://hyperscript.org/</a></li>
</ul>
</description>
<pubDate>Sat, 09 Nov 2024 18:44:56 +0000</pubDate>
<guid>https://www.sixhat.net/2024/11-htmx-resources.html</guid>
</item>
<item>
<title>quem não lê é menos livre</title>
<link>https://www.sixhat.net/2024/03-quem-nao-le-e-menos-livre.html</link>
<description>
<h2 id="quem-não-lê-é-menos-livre"><a class="header" href="#quem-não-lê-é-menos-livre">"Quem não lê é menos livre"</a></h2>
<blockquote>
<p>É que quem não lê, não fala e não escreve decentemente, é menos livre, mais facilmente manipulado, menos ecaz em coisa alguma importante, mais fácil de ser mandado e de não mandar nem em si próprio.<br/>
—Pacheco Pereira no <a href="https://www.publico.pt/2024/03/09/opiniao/opiniao/ler-conta-liberdade-2083066?ref=sixhat.net">Público</a> em dia de reflexão.</p>
</blockquote>
<p>Ainda bem que há quem seja capaz de chamar as coisas pelos nomes.</p>
<p>E diz ainda </p>
<blockquote>
<p>(...) quem defende esta nova forma de ignorância agressiva e de deslumbramento tecnológico não são os novos ignorantes, mas os antigos ignorantes, a quem as redes sociais dão uma ilusão de igualdade e uma presunção de saber (...)</p>
</blockquote>
<p>o termo <code>ignorância agressiva</code> sintetiza claramente os tempos actuais.</p>
</description>
<pubDate>Sat, 09 Mar 2024 17:58:42 +0000</pubDate>
<guid>https://www.sixhat.net/2024/03-quem-nao-le-e-menos-livre.html</guid>
</item>
<item>
<title>haiper video generation</title>
<link>https://www.sixhat.net/2024/03-haiper-after-sora.html</link>
<description>
<h2 id="after-sora-theres-a-haiper-reality"><a class="header" href="#after-sora-theres-a-haiper-reality">After Sora theres a Haiper Reality</a></h2>
<p>Prompt to Vídeo is becoming a fantastic sub-field of all the generative AI world.
The latest entry is <a href="https://haiper.ai/?ref=sixhat.net">Haiper</a>.
This company promises to bring the fight to OpenAI and contrary to ~~Open~~CloseAI, they allow you to try the generation by yourself—although with some limits (2s, lower res).</p>
</description>
<pubDate>Sat, 09 Mar 2024 16:15:21 +0000</pubDate>
<guid>https://www.sixhat.net/2024/03-haiper-after-sora.html</guid>
</item>
<item>
<title>array based languages</title>
<link>https://www.sixhat.net/2024/array-based-languages.html</link>
<description>
<h2 id="array-based-languages"><a class="header" href="#array-based-languages">array based languages</a></h2>
<p>recently i came across <a href="https://www.uiua.org/?ref=sixhat.net">uiua</a>, a rust based <em>array-oriented language</em> that i fell in love immediately because of the aesthetics of the code.
see the following online pad:</p>
<html><iframe frameborder="0" height="740px" src="https://www.uiua.org/pad?src=0_7_1__IyBTdW0gdGhlIG51bWJlcnMgYmlnZ2VyIHRoYW4gNTAgCiMgaW4gdGhlIHJhbmdlIFswLDEwMCkKLyvDlz41MC7ih6ExMDAK" width="100%"></iframe></html>
<p>it computes the sum of the numbers in the range <code>[0,100)</code> that are greater than 50.
although at first it looks a bit confusing, the use of unicode symbols makes is very pretty, in my eyes.</p>
<p>this is very different from <a href="https://www.jsoftware.com/?ref=sixhat.net">j</a>, that only uses ascii characters and makes reading the code a bit harder—just my opinion.
this doesn't take merit from j, which has many features and is more mature, but aesthetics count.</p>
<p>code in `uiua`` reads left to right (ltr) and operators are to the left of their operands, making it resemble lisp like prefix notation.
but it executes things in right to left (rtl).
the authors even have a <a href="https://www.uiua.org/docs/rtl?ref=sixhat.net">faq entry about rtl</a>.</p>
<p>i'm not an expert in these kinds of languages, but i'm going to look more into them during 2024.
i'm tired of imperative programming languages and want to expand into different territories.</p>
<p>i'm considering being drastic about it.
force myself to write everything i need during 2024 in any of the following:</p>
<ul>
<li>lisp style lang (common lisp, scheme, racket, clojure, ...) – yes, i don't mind the parenthesis.</li>
<li>array based languages (uiua, j, bqn, ...)</li>
</ul>
<p>and if something fails me i'll use my loved python, but only in despair.</p>
</description>
<pubDate>Sat, 09 Mar 2024 16:15:10 +0000</pubDate>
<guid>https://www.sixhat.net/2024/array-based-languages.html</guid>
</item>
<item>
<title>where has my disk space gone?</title>
<link>https://www.sixhat.net/2024/02-where-has-my-disk-space-gone.html</link>
<description>
<h1 id="where-has-my-disk-space-gone"><a class="header" href="#where-has-my-disk-space-gone">where has my disk space gone?</a></h1>
<p>I can't really come around in my mind that modern computers are so bloated with stuff (99% unnecessary) that free space just disappears before your eyes at alarming rates.</p>
<p>For many yeas I've been using some sort of treemapping app to visualize where has everything gone.</p>
<p>My favorite is <a href="https://www.omnigroup.com/more/?ref=sixhat.net">OmniDiskSweepper</a>. Takes little space, does its thing, allows you to act upon findings. What else would one need?</p>
</description>
<pubDate>Sat, 09 Mar 2024 16:14:17 +0000</pubDate>
<guid>https://www.sixhat.net/2024/02-where-has-my-disk-space-gone.html</guid>
</item>
<item>
<title>stealing this one, on how to study</title>
<link>https://www.sixhat.net/2024/02-shamelessly-stealing-this.html</link>
<description>
<h1 id="im-shamelessly-stealing-from-this-reference"><a class="header" href="#im-shamelessly-stealing-from-this-reference">I'm shamelessly stealing from this reference</a></h1>
<p>while teaching I need sometimes to point students to good studying practices.
I just found a great gem.
I'm going to steel this one. Professor <a href="https://cse.buffalo.edu/~rapaport/howtostudy.html?ref=sixhat.net">Rapaport</a> wrote a long page on how to study efficiently and I agree 100% with the content.
It is really funny, useful and filled with some gems (i'm recursing gems in gem in gem in gem... ok... who cares, my text)
Next time I get a student struggling with getting results because of bad habits, this will at the top of my list of recommendations.</p>
</description>
<pubDate>Sat, 09 Mar 2024 16:13:59 +0000</pubDate>
<guid>https://www.sixhat.net/2024/02-shamelessly-stealing-this.html</guid>
</item>
<item>
<title>AI Aesthetics</title>
<link>https://www.sixhat.net/2024/06-ai-aesthetics.html</link>
<description>
<h3 id="ai-aesthetics"><a class="header" href="#ai-aesthetics">AI aesthetics</a></h3>
<p>The latest Apple Developer Conference invitation succumbed to the Neon AI Aesthetic. Neon is everywhere, we cannot really get rid of it, and my eyes are hurting. Generative AI so median quality. So average. Yes with lot's of work into being that, just that, always that. There's no sense of beauty in what generative AI produces.</p>
</description>
<pubDate>Sat, 08 Jun 2024 22:02:40 +0000</pubDate>
<guid>https://www.sixhat.net/2024/06-ai-aesthetics.html</guid>
</item>
<item>
<title>Random Number Generator, Anyone</title>
<link>https://www.sixhat.net/2024/06-random-number-generator-anyone.html</link>
<description>
<h3 id="random-number-generator-anyone"><a class="header" href="#random-number-generator-anyone">Random number generator, anyone</a></h3>
<p>Nice trick from <a href="https://codeandbitters.com/main-as-usize/?ref=sixhat.net">codeandbitters</a>:</p>
<pre><pre class="playground"><code class="language-rust">fn main(){
let rand = main as usize;
dbg!(rand);
}</code></pre></pre>
</description>
<pubDate>Sat, 08 Jun 2024 22:01:18 +0000</pubDate>
<guid>https://www.sixhat.net/2024/06-random-number-generator-anyone.html</guid>
</item>
<item>
<title>qualcomm ai hub</title>
<link>https://www.sixhat.net/2024/03-qualcomm-ai-hub.html</link>
<description>
<h2 id="qualcomm-ai-hub"><a class="header" href="#qualcomm-ai-hub">Qualcomm AI HUB</a></h2>
<p>Qualcomm just launched an AI Store (<a href="https://aihub.qualcomm.com/?ref=sixhat.net">AI HUB</a>) with pre-trained models for different tasks, including:</p>
<ul>
<li>Segmentation</li>
<li>Stable-diffusion</li>
<li>Whisper-base (Automatic Speech Recognition)</li>
<li>TrOCR (transformer based OCR)</li>
<li>Face Detection</li>
<li>YOLO-Detection (object detection)</li>
<li>Open AI clip</li>
</ul>
</description>
<pubDate>Sat, 08 Jun 2024 21:36:38 +0000</pubDate>
<guid>https://www.sixhat.net/2024/03-qualcomm-ai-hub.html</guid>
</item>
<item>
<title>making checklists and ai</title>
<link>https://www.sixhat.net/2024/04-making-checklists-and-ai.html</link>
<description>
<h2 id="making-checklists-and-ai"><a class="header" href="#making-checklists-and-ai">Making checklists and AI</a></h2>
<p><em>(a developing thought)</em></p>
<p>I'm reading <a href="https://www.amazon.com/Power-Prediction-Disruptive-Artificial-Intelligence/dp/1647824192">Power and Prediction</a> and at a certain point the authors confront rule based systems with AI as if one is bad/old/démodé and the other is better. You can easily guess which is which. One example provided is the use of checklists. The authors argue that, eventually, AI could replace checklist, but forget that a checklist is made, read and executed by humans, and this process adds value to that person in the form of knowledge about the system. As a process with a required effort, it allows for a reflexive subprocess where the person goes through items and assigns a cognitive load to each task. If AI merely transforms the process in a guidance mechanism that at any given point feeds the user a task to process, then the AI system, although giving the illusion of productivity enhancer is in fact forcing a bigger loss. That loss is the value of persons free will in the decision process. If AI is just a one item checklist generator then AI aim will be to replace the human entirely as soon as the task is accomplished by a companion roboticised system. </p>
</description>
<pubDate>Sat, 06 Apr 2024 18:40:02 +0000</pubDate>
<guid>https://www.sixhat.net/2024/04-making-checklists-and-ai.html</guid>
</item>
<item>
<title>Apple buys Pixelmator</title>
<link>https://www.sixhat.net/2024/11-apple-buys-pixelmator.html</link>
<description>
<h1 id="apple-buys-pixelmator"><a class="header" href="#apple-buys-pixelmator">Apple buys Pixelmator</a></h1>
<p>I was big a fan of Pixelmator when they were just starting out. A simple Photoshop replacement, without being a mammoth of a download. Unfortunately the software never really took off. It worked (almost always, emphasis on 'almost'). This was to be the photo editor with advanced features without the fat of PS. Never really got there.</p>
<p>Eventually I moved on and went with <a href="https://krita.org/en/">Krita</a>, an open-source bitmap editor that fulfills the same role as Pixelmator (and eventually does it better, some might argue)</p>
<p>This week Apple bought Pixelmator. The question is for what? What is Apple going to do do with it? Absorve the technologies into iCloud Photos? Launch a new bitmap editor of its own that is a rebranded and improved version of Pixelmator. Keep Pixelmator for a couple of years until it gets the axe? I don't know but my feeling is that Pixelmator is going the end up like Atom when Microsoft bought it to include some of its tech in VSCode and then effectively killed it.</p>
</description>
<pubDate>Sat, 02 Nov 2024 11:55:52 +0000</pubDate>
<guid>https://www.sixhat.net/2024/11-apple-buys-pixelmator.html</guid>
</item>
<item>
<title>Working on a blog generator in bash</title>
<link>https://www.sixhat.net/2024/11-working-on-a-blog-generator-in-bash.html</link>
<description>
<h1 id="working-on-a-blog-generator-in-bash"><a class="header" href="#working-on-a-blog-generator-in-bash">Working on a blog generator in bash</a></h1>
<p>I've been working on a blog generator in bash. I've meant to do it before, but
work has been overwhelming. It is named BLiG, for now.</p>
<p>The main features I want are:</p>
<ul>
<li>simple bash script</li>
<li>simple structure</li>
<li>markdown</li>
<li>separate writing and editing from rendering allowing anyone to integrate
their own generator (just edit the "render" function to suit your needs)</li>
</ul>
<p>The principle is that any folder can be a blog. The only requirement is
that posts are organized in an yearly folder inside a <strong>src</strong> folder (eg
<code>src/2024/some-title.md</code>). The output will depend on the software used for
the static generation. In my case I use
<a href="https://rust-lang.github.io/mdBook/">mdbook</a>, so my render function just has
<code>mdbook build</code> and <code>mdbook serve</code> commands.</p>
<p>It is still very alpha but if you want to try it out please <a href="https://github.com/sixhat/swipe/tree/master/bash/blig">check it out</a></p>
</description>
<pubDate>Sat, 02 Nov 2024 11:54:39 +0000</pubDate>
<guid>https://www.sixhat.net/2024/11-working-on-a-blog-generator-in-bash.html</guid>
</item>
<item>
<title>I'm giving Zed a try</title>
<link>https://www.sixhat.net/2024/02-im-giving-zed-a-try.html</link>
<description>
<h2 id="im-giving-zed-a-try"><a class="header" href="#im-giving-zed-a-try">I'm giving Zed a try.</a></h2>
<pre><code>- Whose program is this?
- It's editor, baby.
- Whose editor is this?
- It's VSCode's.
- Who's VSCode?
- VSCode's dead baby, VSCode's dead.
</code></pre>
<p>In this case, the dead one might be VSCode.</p>
<p>I've been a <strong>vim/neovim/vscode</strong> user for a couple of yeas now.
In all this time the amount of bloat VSCode has installed as grown to a point that I find myself thinking why I keep dragging this anchor of an editor.</p>
<p>It can't even delete extensions from the support folder when I uninstall them.</p>
<p>Then there's all those <code>node_modules</code> folders filled with rogue JS for each plugin I use (that vscode never deletes).</p>
<p>So I decided to try <a href="https://zed.dev/?ref=sixhat.net">Zed</a>.</p>
<p>It comes from the creators of Atom.
After <a href="https://zed.dev/blog/we-have-to-start-over?ref=sixhat.net">repenting</a> they went with <a href="https://www.rust-lang.org/?ref=sixhat.net">rust</a> for Zed.</p>
<p>If the last two facts weren't enough to get my attention, the <a href="https://github.com/zed-industries/zed-fonts?ref=sixhat.net">Zed mono and Zed sans fonts</a> used in the editor got me 100%.</p>
<p>And for what I use a text editor for, zed is just blazing fast.
What a surprise. Happy times in 2024!</p>
</description>
<pubDate>Sat, 02 Nov 2024 00:11:46 +0000</pubDate>
<guid>https://www.sixhat.net/2024/02-im-giving-zed-a-try.html</guid>
</item>
<item>
<title>toggling light bulb problem</title>
<link>https://www.sixhat.net/2023/toggling-light-bulb-problem.html</link>
<description>
<h2 id="toggling-light-bulb-problem"><a class="header" href="#toggling-light-bulb-problem">Toggling Light Bulb Problem</a></h2>
<p>The video published by numberphile a few days ago about the <a href="https://www.youtube.com/watch?v=-UBDRX6bk-A">Toggling Light Bulb Problem</a> is very interesting. Here's some code to visualize it in the console. </p>
<pre><code class="language-python">lights = [0]*100
print("".join(map(str,lights)))
for person in range(1,101):
prt = [chr(32)]*100
for l in range(1,101):
if l % person == 0:
lights[l-1]=1-lights[l-1]
if lights[l-1]==1:
prt[l-1] = chr(9608)
print("".join(prt))
</code></pre>
<html><link href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/default.min.css" rel="stylesheet"/>
<script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/highlight.min.js"> </script>
<script>hljs.highlightAll();</script></html>
</description>
<pubDate>Mon, 18 Dec 2023 18:47:10 +0000</pubDate>
<guid>https://www.sixhat.net/2023/toggling-light-bulb-problem.html</guid>
</item>
<item>
<title>is this the only css you'll ever need?</title>
<link>https://www.sixhat.net/2023/css-is-this-the-only-css-youll-ever-need.html</link>
<description>
<h2 id="is-this-this-the-only-css-youll-ever-need"><a class="header" href="#is-this-this-the-only-css-youll-ever-need">Is this this the only CSS you'll ever need?</a></h2>
<pre><code class="language-css">body {
margin: 40px auto;
padding: 0 10px;
max-width: 66ch;
line-height: 1.6;
font-size: 18px;
color: #ccc;
background: #234;
}
h1,
h2,
h3 {
line-height: 1.2;
}
a {
color: #abf;
}
</code></pre>
<html><link href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/default.min.css" rel="stylesheet"/>
<script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/highlight.min.js"> </script>
<script>hljs.highlightAll();</script></html>
</description>
<pubDate>Mon, 18 Dec 2023 18:43:59 +0000</pubDate>
<guid>https://www.sixhat.net/2023/css-is-this-the-only-css-youll-ever-need.html</guid>
</item>
<item>
<title>code snippet for highlight.js in markdown pages</title>
<link>https://www.sixhat.net/2023/code-snippet-for-highlightjs-in-markdown-pages.html</link>
<description>
<h2 id="code-snippet-for-highlightjs-in-markdown-pages"><a class="header" href="#code-snippet-for-highlightjs-in-markdown-pages">Code Snippet for Highlight.js in Markdown Pages</a></h2>
<p>This is just the snippet of code I use on pages that need syntax highlighting.
I don't add them directly to the template as other pages don't need to load them.</p>
<pre><code class="language-html">&lt;html&gt;&lt;link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/default.min.css"&gt;
&lt;script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/highlight.min.js"&gt; &lt;/script&gt;
&lt;script&gt;hljs.highlightAll();&lt;/script&gt;&lt;/html&gt;
</code></pre>
<p>Just copy and paste this at the end of markdown file</p>
<html><link href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/default.min.css" rel="stylesheet"/>
<script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/highlight.min.js"></script>
<script>hljs.highlightAll();</script></html>
</description>
<pubDate>Mon, 18 Dec 2023 18:30:09 +0000</pubDate>
<guid>https://www.sixhat.net/2023/code-snippet-for-highlightjs-in-markdown-pages.html</guid>
</item>
<item>
<title>a simple css trick 4 dithered images</title>
<link>https://www.sixhat.net/2023/a-simple-css-trick-por-dithered-images.html</link>
<description>
<h2 id="a-simple-css-trick-4-dithered-images"><a class="header" href="#a-simple-css-trick-4-dithered-images">A simple CSS trick 4 dithered images</a></h2>
<pre><code class="language-css">p.img {
background: linear-gradient(0deg, #c91136 0%, #4343b5 35%, #eebf18 100%);
display: inline-flex;
}
p.img &gt; img {
mix-blend-mode: hard-light;
}
</code></pre>
<html><link href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/default.min.css" rel="stylesheet"/>
<script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/highlight.min.js"></script>
<script>hljs.highlightAll();</script></html>
</description>
<pubDate>Mon, 18 Dec 2023 18:24:16 +0000</pubDate>
<guid>https://www.sixhat.net/2023/a-simple-css-trick-por-dithered-images.html</guid>
</item>
<item>
<title>me</title>
<link>https://www.sixhat.net/about.html</link>
<description>
<h2 id="about-this-notebook"><a class="header" href="#about-this-notebook">About this notebook</a></h2>
<p>Hi, I'm David. You've stumbled upon my personal online notebook. The notes you find here are my working notes about things that interest me in some manner. </p>
<p>These notes are mostly written for myself. This website is a notebook where I take notes, draft ideas and try a few things. It is not a full publication website –– My <a href="https://www.davidrodrigues.org/">academic Website sits here</a> -- and don't expect it to be highly edited or even correct most of the time. It is just my personal journal to take notes for my future reference. These are evergreen notes, that will evolve over time or be forgotten forever in these pages.</p>
<p>There is a main <a href="index.html">index</a> but it might not include all notes available in the website as some will just live under hierarchies of ideas. I'm very fond of Zettlekastens but usually lack the rigorous mind to keep them organised. </p>
</description>
<pubDate>Mon, 18 Dec 2023 17:54:20 +0000</pubDate>
<guid>https://www.sixhat.net/about.html</guid>
</item>
<item>
<title>websites que funcionam apenas em modo texto</title>
<link>https://www.sixhat.net/2023/websites-que-funcionam-apenas-em-modo-texto.html</link>
<description>
<h2 id="websites-que-funcionam-apenas-em-modo-texto"><a class="header" href="#websites-que-funcionam-apenas-em-modo-texto">Websites que funcionam apenas em modo texto</a></h2>
<p>Já tinha uma <a href="https://www.sixhat.net/article/uncategorized/2022/Ainda-e-possivel-navegar-a-Web-com-o-Javascript-Desligado.html">lista de websites</a> que funcionam com o JavaScript desligado, e portanto muito mais rápidas. </p>
<p>Agora encontrei uma <a href="https://sjmulder.nl/en/textonly.html">lista do Sijmen Mulder</a> que tem ainda mais recursos. Alguns já eram conhecidos mas é interessante explorar as novidades.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:39:14 +0000</pubDate>
<guid>https://www.sixhat.net/2023/websites-que-funcionam-apenas-em-modo-texto.html</guid>
</item>
<item>
<title>two main developments in the ai generators world</title>
<link>https://www.sixhat.net/2023/two-main-developments-in-the-ai-generators-world.html</link>
<description>
<h2 id="two-main-developments-in-the-ai-generators-world"><a class="header" href="#two-main-developments-in-the-ai-generators-world">Two main developments in the AI generators world</a></h2>
<p><a href="https://bard.google.com/">Bard</a> is entering public beta, although not globally. &gt; It sucks big time.</p>
<p>Adobe is entering this crazy world with <a href="https://www.adobe.com/sensei/generative-ai/firefly.html">Firefly</a>. Goodbye Designers, Hello cousin with an internet connection and an Adobe subscription.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:39:08 +0000</pubDate>
<guid>https://www.sixhat.net/2023/two-main-developments-in-the-ai-generators-world.html</guid>
</item>
<item>
<title>tools for modern research</title>
<link>https://www.sixhat.net/2023/tools-for-modern-research.html</link>
<description>
<h2 id="tools-for-modern-research"><a class="header" href="#tools-for-modern-research">Tools for Modern Research</a></h2>
<p>With all the ChatGPT Buzz out there the truth is that there are many very interesting tools to help us write/do better science. Here's a short list:</p>
<p><a href="https://elicit.org/">https://elicit.org/</a></p>
<p><a href="https://chat.openai.com/chat">https://chat.openai.com/chat</a></p>
<p><a href="https://typeset.io/">https://typeset.io/</a></p>
<p><a href="https://app.wordtune.com/account/signup?product=read">https://app.wordtune.com/account/signup?product=read</a></p>
<p><a href="https://www.explainpaper.com/">https://www.explainpaper.com/</a></p>
<p><a href="https://www.perplexity.ai/">https://www.perplexity.ai/</a></p>
<p><a href="https://www.copy.ai/">https://www.copy.ai/</a></p>
<p><a href="https://originality.ai/">https://originality.ai/</a></p>
<p><a href="https://openai-openai-detector.hf.space/">https://openai-openai-detector.hf.space/</a></p>
<p><a href="https://gptzero.me/">https://gptzero.me/</a></p>
</description>
<pubDate>Mon, 18 Dec 2023 17:39:03 +0000</pubDate>
<guid>https://www.sixhat.net/2023/tools-for-modern-research.html</guid>
</item>
<item>
<title>the ai races for march 22:</title>
<link>https://www.sixhat.net/2023/the-ai-races-for-march-22.html</link>
<description>
<h2 id="the-ai-races-for-march-22"><a class="header" href="#the-ai-races-for-march-22">The AI Races for March 22:</a></h2>
<ul>
<li>Bill Gates published an extensive article on is blog about how he thinks this is a totally game changer technology. <a href="https://www.gatesnotes.com/The-Age-of-AI-Has-Begun?WT.mc_id=20230321100000_Artificial-Intelligence_BG-TW_&amp;WT.tsrc=BGTW">The Age of AI has begun</a> is a must read for anyone interested in these subjects. </li>
<li><a href="https://blog.mozilla.org/en/mozilla/introducing-mozilla-ai-investing-in-trustworthy-ai/">Mozilla presented</a> the <a href="https://mozilla.ai/">Moz://a.ai</a>, a startup community aiming at building trustworthy and open-source AI. With the rapid advances in the field this will become a must if we don't want some kind of violent reaction to it.</li>
</ul>
</description>
<pubDate>Mon, 18 Dec 2023 17:38:42 +0000</pubDate>
<guid>https://www.sixhat.net/2023/the-ai-races-for-march-22.html</guid>
</item>
<item>
<title>svelte link dump.</title>
<link>https://www.sixhat.net/2023/svelte-link-dump.html</link>
<description>
<h2 id="svelte-link-dump"><a class="header" href="#svelte-link-dump">Svelte link dump.</a></h2>
<ul>
<li><a href="https://svelte.dev/">Svelte</a> - The App Framework </li>
<li><a href="https://kit.svelte.dev/">Sveltekit</a> - The UI Framework (provides routers, etc...)</li>
<li><a href="https://threlte.xyz/">Threlte</a> - The ThreeJS bit that makes this great.</li>
</ul>
</description>
<pubDate>Mon, 18 Dec 2023 17:38:38 +0000</pubDate>
<guid>https://www.sixhat.net/2023/svelte-link-dump.html</guid>
</item>
<item>
<title>still reading about ai and gpt and what is next in this space</title>
<link>https://www.sixhat.net/2023/still-reading-about-ai-and-gpt-and-whats-next-in-this-space.html</link>
<description>
<h2 id="still-reading-about-ai-and-gpt-and-whats-next-in-this-space"><a class="header" href="#still-reading-about-ai-and-gpt-and-whats-next-in-this-space">Still reading about AI and GPT and what's next in this space</a></h2>
<ul>
<li><a href="https://rodneybrooks.com/what-will-transformers-transform/">Let's cool dow a
bit</a>- A very "food
for thought" essay where the euphoria about transformers technology (read GPTs)
is brought to sane levels, exposing some of its bias, mistakes, and
highlighting the ontological differences between what they are and people's
expectations of them.</li>
<li><a href="https://aisnakeoil.substack.com/p/openais-policies-hinder-reproducible">Open AI and Reproducible
science</a>
Open AI is decommissioning Codex (the one that powers Copilot for example), and
this creates a serious problem: What happens to the science produced with /
based on Codex? How will someone be able to reproduce the scientific results? </li>
<li><a href="https://www.tabnine.com/">Tabnine</a> - This is a code completion / assistant
<em>a la</em> Copilot. Just trying this out now and ... well as Copilot, I'm really
amazed how AI will transform programming if you keep a human in the AI loop.
Tabnine and Copilot are not the only ones and are in what might be called gen.
1 AI assistants. I can't imagine what these space will be in 5 years. But Wow.
I want in. </li>
</ul>
</description>
<pubDate>Mon, 18 Dec 2023 17:38:32 +0000</pubDate>
<guid>https://www.sixhat.net/2023/still-reading-about-ai-and-gpt-and-whats-next-in-this-space.html</guid>
</item>
<item>
<title>spell checker for shell scripts</title>
<link>https://www.sixhat.net/2023/spellcheck-shell-sripts.html</link>
<description>
<h2 id="spell-checker-for-shell-scripts"><a class="header" href="#spell-checker-for-shell-scripts">Spell Checker for Shell Scripts</a></h2>
<p>I'm not a prolific shell programmer (although my <a href="https://www.davidrodrigues.org/research.html">master's dissertation</a> leveraged bash + gawk and made the <em>programming</em> part easy) but I used them for many small tasks that are repetitive and can be simplified via a script. </p>
<p>Writing those scripts for the shell (bash, or whatever) is hard because of some quirks in the language/system, like <a href="https://www.gnu.org/software/bash/manual/html_node/Shell-Expansions.html">shell expansions</a> and <a href="https://en.wikipedia.org/wiki/Glob_(programming)">file globing </a>
Therefore, it is nice to have some sort of syntax checker for most small pickles. <a href="https://www.shellcheck.net/">https://www.shellcheck.net/</a> mitigates the problem by analysing your scripts for some common buggy patterns.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:38:23 +0000</pubDate>
<guid>https://www.sixhat.net/2023/spellcheck-shell-sripts.html</guid>
</item>
<item>
<title>small is beautiful</title>
<link>https://www.sixhat.net/2023/small-is-beautiful.html</link>
<description>
<h2 id="small-is-beautiful"><a class="header" href="#small-is-beautiful">Small is beautiful</a></h2>
<p>I've been around for a long time to remember...</p>
<p>-- Dave, are you going to write that?</p>
<p>... as I was saying, to remember when all we had was small things. Code compiled
to small sizes, websites were a few kB, computers were slow, but the slownness
didn't come from size of code, slow was just because of lack of chip power.
These days we are all spoiled by the abundance of <em>processing power</em>. </p>
<p>That is why this <a href="https://benhoyt.com/writings/the-small-web-is-beautiful/">text rings so many rights in a world full of wrongs</a>.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:36:35 +0000</pubDate>
<guid>https://www.sixhat.net/2023/small-is-beautiful.html</guid>
</item>
<item>
<title>senhor clemente, que oportunidade perdida para o arrependimento.</title>
<link>https://www.sixhat.net/2023/senhor-clemente-que-oportunidade-perdida-para-o-arrependimento.html</link>
<description>
<h2 id="senhor-clemente-que-oportunidade-perdida-para-o-arrependimento"><a class="header" href="#senhor-clemente-que-oportunidade-perdida-para-o-arrependimento">Senhor Clemente, que oportunidade perdida para o arrependimento.</a></h2>
<p>A igreja deixou andar a comissão, e agora fez aquilo que já se esperava de uma organização podre até ao tutano. Matou os resultados de uma assentada. Cuspiu na cara das vítimas e ainda se riu. E o pior é que o fez não por uma razão de fé, ou de crença ou qualquer outra coisa, refugiou-se numa artimanha de um estado de direito que é a "produção de prova". </p>
<blockquote>
<p>"Se essa lista de nomes for preenchida por factos, tanto nós como as autoridades civis podemos actuar"</p>
</blockquote>
<p>O Sr cardeal-patriarca de Lisboa decidiu matar o processo de investigação aos pedófilos da igreja e argumentando que na justiça civil só o é o que se prova. Criou um artefacto para salvar a pele aos amigos pedófilos e predadores sexuais que a igreja portuguesa acoita. </p>
<p><strong>Sr cardeal-patriarca, por vezes sabem-se coisas que não se podem provar.</strong></p>
<p>Ou isso ou então está com medo de PAGAR. Lembre-se que nos <a href="https://en.wikipedia.org/wiki/Catholic_Church_sex_abuse_cases_in_the_United_States">estados unidos a igreja terá pago mais de 4 mil milhões</a> (uma tapezinha) desde 1980 a cerca de 17000 vítimas. Será que o este "shutdown" procura servir quem? os cofres da igreja portuguesa? Algumas dioceses nos EUA tiveram que pedir falência por causa dos escândalos, será que por cá já se estão a fazer as contas? é o que parece. </p>
<p>As vítimas sabem coisas que não podem provar e lá porque não se podem provar não quer dizer que não tenham acontecido e que os nomes... sim os nomes que o Sr quer esconder não passam de criminosos que a Igreja, abusando da lei do estado, procura esconder. Talvez, tal como noutros países não se consigam provar todos actos a ponto de condenar, mas talvez como noutros países se possam condenar pelo encobrimento os bispos, arcebispos ou cardeais que o praticam. E se não for possível pela justiça laica de que abusam, pelo menos da justiça divina da sua igreja não escaparão.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:36:29 +0000</pubDate>
<guid>https://www.sixhat.net/2023/senhor-clemente-que-oportunidade-perdida-para-o-arrependimento.html</guid>
</item>
<item>
<title>red led at 13, blue led at 12 - police lights</title>
<link>https://www.sixhat.net/2023/red-led-at-13-blue-led-at-12---police-lights.html</link>
<description>
<h2 id="red-led-at-13-blue-led-at-12---police-lights"><a class="header" href="#red-led-at-13-blue-led-at-12---police-lights">RED LED at 13, BLUE LED at 12 - Police Lights</a></h2>
<pre><code class="language-cpp">int main() {
DDRB = B00110000;
PINB = 1 &lt;&lt; PINB5;
while (true) {
PINB = B00110000;
for (long i = 0; i &lt; 500000; i++) {
asm("");
}
}
}
</code></pre>
<p>Tags: Arduino, AVR, ATmega</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:36:23 +0000</pubDate>
<guid>https://www.sixhat.net/2023/red-led-at-13-blue-led-at-12---police-lights.html</guid>
</item>
<item>
<title>readings on strange programming, art and electronics</title>
<link>https://www.sixhat.net/2023/readings-on-strange-programming-art-and-electronics.html</link>
<description>
<h2 id="readings-on-strange-programming-art-and-electronics"><a class="header" href="#readings-on-strange-programming-art-and-electronics">Readings on strange programming, art and electronics</a></h2>
<ul>
<li>Although from 3 years ago, this <a href="https://raganwald.com/2020/05/03/fractran.html">FRACTRAN page</a> by <a href="https://braythwayt.com/">Reg Braithwaite</a> is pure intellectual joy. Dive in at your own peril. (also if you use JQuery check <a href="https://github.com/raganwald/JQuery-Combinators">JQuery-Combinators</a>)</li>
<li>Amy Goodchild has very interesting post about <a href="https://www.amygoodchild.com/blog/computer-art-50s-and-60s">computer art of the 50s and 60s</a>. </li>
<li>An interesting project to create an <a href="https://ultimateelectronicsbook.com/#">Ultimate Electronics Book</a>. Another one? Well they all need 2 things to work, be clear and be interactive. If not I can't give them to my students.</li>
</ul>
</description>
<pubDate>Mon, 18 Dec 2023 17:36:18 +0000</pubDate>
<guid>https://www.sixhat.net/2023/readings-on-strange-programming-art-and-electronics.html</guid>
</item>
<item>
<title>organising stuff is hard until it is not</title>
<link>https://www.sixhat.net/2023/organising-stuff-is-hard-until-it-isnt.html</link>
<description>
<h2 id="organising-stuff-is-hard-until-it-isnt"><a class="header" href="#organising-stuff-is-hard-until-it-isnt">Organising stuff is hard until it isn't</a></h2>
<p><a href="https://bulletjournal.com/">Bullet Journal</a> - A very flexible analog system on note taking (that can also be used with digital notetaking tools)</p>
<p><a href="https://johnnydecimal.com/">Johnny.Decimal</a> - An interesting way of organising stuff in an hierarchical manner.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:36:12 +0000</pubDate>
<guid>https://www.sixhat.net/2023/organising-stuff-is-hard-until-it-isnt.html</guid>
</item>
<item>
<title>o fim das trotinetes de aluguer?</title>
<link>https://www.sixhat.net/2023/o-fim-das-trotinetes-de-aluguer.html</link>
<description>
<h2 id="o-fim-das-trotinetes-de-aluguer"><a class="header" href="#o-fim-das-trotinetes-de-aluguer">O fim das trotinetes de aluguer?</a></h2>
<p>Paris <a href="https://www.france24.com/en/live-news/20230402-paris-votes-on-ban-for-rental-e-scooters">baniu</a> as trotinetes eléctricas de aluguer. Atendendo ao que representa esta medida não me admirava que em breve também Lisboa seguisse o exemplo. O modelo de funcionamento destes negócios representa o abuso completo do espaço público, com total desrespeito pelo espaço pedonal. Se ainda por cima o se tornam inseguras, não há como dar a volta ao texto e as gestões urbanas acabarão por impor restrições ou proibições completas. Sim, a culpa não é das marcas, mas dos utilizadores irresponsáveis. Mentira, também é das marcas que desenham um negócio baseado na apropriação do espaço público, como as esplandas do Covid que nunca mais desapareceram e que comem em muitos sítios mais de metade dos passeios. Mas há licenças. Então terminem-se.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:36:04 +0000</pubDate>
<guid>https://www.sixhat.net/2023/o-fim-das-trotinetes-de-aluguer.html</guid>
</item>
<item>
<title>no arrendamento, quem se lixa é quem cumpre e já aluga</title>
<link>https://www.sixhat.net/2023/no-arrendamento-quem-se-lixa-e-quem-cumpre-e-ja-aluga.html</link>
<description>
<h2 id="no-arrendamento-quem-se-lixa-é-quem-cumpre-e-já-aluga"><a class="header" href="#no-arrendamento-quem-se-lixa-é-quem-cumpre-e-já-aluga">No arrendamento, quem se lixa é quem cumpre e já aluga</a></h2>
<ul>
<li>Quem converter AL em arrendamento tem benefícios e isenções</li>
<li>Off-shores que venderem casas ao estado tem isenções</li>
</ul>
<p>Esta salganhada do arrendamento que o governo se prepara para criar mostra claramente que vai haver quem vai pagar a fatura, e esse alguém parecem ser os actuais senhorios que já alugam as casas. Então agora se um AL se converter em arrendamento do estado fica com isenção de IRS até 2030? Então alguém que está num mercado competitivo vai ficar com uma vantagem comercial de 25%? A lei da concorrência fica completamente desvirtuada. Os que já fazem aquilo que o estado pretende, pelos vistos podem ser sacrificados pelo estado e pagar a fatura. Vamos bem. </p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:56 +0000</pubDate>
<guid>https://www.sixhat.net/2023/no-arrendamento-quem-se-lixa-e-quem-cumpre-e-ja-aluga.html</guid>
</item>
<item>
<title>models of creativity</title>
<link>https://www.sixhat.net/2023/models-of-creativity-.html</link>
<description>
<h2 id="models-of-creativity"><a class="header" href="#models-of-creativity">Models of Creativity</a></h2>
<p>Are there any models on creativity? Not creativity in a utilitarian way, like innovation, but creativity, just a theory of how creativity exists.</p>
<ul>
<li>Exploring Creativity and Urban Development with Agent-Based Modelling - <a href="https://jasss.soc.surrey.ac.uk/18/2/12.html">https://jasss.soc.surrey.ac.uk/18/2/12.html</a></li>
</ul>
<p>Different factors are studied with ABM (transport, land use, residential segregation) and the authors found that there is a trade-off between economic progress and socioeconomic equity. <strong>Is Creativity elitist?</strong> #FullRead</p>
<ul>
<li>creative urban environments
<ul>
<li>mixed land-use and population density</li>
<li>affordable urban and regional mobility</li>
<li>high levels of societal tolerance.</li>
</ul>
</li>
</ul>
<p>Notes:</p>
<blockquote>
<p>"... the starting point for theory building is establishing relevant stylised facts"</p>
</blockquote>
<p>KALDOR, N. (1957). A Model of Economic Growth. The Economic Journal, 67 (268): 591–624. [doi:10.2307/2227704]</p>
<ul>
<li>"... provides the intellectual platform underlying the ... approach..."§</li>
<li>§ 2.1 - "... maximize the benefits of agglomeration by reducing the inefficiencies caused by congestion."</li>
<li>§ 2.2 - a notion of 'creativity' conceptualises the relationship between urban morphology and economic productivity ... </li>
<li>§2.4 "doubling the employment density (jobs per unit area) is associated with 2-6 percent rise in productivity". "high density of interactions is associated with high levels of economic productivity" (potential connection with imperfect information and in-person interactions)</li>
<li>§2.6 Innovation &lt;- in-person interactions. Tension between Centrifugal (causing sprawl) and centripetal (-&gt; density).</li>
<li>§2.8 "observe global phenomena through local-level interactions"</li>
<li>§3.2 "Axtell and Epstein (1994) classify ABMs from "level 0" types that broadly caricature real-world agent behavior to "level 3" where they are in quantitative agreement with both the macro- and "micro-structures" of the target system."
<ul>
<li>AXTELL, R. &amp; Epstein, J., (1994). Agent-based Modelling: Understanding Our Creations. <em>The Bulletin of the Santa Fe Institute</em>, (Winter): 28–32.</li>
</ul>
</li>
<li>the model <a href="https://www.openabm.org/model/4396/version/1/view">https://www.openabm.org/model/4396/version/1/view</a> is available for download and also includes de <a href="grimm-odd.html">ODD</a> description</li>
<li>§4.1 emergence of creative clusters</li>
</ul>
<p>Like so many other properties of life, measuring creativity needs indicies:</p>
<ul>
<li>Creativity and Prosperity: The Global Creativity Index https://tspace.library.utoronto.ca/handle/1807/80125 by Richard Florida et al; In this one the index (CGI - Global Creativity Index) is based on 3Ts of economic development (Talent, Technology, and Tolerance [What is the opposite of Tolerance?] .)</li>
</ul>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:48 +0000</pubDate>
<guid>https://www.sixhat.net/2023/models-of-creativity-.html</guid>
</item>
<item>
<title>marginalia in the modern digital world. is it possible?</title>
<link>https://www.sixhat.net/2023/marginalia-in-the-modern-digital-world-is-it-possible.html</link>
<description>
<h2 id="marginalia-in-the-modern-digital-world-is-it-possible"><a class="header" href="#marginalia-in-the-modern-digital-world-is-it-possible">Marginalia in the modern digital world. Is it possible?</a></h2>
<p>Read a paper on the process of writing Marginalia in books, both in print and digital versions, and how the e-reader versions makes annotating books much more difficult and having a very different nature and perceived value by the readers. The paper is <a href="https://www.sciencedirect.com/science/article/abs/pii/S0740818817300099">"Marginalia in the digital age: Are digital reading devices meeting the needs of today's readers?" by Melanie Ramdarshan Bold and Kiri L. Wagstaff</a>. They conducted a survey and the users self-reported there views on marginalia. There are many interesting points, namely the print vs digital, the pedagogical value of marginalia, the leisure vs educational marginalia, and one of the most fascinating, the individualistic vs social aspect of writing in the margins of books. </p>
<p class="mar">PS: Should I re-implement a marginalia system for this blog?</p>
<p>Tags: books, reading</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:43 +0000</pubDate>
<guid>https://www.sixhat.net/2023/marginalia-in-the-modern-digital-world-is-it-possible.html</guid>
</item>
<item>
<title>m3, m3 pro and m3 max</title>
<link>https://www.sixhat.net/2023/m3-m3-pro-and-m3-max.html</link>
<description>
<h2 id="m3-m3-pro-and-m3-max"><a class="header" href="#m3-m3-pro-and-m3-max">M3, M3 Pro and M3 Max</a></h2>
<p>Isn't this trio of performance chips similar to what intel has been doing forever with the i3, i5 and i7 chips? What is this trio paranoia? </p>
<p>Tags: apple</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:38 +0000</pubDate>
<guid>https://www.sixhat.net/2023/m3-m3-pro-and-m3-max.html</guid>
</item>
<item>
<title>in the slow movement you find great pearls of wisdom.</title>
<link>https://www.sixhat.net/2023/in-the-slow-movement-you-find-great-pearls-of-wisdom.html</link>
<description>
<h2 id="in-the-slow-movement-you-find-great-pearls-of-wisdom"><a class="header" href="#in-the-slow-movement-you-find-great-pearls-of-wisdom">In the slow movement you find great pearls of wisdom.</a></h2>
<p>I'm really impressed how much love some people put into small works that are really well crafted.
Two websites that have impressed me lately are the <a href="https://solar.lowtechmagazine.com/">Low&lt;–Tech Magazine</a>, a really well thought website with long reads that are inspiring and well written. More, it is powered by solar power! The second website that really impressed me recently is <a href="https://rawtext.club/~sloum/">Sloum's</a>, mainly because of the little software utilities he has coded and published. If you have time (and you should have time, or get time) take a look at these. </p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:31 +0000</pubDate>
<guid>https://www.sixhat.net/2023/in-the-slow-movement-you-find-great-pearls-of-wisdom.html</guid>
</item>
<item>
<title>I am back to social networks, and it is mastodon.</title>
<link>https://www.sixhat.net/2023/im-back-to-social-networks-and-it-is-mastodon.html</link>
<description>
<h2 id="im-back-to-social-networks-and-it-is-mastodon"><a class="header" href="#im-back-to-social-networks-and-it-is-mastodon">I'm back to social networks, and it is Mastodon.</a></h2>
<p>As the blue bird melts down I've decided to get back to social networks. I joined Mastodon. If you want to get out of the toxic/dictatorial world that Twitter has become and want to go back to fundamental community sharing in a decentralised way that no one can buy and destroy, give Mastodon a try. You can find me <a href="https://datasci.social/@sixhat">@sixhat@datasci.social</a></p>
<p>The main reason for choosing Mastodon is that it is a federated not owned centrally by one entity. For me, this decentralisation is of primary importance. If you don't find a community that you like, you can join with your server, build your small community (even if it is only your family) interested in some particular topic, and still take part in a greater public discussion. </p>
<p>To start, <a href="https://joinmastodon.org/servers">choose a server</a> on a topic that you are keen on and start following some people (you can follow people from other servers, you are not in a silo, or have reading limits) and some #hashtags. </p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:27 +0000</pubDate>
<guid>https://www.sixhat.net/2023/im-back-to-social-networks-and-it-is-mastodon.html</guid>
</item>
<item>
<title>i have to many rss feeds in my reader</title>
<link>https://www.sixhat.net/2023/i-have-to-many-rss-feeds-in-my-reader.html</link>
<description>
<h2 id="i-have-to-many-rss-feeds-in-my-reader"><a class="header" href="#i-have-to-many-rss-feeds-in-my-reader">I have to many RSS feeds in my reader</a></h2>
<p>I've been using RSS since its inception, and no, <a href="https://www.google.com/search?q=site%253Asixhat.net++RSS+is+not+dead">RSS is not dead</a>, but my opml has a lot of dead wood in it and that is annoying. I'd like to have an RSS reader that would give me statistics on my feeds, readings, prefered feeds, etc... could it be an enterprise for personalized AI?</p>
<p>By the way, I mostly use newsboat as my reader.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:24 +0000</pubDate>
<guid>https://www.sixhat.net/2023/i-have-to-many-rss-feeds-in-my-reader.html</guid>
</item>
<item>
<title>how we learn and how to organise a reading inbox</title>
<link>https://www.sixhat.net/2023/how-we-learn-and-how-to-organise-a-reading-inbox.html</link>
<description>
<h2 id="how-we-learn-and-how-to-organise-a-reading-inbox"><a class="header" href="#how-we-learn-and-how-to-organise-a-reading-inbox">How we learn and how to organise a Reading Inbox</a></h2>
<p>I'm always interested in opportunities to learn new things, but like many, I struggle with information overload, in the digital world <em>entertainment</em> is so prevalent that the drive to learn new things becomes subdued to the joy of rapid exploration.</p>
<p>In any case, I've come across a couple of references that I want to revisit in detail.</p>
<ul>
<li><a href="https://giansegato.com/essays/edutainment-is-not-learning">How to learn better in the Digital Age</a> This points in some interesting directions: the biological aspect of learning, the physical effort needed, and the association between retention and physical action is something often forgotten. The need for <em>effort</em> for learning to be successful. Nothing easy is worth much.</li>
<li><a href="https://notes.andymatuschak.org/A_reading_inbox_to_capture_possibly-useful_references">The Reading Inbox</a>This points (in the GTD style?) to the problem of the Reading Inbox — a problem I have (my Reading Inbox is over capacity). Maybe one should implement a Reading Inbox as a circular LinkedList with a fixed capacity. If it grows it starts overwriting older entries. Hm...</li>
</ul>
<p>Tags: learning, gtd, digital-life</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:21 +0000</pubDate>
<guid>https://www.sixhat.net/2023/how-we-learn-and-how-to-organise-a-reading-inbox.html</guid>
</item>
<item>
<title>grimm's odd standard protocol for describing individual-based and agent based models</title>
<link>https://www.sixhat.net/2023/grimm-odd.html</link>
<description>
<h2 id="grimms-odd"><a class="header" href="#grimms-odd">Grimm's ODD</a></h2>
<p>GRIMM, V., Berger, U., Bastiansen, F., Eliassen, S., Ginot, V., Giske, J., Goss-Custard, J., Grand, T., Heinz, S., Huse, G., Huth, A., Jepsen, J., Jorgensen, C., Mooij, W., Muller, B., Pe'er, G., Piou, C., Railsback, S., Robbins, A., Robbins, M., Rossmanith, E., Ruger, N., Strand, E., Souissi, S., Stillman, R., Vabo, R., Visser, U. &amp; Deangelis, D. (2006). A Standard Protocol for Describing Individual-Based and Agent-Based Models. Ecological Modelling, 198 (1–2): 115–126. <a href="https://www.sciencedirect.com/science/article/abs/pii/S0304380006002043">doi:10.1016/j.ecolmodel.2006.04.023</a></p>
<p>A standard protocol for describing individual-based and agent-based models, the authors describe the ODD protocol that has become the <em>de facto</em> standard to describe agent based models. </p>
<p>In papers these ODD descriptions either appear inline (rare), but mostly appear as a supplementary material (common), and in some cases they also follow the software distribuition.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:18 +0000</pubDate>
<guid>https://www.sixhat.net/2023/grimm-odd.html</guid>
</item>
<item>
<title>gpt4 experiments - sparks of agi</title>
<link>https://www.sixhat.net/2023/gpt4-experiments---sparks-of-agi.html</link>
<description>
<h2 id="gpt4-experiments---sparks-of-agi"><a class="header" href="#gpt4-experiments---sparks-of-agi">GPT4 experiments - Sparks of AGI</a></h2>
<p><a href="https://arxiv.org/abs/2303.12712">Sparks of Artificial General Intelligence: Early experiments with GPT-4</a> the paper and this video on <a href="https://www.youtube.com/watch?v=qbIk7-JPB2c">Youtube by Bubeck</a> at the MIT on March 22, 2023 are a must read/watch.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:15 +0000</pubDate>
<guid>https://www.sixhat.net/2023/gpt4-experiments---sparks-of-agi.html</guid>
</item>
<item>
<title>fuzzy logic shell prompt alias</title>
<link>https://www.sixhat.net/2023/fuzzy-logic-shell-alias.html</link>
<description>
<h2 id="fuzzy-logic-shell-alias"><a class="header" href="#fuzzy-logic-shell-alias">fuzzy logic shell alias</a></h2>
<p>I've used <a href="https://github.com/junegunn/fzf">fzf</a> for a long time now. It works great and I'd recommend every power user to use fzf (or any other fuzzy logic search tool)</p>
<p>One of the things I like about using it is to go to the folder of some file I'm working on. For example I use this alias <code>cdf</code> to cd into the file directory. </p>
<p>Just put the following alias in your shell</p>
<pre><code>alias cdf='cd `dirname $(fzf)`'
</code></pre>
<p>Then you can cdf, search, press enter and cd into the file's directory. (Well, there's also ALT-C, but who cares)</p>
<p>You migth want to check <a href="https://andrew-quinn.me/fzf/">this blog post</a> about fzf.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:12 +0000</pubDate>
<guid>https://www.sixhat.net/2023/fuzzy-logic-shell-alias.html</guid>
</item>
<item>
<title>finder, explorer, nautilus, rox, ... spacedrive</title>
<link>https://www.sixhat.net/2023/finder-explorer-nautilus-rox--spacedrive.html</link>
<description>
<h2 id="finder-explorer-nautilus-rox--spacedrive"><a class="header" href="#finder-explorer-nautilus-rox--spacedrive">Finder, Explorer, Nautilus, Rox, ... Spacedrive</a></h2>
<p>Just came across Spacedrive — a new file explorer that is cross platform and aims to bring the same experience across OSes. It reminds a lot of apple Finder, and at this moment it doesn't bring anything that would convince me to change to SpaceDrive. For Windows users and eventually Linux users it might provide a better user experience, but until it can offer something that blows Finder away... it has a steep mountain to climb. In any case it is a nice project to follow as the current release is still alpha — <a href="https://www.spacedrive.com/">Spacedrive.com</a> <a href="https://github.com/spacedriveapp/spacedrive">SpaceDrive Github</a></p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:07 +0000</pubDate>
<guid>https://www.sixhat.net/2023/finder-explorer-nautilus-rox--spacedrive.html</guid>
</item>
<item>
<title>axtel and epstein's levels for agent based performance</title>
<link>https://www.sixhat.net/2023/epsteins-levels-for-agent-based-models-performance.html</link>
<description>
<h2 id="robert-axtell-and-joshua-epstein-levels-of-agent-based-models-performance"><a class="header" href="#robert-axtell-and-joshua-epstein-levels-of-agent-based-models-performance">Robert Axtell and Joshua Epstein levels of agent based models performance</a></h2>
<p>Axtell and Epstein[1] stated that an agent-based model performance can be characterised a a level, with any further level encompassing the previous:</p>
<p><strong>Level 0</strong>: a <em>caricature</em> of reality, as established through the use of simple graphical devices (allowing the visualisation of agent motion)
<strong>Level 1</strong>: The model is in <em>qualitative agreement with empirical macro-structures</em>, as established by plotting distributions of population properties
<strong>Level 2</strong>: The model is in <em>quantitative agreement with empirical macro-structures</em> as established by plotting on-board statistical estimation routines.
<strong>Level 3</strong>: The model is in quantitative agreement with empirical micro-structures, as determined by cross-sectional and longitudinal analysis of the agent population.</p>
<p>1 - Axtell, R., &amp; Epstein, J. (1994). Agent-based modeling: Understanding our creations. <em>The Bulletin of the Santa Fe Institute</em>, <em>9</em>(4), 28-32.</p>
<p>Related to <a href="grimm-odd.html">grimm's ODD</a></p>
</description>
<pubDate>Mon, 18 Dec 2023 17:35:03 +0000</pubDate>
<guid>https://www.sixhat.net/2023/epsteins-levels-for-agent-based-models-performance.html</guid>
</item>
<item>
<title>disable macbook air autoboot when dis/connecting to power or open lid</title>
<link>https://www.sixhat.net/2023/disable-macbook-air-autoboot-when-disconnecting-to-power-or-open-lid.html</link>
<description>
<h2 id="disable-macbook-air-autoboot-when-disconnecting-to-power-or-open-lid"><a class="header" href="#disable-macbook-air-autoboot-when-disconnecting-to-power-or-open-lid">Disable MacBook Air Autoboot when dis/connecting to power or open lid</a></h2>
<pre><code>// disable
sudo nvram AutoBoot=%00
// re-enable
sudo nvram AutoBoot=%03
</code></pre>
<p>Note: And it doesn't work in newer M1, M2 apple computers. </p>
<p>Tags: apple</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:34:58 +0000</pubDate>
<guid>https://www.sixhat.net/2023/disable-macbook-air-autoboot-when-disconnecting-to-power-or-open-lid.html</guid>
</item>
<item>
<title>data science handbook</title>
<link>https://www.sixhat.net/2023/data-science-handbook.html</link>
<description>
<h2 id="data-science-handbook"><a class="header" href="#data-science-handbook">Data Science Handbook</a></h2>
<p>Standford has an online <a href="https://stanforddatascience.github.io/best-practices/">Data Science Handbook</a> aimed at <strong>open, rigorous and reproducible research: A practitioner's handbook</strong>.</p>
<p>This kind of asset is so valuable in a time when we have students wanting to do open, transparent and reproducible research. Good for any type of research but focused on Data Science. </p>
</description>
<pubDate>Mon, 18 Dec 2023 17:34:55 +0000</pubDate>
<guid>https://www.sixhat.net/2023/data-science-handbook.html</guid>
</item>
<item>
<title>computer related stuff---how these machines work</title>
<link>https://www.sixhat.net/2023/computer-related-stuff---how-these-machines-work.html</link>
<description>
<h2 id="computer-related-stuff---how-these-machines-work"><a class="header" href="#computer-related-stuff---how-these-machines-work">Computer related stuff---How these machines work</a></h2>
<p>I might convert this post into a list of links that are good references about computers if I remember to add more links.</p>
<ul>
<li><a href="https://cpu.land/">https://cpu.land/</a></li>
</ul>
<p>Tags: computer</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:34:45 +0000</pubDate>
<guid>https://www.sixhat.net/2023/computer-related-stuff---how-these-machines-work.html</guid>
</item>
<item>
<title>creating indices with tree</title>
<link>https://www.sixhat.net/2023/creating-html-indices-with-tree.html</link>
<description>
<h2 id="creating-html-indices-with-tree"><a class="header" href="#creating-html-indices-with-tree">Creating html indices with tree</a></h2>
<p>Many times I need to create an index of certain files in my computer, namely when they are spreaded over many subfolders. Using the utility <strong>tree</strong> generates indices quickly and in a format that makes them very useful.</p>
<p>For example, to find all PDF files: </p>
<pre><code>tree -P '*.pdf' --prune -H . &gt; index.html
</code></pre>
</description>
<pubDate>Mon, 18 Dec 2023 17:34:42 +0000</pubDate>
<guid>https://www.sixhat.net/2023/creating-html-indices-with-tree.html</guid>
</item>
<item>
<title>blogs without server side rendering</title>
<link>https://www.sixhat.net/2023/blogs-without-server-side-rendering.html</link>
<description>
<h2 id="blogs-without-server-side-rendering"><a class="header" href="#blogs-without-server-side-rendering">Blogs without server side rendering</a></h2>
<p>I love <strong>static websites</strong>, but what about if the <em>static</em> was just markdown files and I let the client render everything via a Javascript library? </p>
<ul>
<li><a href="https://dynalon.github.io/mdwiki/#!index.md">MDWiki</a> - This is used to run this website. Makes it very simple to use. One index.html file and you're done.</li>
<li><a href="https://chrisdiana.github.io/cms.js/">CMS.js</a> - Interesting, with more configuration/conventions to follow. An alternative in any case.</li>
</ul>
<p>(This website uses <a href="https://rust-lang.github.io/mdBook/">MDBook</a> - This is highly recommended, even for bigger things other than a simple blog like this, although it is pre-rendered)</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:34:37 +0000</pubDate>
<guid>https://www.sixhat.net/2023/blogs-without-server-side-rendering.html</guid>
</item>
<item>
<title>beja e alverca</title>
<link>https://www.sixhat.net/2023/beja-e-alverca.html</link>
<description>
<h2 id="beja-e-alverca"><a class="header" href="#beja-e-alverca">Beja e Alverca</a></h2>
<p>Se cada um destes locais demorar 6 meses a avaliar, lá vai a comissão de avaliação ter que trabalhar mais um ano. Que chatice dirá a comissão. Já ninguém acredita que haja aeroporto novo de Lisboa, agora se calhar já niguém vai acretidar que a comissão consiga acabar o trabalho.</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:34:34 +0000</pubDate>
<guid>https://www.sixhat.net/2023/beja-e-alverca.html</guid>
</item>
<item>
<title>avoid long urls extending past the margins of text in latex with xurl</title>
<link>https://www.sixhat.net/2023/avoid-long-urls-extending-past-the-margins-of-text-in-latex-with-xurl.html</link>
<description>
<h2 id="avoid-long-urls-extending-past-the-margins-of-text-in-latex-with-xurl"><a class="header" href="#avoid-long-urls-extending-past-the-margins-of-text-in-latex-with-xurl">Avoid long urls extending past the margins of text in LaTeX with xurl</a></h2>
<p>Use package <em>xurl</em> instead of <em>url</em> in LaTeX documents to avoid having long urls going into the margin of the documents. This will break the long url into multiple lines without creating any issues. </p>
<p>Just load <em>xurl</em> with <em>usepackage</em> instead of <em>url</em> and every time <em>url</em> is called <em>xurl</em> will take over. </p>
<p>Don't know why this one escaped my knowledge for over 15 years, but well, learn something every day. </p>
<p>Tags: latex, typesetting</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:34:31 +0000</pubDate>
<guid>https://www.sixhat.net/2023/avoid-long-urls-extending-past-the-margins-of-text-in-latex-with-xurl.html</guid>
</item>
<item>
<title>ants are amazing - what about organizations - reading</title>
<link>https://www.sixhat.net/2023/ants-are-amazing---what-about-organizations----reading.html</link>
<description>
<h2 id="ants-are-amazing---what-about-organizations----reading"><a class="header" href="#ants-are-amazing---what-about-organizations----reading">Ants are amazing - what about organizations - reading</a></h2>
<ul>
<li><a href="https://www.kooslooijesteijn.net/blog/organizations-are-ant-colonies">https://www.kooslooijesteijn.net/blog/organizations-are-ant-colonies</a> - Organizations are ant colonies. </li>
</ul>
</description>
<pubDate>Mon, 18 Dec 2023 17:34:23 +0000</pubDate>
<guid>https://www.sixhat.net/2023/ants-are-amazing---what-about-organizations----reading.html</guid>
</item>
<item>
<title>and we are in 2023</title>
<link>https://www.sixhat.net/2023/and-we-are-in-2023.html</link>
<description>
<h2 id="and-we-are-in-2023"><a class="header" href="#and-we-are-in-2023">And we are in 2023</a></h2>
<p>And as usual I change things, this time went with something even
simpler.</p>
<p>The old blog still exists and you can read it <a href="/index_2022.html">here</a>. </p>
<p>2023 has to be the year of simplicity and minimalism. Sorry. No time for
complex stuff. Life is changing quickly and this blog will be more of a
refence point of things that interest me, or that I'll need in the</p>
</description>
<pubDate>Mon, 18 Dec 2023 17:33:09 +0000</pubDate>
<guid>https://www.sixhat.net/2023/and-we-are-in-2023.html</guid>
</item>
<item>
<title>and it goes dark</title>
<link>https://www.sixhat.net/2023/and-it-goes-dark.html</link>
<description>
<h2 id="and-it-goes-dark"><a class="header" href="#and-it-goes-dark">And it goes dark</a></h2>
<pre><code class="language-css">:root{color-scheme:only dark}
</code></pre>
</description>
<pubDate>Mon, 18 Dec 2023 17:33:04 +0000</pubDate>
<guid>https://www.sixhat.net/2023/and-it-goes-dark.html</guid>
</item>
<item>
<title>a simple js range one-liner</title>
<link>https://www.sixhat.net/2023/a-simple-js-range-one-liner.html</link>
<description>
<h2 id="a-simple-js-range-one-liner"><a class="header" href="#a-simple-js-range-one-liner">A simple JS range one-liner</a></h2>
<pre><code>const range=n=&gt;[...Array(n).keys()]
</code></pre>
</description>
<pubDate>Mon, 18 Dec 2023 17:32:57 +0000</pubDate>
<guid>https://www.sixhat.net/2023/a-simple-js-range-one-liner.html</guid>
</item>
<item>
<title>gpt-4o, the crazy over helpful assistant</title>
<link>https://www.sixhat.net/2024/05-chatgpt4o.html</link>
<description>
<h1 id="gpt-4o---what-an-awful-personality"><a class="header" href="#gpt-4o---what-an-awful-personality">GPT-4o - What an awful personality</a></h1>
<p>OpenAI just launched GPT-4o, a multimodal model capable of reasoning about text, voice and video in real time. </p>
<p>The demo videos are really impressive about the reasoning, but one has to be be really scared about the personality of the 'persona' in the machine. </p>
<p>It is really over-positively-enthusiastic about everything and about drying paint. It is the hyperactive overly eager friend that tries so hard to please that it becomes completely obnoxious. And it is chatty, trying to convey emotion to every answer, to every interaction. What a pain to interact with. </p>
<p>The researchers from Open AI where obviously enamoured with their creation in the videos, but the 'persona' they created is nothing more than a Frankenstein striving to much to become human.</p>
<p>Someone had to choose that style for the persona—I don't want to start thinking that this personality is a byproduct of the training data. The latter would be very unsettling.</p>
</description>
<pubDate>Mon, 13 May 2024 21:02:19 +0000</pubDate>
<guid>https://www.sixhat.net/2024/05-chatgpt4o.html</guid>
</item>
<item>
<title>Some Games I Like</title>
<link>https://www.sixhat.net/2024/06-some-games-i-like.html</link>
<description>
<h3 id="some-games-i-like-to-be-updated"><a class="header" href="#some-games-i-like-to-be-updated">Some games I like (to be updated)</a></h3>
<p>This can be played on a Go Board - <a href="https://www.puzzle-binairo.com/">Binario</a></p>
<p>This a much more visual sudoku puzzle - <a href="https://www.puzzle-nonograms.com/">Nonograms</a></p>
</description>
<pubDate>Mon, 10 Jun 2024 09:05:28 +0000</pubDate>
<guid>https://www.sixhat.net/2024/06-some-games-i-like.html</guid>
</item>
<item>
<title>reading</title>
<link>https://www.sixhat.net/river.html</link>
<description>
<h2 id="worth-reading-take-your-time"><a class="header" href="#worth-reading-take-your-time">Worth reading, take your time.</a></h2>
<ul>
<li><a href="https://www.bitsofwonder.co/?ref=sixhat.net">Bits of Wonder</a> - deepdives into science, philosophy, and how to be a human</li>
<li><a href="https://arne.me/?ref=sixhat.net">Arne Bahlo</a></li>
<li><a href="https://stpeter.im/journal/?ref=sixhat.net">One Small Voice</a> by Peter Saint-Andre</li>
<li><a href="https://timothygartonash.substack.com/?ref=sixhat.net">History of the Present</a> by Timothy Garton Ash</li>
<li><a href="https://manlius.substack.com/?ref=sixhat.net">Complexity Thoughts</a> by Manlio De Domenico</li>
<li><a href="https://benhoyt.com/writings/?ref=sixhat.net">Ben Hoyt</a></li>
<li><a href="https://studio.ribbonfarm.com/?ref=sixhat.net">Ribonfarm Studio</a> by Venkatesh Rao</li>
<li><a href="https://sive.rs/?ref=sixhat.net">Derek Sivers</a></li>
<li><a href="https://stephango.com/?ref=sixhat.net">Steph Ango</a> CEO of Obsidian, full of great ideas worth thinking about.</li>
<li><a href="https://www.computerenhance.com/?ref=sixhat.net">Computer, Enhance!</a> - More quality information on low level stuff than I might have time to learn, but very cool work.</li>
<li><a href="https://aethermug.com/?ref=sixhat.net">Aethermug</a> and <a href="https://planktonvalhalla.com/?ref=sixhat.net">Plankton Valhalla</a> - good writing always trumps, by Marco Giancotti. </li>
</ul>
</description>
<pubDate>Mon, 10 Jun 2024 08:39:55 +0000</pubDate>
<guid>https://www.sixhat.net/river.html</guid>
</item>
<item>
<title>smartmatch</title>
<link>https://www.sixhat.net/https://perldoc.perl.org/perlop#Smartmatch-Operator</link>
<description>
No description for smartmatch
Please visit the website
</description>
<pubDate>Mon, 09 Dec 2024 16:19:24 +0000</pubDate>
<guid>https://www.sixhat.net/https://perldoc.perl.org/perlop#Smartmatch-Operator</guid>
</item>
<item>
<title>Reply</title>
<link>https://www.sixhat.net/https://metacpan.org/pod/Reply</link>
<description>
No description for Reply
Please visit the website
</description>
<pubDate>Mon, 09 Dec 2024 16:19:24 +0000</pubDate>
<guid>https://www.sixhat.net/https://metacpan.org/pod/Reply</guid>
</item>
<item>
<title>ai</title>
<link>https://www.sixhat.net/ai.html</link>
<description>
<h2 id="ai-river"><a class="header" href="#ai-river">AI River</a></h2>
<ul>
<li><a href="https://eugeneyan.com/writing/evals/">Task-Specific LLM Evals That Do and Don't Work</a></li>
<li><a href="https://www.theverge.com/2024/12/9/24316969/mustafa-suleyman-sam-altman-microsoft-openai-agi">Microsoft’s AI boss and Sam Altman disagree on what it takes to get to AGI</a></li>
<li><a href="https://www.theverge.com/24314821/microsoft-ai-ceo-mustafa-suleyman-google-deepmind-openai-inflection-agi-decoder-podcast">Microsoft AI CEO Mustafa Suleyman says conversational AI is the next web browser</a></li>
<li><a href="https://www.theverge.com/2024/12/9/24314445/reddit-answers-ai-powered-search-tool">Reddit’s new AI search tool helps you find Reddit answers without Google</a></li>
<li><a href="https://go.theregister.com/feed/www.theregister.com/2024/12/09/microsoft_llm_prompt_injection_challenge/">Microsoft dangles $10K for hackers to hijack LLM email service</a></li>
<li><a href="https://go.theregister.com/feed/www.theregister.com/2024/12/09/palantir_anduril_alliance/">We can't make this stuff up: tech companies Palantir and Anduril form fellowship for AI adventures</a></li>
<li><a href="https://go.theregister.com/feed/www.theregister.com/2024/12/09/asia_tech_news_roundup/">China launches AI that writes politically correct docs for bureaucrats</a></li>
<li><a href="https://www.wired.com/story/48-hours-in-tokyo-with-my-ai-travel-companion/">48 Hours in Tokyo With My AI Travel Companion</a></li>
<li><a href="https://slashdot.org/story/24/12/09/159202/ai-boosts-materials-discovery-by-44-at-major-us-lab?utm_source=rss1.0mainlinkanon&amp;utm_medium=feed">AI Boosts Materials Discovery By 44% at Major US Lab</a></li>
<li><a href="https://blog.cloudflare.com/radar-2024-year-in-review-internet-services">From ChatGPT to Temu: ranking top Internet services in 2024</a></li>
<li><a href="https://www.techradar.com/pro/three-approaches-to-generative-ai-which-approach-will-you-take">Three approaches to generative AI - which approach will you take?</a></li>
<li><a href="https://www.techradar.com/computing/artificial-intelligence/goodbye-unreliable-weather-forecasts-google-deepminds-ai-model-sets-new-benchmark-for-15-day-predictions">Goodbye, unreliable weather forecasts? Google DeepMind's AI model sets new benchmark for 15-day predictions</a></li>
<li><a href="https://www.theshortcut.com/p/call-of-duty-black-ops-6-developer-is-being-accused-of-using-ai-generated-artwork">Call of Duty: Black Ops 6 developer is being accused of using AI generated artwork</a></li>
<li><a href="https://theconversation.com/how-utilities-are-working-to-meet-ai-data-centers-voracious-appetite-for-electricity-240196">How utilities are working to meet AI data centers’ voracious appetite for electricity</a></li>
<li><a href="https://qz.com/reddit-answers-ai-chatbot-google-openai-chatgpt-search-1851716486">Reddit is taking on Google and ChatGPT with its own AI chatbot</a></li>
<li><a href="https://qz.com/ai-agents-chatbots-artificial-intelligence-work-1851714587">2025 could be the year AI grows up</a></li>
<li><a href="https://www.cnbc.com/2024/12/09/ant-group-gets-new-ceo-as-billionaire-founder-jack-ma-talks-up-ai.html">Ant Group gets new CEO as billionaire founder Jack Ma talks up AI in rare appearance</a></li>
<li><a href="https://www.fastcompany.com/91242961/nvidia-stock-price-tumbles-china-investigation-news-ai-chip-giant">Nvidia stock tumbles on news of China investigation: Is the AI chip giant in trouble?</a></li>
<li><a href="https://www.fastcompany.com/91237841/helen-toners-openai-exit-only-made-her-a-more-powerful-force-for-responsible-ai">Helen Toner’s OpenAI exit only made her a more powerful force for responsible AI</a></li>
<li><a href="https://www.fastcompany.com/91242382/5-ways-ai-is-helping-spur-the-ev-era">5 ways AI is helping spur the EV era</a></li>
<li><a href="https://www.scientificamerican.com/article/ai-will-turn-our-lives-into-the-truman-show/">AI Will Turn Our Lives into The Truman Show</a></li>
<li><a href="https://www.newscientist.com/article/2459368-chips-linked-with-light-could-train-ai-faster-while-using-less-energy/?utm_campaign=RSS%7CNSNS&amp;utm_source=NSNS&amp;utm_medium=RSS&amp;utm_content=home">Chips linked with light could train AI faster while using less energy</a></li>
<li><a href="https://www.pcgamer.com/software/ai/fbi-recommends-coming-up-with-a-secret-word-or-phrase-to-make-sure-your-family-know-youre-you-and-not-some-hellish-ai-copycat">FBI recommends coming up with a 'secret word or phrase' to make sure your family know you're you and not some hellish AI copycat</a></li>
<li><a href="https://mashable.com/deals/dec-9-desksense-ai-assistant">Get this AI personal assistant with lifetime access to GPT-4, Gemini Pro, and more for under $40</a></li>
<li><a href="/category/artificial-intelligence/">AI</a></li>
<li><a href="/category/artificial-intelligence/">AI</a></li>
<li><a href="/category/robotics/">Robotics</a></li>
<li><a href="https://techcrunch.com/tag/generative-ai/">generative AI</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/09/openai-2024-event-how-to-watch-new-chatgpt-product-reveals-and-demos/">OpenAI 2024 event: How to watch new ChatGPT product reveals and demos</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/09/openais-sora-video-generator-might-not-be-available-in-the-eu-at-launch/">OpenAI’s Sora video generator might not be available in the EU at launch</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/09/reddit-tests-a-conversational-ai-search-tool/">Reddit tests a conversational AI search tool</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/09/study-claims-ai-could-boost-detection-of-breast-cancer-by-21/">Study claims AI could boost detection of breast cancer by 21%</a></li>
<li><a href="https://techcrunch.com/2024/12/08/ucla-offers-comp-lit-course-developed-by-ai/">UCLA offers comp lit course developed by AI</a></li>
<li><a href="https://techcrunch.com/2024/12/07/google-says-its-new-ai-model-outperforms-the-top-weather-forecast-system/">Google says its new AI model outperforms the top weather forecast system</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/06/if-you-can-make-this-ai-bot-fall-in-love-you-could-win-thousands-of-dollars/">If you can make this AI bot fall in love, you could win thousands of dollars</a></li>
<li><a href="https://techcrunch.com/2024/12/06/meta-unveils-a-new-more-efficient-llama-model/">Meta unveils a new, more efficient Llama model</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/05/ex-paypal-coo-david-sacks-is-trumps-new-crypto-and-ai-czar/">Ex-PayPal COO David Sacks is Trump’s new crypto and AI ‘czar’</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/05/elon-musks-xai-lands-billions-in-new-cash-to-fuel-ai-ambitions/">Elon Musk’s xAI lands $6B in new cash to fuel AI ambitions</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/05/chatgpt-everything-to-know-about-the-ai-chatbot/">ChatGPT: Everything you need to know about the AI-powered chatbot</a></li>
<li><a href="https://techcrunch.com/2024/12/05/openai-confirms-its-new-200-plan-chatgpt-pro-which-includes-reasoning-models-and-more/">OpenAI confirms new $200 monthly subscription, ChatGPT Pro, which includes its o1 reasoning model</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/storyline/openais-2024-event-live-updates-for-chatgpt-product-reveals-and-demos/">OpenAI’s 2024 event: Live updates for ChatGPT product reveals and demos</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/05/google-says-its-new-open-models-can-identify-emotions-and-that-has-experts-worried/">Google says its new AI models can identify emotions — and that has experts worried</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/05/copilot-vision-microsofts-ai-tool-that-can-read-your-screen-launches-in-preview/">Copilot Vision, Microsoft’s AI tool that can read your screen, launches in preview</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/05/openai-may-be-planning-a-chatgpt-pro-plan-for-200-per-month/">OpenAI may be planning a ChatGPT Pro plan for $200 per month</a></li>
<li><a href="https://techcrunch.com/category/artificial-intelligence/">AI</a></li>
<li><a href="https://techcrunch.com/2024/12/05/twos-is-a-to-do-app-that-uses-ai-to-suggest-actions-to-complete-your-tasks/">Twos is a to-do app that uses AI to suggest actions to complete your tasks</a></li>
<li><a href="https://techcrunch.com/2024/12/07/openai-bets-youll-pay-200-a-month-for-chatgpt/">OpenAI bets you’ll pay $200 a month for ChatGPT</a></li>
<li><a href="https://techcrunch.com/2024/12/03/why-does-the-name-david-mayer-crash-chatgpt-digital-privacy-requests-may-be-at-fault/">David Mayer ChatGPT</a></li>
<li><a href="https://techcrunch.com/video/surreal-elderhood-using-openais-text-to-video-model-sora/">OpenAI’s Sora</a></li>
<li><a href="https://techcrunch.com/2024/12/04/chatgpt-everything-to-know-about-the-ai-chatbot/">ChatGPT</a></li>
</ul>
<p>Generated Mon Dec 9 16:19:23 2024</p>
</description>
<pubDate>Mon, 09 Dec 2024 16:19:23 +0000</pubDate>
<guid>https://www.sixhat.net/ai.html</guid>
</item>
<item>
<title>2024 Advent Of Code in Perl</title>
<link>https://www.sixhat.net/2024/12-2024-advent-of-code-in-perl.html</link>
<description>
<h1 id="2024-advent-of-code-in-perl"><a class="header" href="#2024-advent-of-code-in-perl">2024 Advent Of Code in Perl</a></h1>
<p>The title says it all. I've never programmed in Perl. I hacked some existing scripts, but never tried this oldie. But as it is a great language for text processing, it was time to start learning it. Maybe it will come handy some time in the future.</p>
<p>My repo is at <a href="https://github.com/sixhat/adventofcode.com/tree/main/2024/">https://github.com/sixhat/adventofcode.com/tree/main/2024/</a></p>
<h2 id="notes-on-using-perl-for-aoc"><a class="header" href="#notes-on-using-perl-for-aoc">Notes on using Perl for AoC</a></h2>
<p>(or... things I'm learning along the way)</p>
<ul>
<li><code>unless</code> is interesting, but I have to get my head around it. But it works.</li>
<li>$@% ? What? I have a lot to learn on the syntax front of Perl. </li>
<li><code>use strict;</code> and <code>use warnings;</code> are your friend</li>
<li>Pass by value and pass by reference are dependent on a lexical variable being created on the subroutine <code>my @local_var = @_</code>. </li>
<li>Regular Expressions are powerful, but they also introduce some complexity and trying to debug them is harder.</li>
<li><code>Use of uninitialized value</code> warnings reveals some <code>undef</code> entries in an array (for example in my <code>@matches</code>). Trick to check them out is to <code>print join(', ', @matches)</code> to double check.</li>
<li><a href="https://perldoc.perl.org/perlop#Smartmatch-Operator">smartmatch</a> is really nice (day 5). </li>
<li>Install a repl - <a href="https://metacpan.org/pod/Reply">Reply</a> - or any other. It makes testing simple things in the language a breeze. </li>
</ul>
</description>
<pubDate>Mon, 09 Dec 2024 16:19:16 +0000</pubDate>
<guid>https://www.sixhat.net/2024/12-2024-advent-of-code-in-perl.html</guid>
</item>
<item>
<title>my approach to managing scratch projects with bash scripts</title>
<link>https://www.sixhat.net/2023/my-approach-to-managing-scratch-projects-with-bash-scripts.html</link>
<description>
<h2 id="my-approach-to-managing-scratch-projects-with-bash-scripts"><a class="header" href="#my-approach-to-managing-scratch-projects-with-bash-scripts">My approach to managing scratch projects with bash scripts</a></h2>
<p>In R if I just want to test an idea I create a scratch project in /tmp . This will removed automatically on the next boot. This makes it easy to setup scratch folders without leaving much clutter behind. The problem is that the next time you open R after a reboot it will complaint about not finding the previous workspace. I also do the same thing with Markdown files and folders. To help me manage this I have a <code>new_project</code> bash script in my bin folder that goes like this (only .r and .md versions shown but can easily be extended for your needs)</p>
<pre><code class="language-bash">#!/bin/bash
set -euo pipefail
__usage="usage: new_project name &lt;type&gt;
type can be one of md, r, or any other... and accordingly it should open
and setup different folder structures
name will be the folder name where the project exists.
"
make_folder () {
mkdir -p "$1"
cd "$1"
}
project_md (){
make_folder "$1"
local today_file="$(date +%F).md"
if test ! -f "$today_file"; then
touch "$today_file"
fi
open -a "/Applications/Visual Studio Code.app/" "$today_file"
}
project_r (){
make_folder "$1"
local today_file="$(date +%F).r"
if test ! -f "$today_file"; then
touch "$today_file"
fi
open -a "/Applications/RStudio.app" "$today_file"
}
# Start logic bellow this line
if [ "$#" != 2 ]; then
echo "$__usage"
exit 1
fi
if [ "$2" = "r" ]; then
project_r $1
fi
if [ "$2" = "md" ]; then
project_md $1
fi
</code></pre>
<html><link href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/styles/default.min.css" rel="stylesheet"/>
<script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.9.0/highlight.min.js"> </script>
<script>hljs.highlightAll();</script></html>
</description>
<pubDate>Fri, 19 Apr 2024 09:06:56 +0000</pubDate>
<guid>https://www.sixhat.net/2023/my-approach-to-managing-scratch-projects-with-bash-scripts.html</guid>
</item>
<item>
<title>You are not a Parrot, and other AI stories</title>
<link>https://www.sixhat.net/2024/07-you-are-not-a-parrot-and-other-ai-stories.html</link>
<description>
<h1 id="you-are-not-a-parrot-and-other-ai-stories"><a class="header" href="#you-are-not-a-parrot-and-other-ai-stories">You are not a Parrot, and other AI stories</a></h1>
<p><a href="https://nymag.com/intelligencer/article/ai-artificial-intelligence-chatbots-emily-m-bender.html?ref=sixhat.net">You are not a Parrot, and a chatbot is not a human</a>, is a very interesting in the sense of accounting for the words one says.
AI doesn't have that because the words for an LLM mean nothing.
They have no referents that will ground them and thus enforce some kind of moral behaviour.</p>
<p><a href="https://werd.io/2024/stripping-the-web-of-its-humanity?ref=sixhat.net">Stripping the web of its humanity</a> is another opinion on how humans are mistreated by AI tools based on LLMs.
AI products are spitting out LLM generated content from user prompts as factual, where most of the time it is not.
In some situations this might be funny, but often it is not.
The lack of attribution (or its demoting to a footer) makes LLMs a potential case of plagiarism, and a bad one at points. </p>
</description>
<pubDate>Fri, 05 Jul 2024 09:10:50 +0000</pubDate>
<guid>https://www.sixhat.net/2024/07-you-are-not-a-parrot-and-other-ai-stories.html</guid>
</item>
</channel>
</rss>
If you would like to create a banner that links to this page (i.e. this validation result), do the following:
Download the "valid RSS" banner.
Upload the image to your own server. (This step is important. Please do not link directly to the image on this server.)
Add this HTML to your page (change the image src
attribute if necessary):
If you would like to create a text link instead, here is the URL you can use:
http://www.rssboard.org/rss-validator/check.cgi?url=https%3A//www.sixhat.net/rss.xml