<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Thoughts on Steady Monkey</title><link>https://steadymonkey.eu/tags/thoughts/</link><description>Recent content in Thoughts on Steady Monkey</description><generator>Hugo</generator><language>en-gb</language><lastBuildDate>Sat, 24 Jan 2026 09:34:00 +0000</lastBuildDate><atom:link href="https://steadymonkey.eu/tags/thoughts/index.xml" rel="self" type="application/rss+xml"/><item><title>Balancing fast and slow thinking with AI tools</title><link>https://steadymonkey.eu/blog/balancing-fast-and-slow-thinking-with-ai-tools/</link><pubDate>Sat, 24 Jan 2026 09:34:00 +0000</pubDate><guid>https://steadymonkey.eu/blog/balancing-fast-and-slow-thinking-with-ai-tools/</guid><description>&lt;h3 id="the-gist-of-it"&gt;The gist of it&lt;/h3&gt;
&lt;p&gt;Many people said it, and I did too, but without &lt;em&gt;really feeling it&lt;/em&gt;.
You. still. need. to. think. for. yourself.&lt;/p&gt;
&lt;h3 id="the-longer-form"&gt;The longer form&lt;/h3&gt;
&lt;p&gt;I take it as a benefit from the current generation of AI tooling, how working with so-called &lt;em&gt;agentic AI&lt;/em&gt; makes you (eventually) take a step back and ask yourself: &lt;em&gt;do I need to go fast or slow?&lt;/em&gt;&lt;/p&gt;
&lt;p&gt;An example of being fast would be:&lt;/p&gt;</description></item><item><title>My hobby? Speculative Engineering</title><link>https://steadymonkey.eu/blog/my-hobby-speculative-engineering/</link><pubDate>Sat, 13 Dec 2025 21:26:49 +0000</pubDate><guid>https://steadymonkey.eu/blog/my-hobby-speculative-engineering/</guid><description>&lt;p&gt;I was thinking about my &lt;a href="https://famlab.net"&gt;Famlab&lt;/a&gt; project this morning—the &amp;ldquo;homelab for busy families&amp;rdquo; idea where you reclaim data ownership (dare I mention #TechSovereignty?) without becoming a full-time sysadmin.
Specifically, I was thinking about the maintenance of the base system and all services: the update path.
It is an old idea that spawned from A/B updates in Android (where you have a second partition on which you install system updates, and then try to boot from afterwards; if it fails, you mark it as such and reboot into the other, older system partition) and the appeal from immutable operating systems—reproducible, easier deployments, safer.&lt;/p&gt;</description></item><item><title>On inequality and AI risks</title><link>https://steadymonkey.eu/blog/on-inequality-and-ai-risks/</link><pubDate>Sun, 07 Dec 2025 21:03:12 +0000</pubDate><guid>https://steadymonkey.eu/blog/on-inequality-and-ai-risks/</guid><description>&lt;p&gt;Friday morning I read &lt;a href="https://jennykudymowa.substack.com/p/how-the-very-rich-make-the-world"&gt;Jenny&amp;rsquo;s piece on how the ultra-rich make the world worse&lt;/a&gt;.
Friday evening I attended a rationality meetup about &lt;em&gt;&lt;a href="https://ifanyonebuildsit.com"&gt;If Anyone Builds It, Everyone Dies&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20260529023311/https://ifanyonebuildsit.com" title="2026-05-29 archived copy of https://ifanyonebuildsit.com" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
&lt;/em&gt;.
Sunday afternoon, an on-site &lt;a href="https://pauseai.info"&gt;PauseAI&lt;/a&gt; meetup.&lt;/p&gt;
&lt;p&gt;Three events, one weekend.
Some clarity in the end.&lt;/p&gt;
&lt;p&gt;If you haven&amp;rsquo;t read the book mentioned above, or &lt;a href="https://steadymonkey.eu/blog/two-weeks-with-the-extinction-tool/"&gt;read my previous reflections on the matter&lt;/a&gt;, let me quote them directly:&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;If any company or group, anywhere on the planet, builds an artificial superintelligence using anything remotely like current techniques, based on anything remotely like the present understanding of AI, then everyone, everywhere on Earth, will die.&lt;/p&gt;</description></item><item><title>Two Weeks with the Extinction Tool</title><link>https://steadymonkey.eu/blog/two-weeks-with-the-extinction-tool/</link><pubDate>Sat, 25 Oct 2025 12:22:07 +0000</pubDate><guid>https://steadymonkey.eu/blog/two-weeks-with-the-extinction-tool/</guid><description>&lt;blockquote&gt;
&lt;p&gt;I read a book that said AI might kill everyone. I&amp;rsquo;m now using it as a daily tool and &lt;em&gt;enjoying it&lt;/em&gt;, and I don&amp;rsquo;t know what that makes me.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;Two weeks ago I read &lt;a href="https://ifanyonebuildsit.com"&gt;&amp;ldquo;If Anyone Builds It, Everybody Dies&amp;rdquo;&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20260529023311/https://ifanyonebuildsit.com" title="2026-05-29 archived copy of https://ifanyonebuildsit.com" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
.
I could follow the reasoning.
There&amp;rsquo;s no surprise in the book, the title is &lt;em&gt;it&lt;/em&gt;.
The arguments appear serious enough, and that, even if the authors were inflating the risk, we would still be looking at a sheer cliff humanity is bound to fall off.
So, what if they are right?&lt;/p&gt;</description></item><item><title>How I successfully used an LLM, or my Simon Willison moment</title><link>https://steadymonkey.eu/blog/how-i-successfully-used-an-llm-or-my-simon-willison-moment/</link><pubDate>Fri, 26 Apr 2024 13:26:03 +0000</pubDate><guid>https://steadymonkey.eu/blog/how-i-successfully-used-an-llm-or-my-simon-willison-moment/</guid><description>&lt;blockquote&gt;
&lt;p&gt;&lt;a href="https://simonwillison.net/about/"&gt;Simon Willison&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20240410150511/https://simonwillison.net/about/" title="2024-04-10 archived copy of https://simonwillison.net/about/" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
 has a legendary throughput.
No, seriously.
Not only is he a co-creator of &lt;a href="https://www.djangoproject.com/"&gt;Django&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20240425234237/https://www.djangoproject.com/" title="2024-04-25 archived copy of https://www.djangoproject.com/" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
, he has been publishing interesting software (see &lt;a href="https://datasette.io/"&gt;Datasette&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20240422130818/https://datasette.io/" title="2024-04-22 archived copy of https://datasette.io/" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
 and &lt;a href="https://simonwillison.net/tags/datasette/"&gt;its gazillion plugins&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20240327131226/https://simonwillison.net/tags/datasette/" title="2024-03-27 archived copy of https://simonwillison.net/tags/datasette/" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
) at an outstanding pace.
More recently he has taken the task of making LLM&amp;rsquo;s easy to interact from the CLI with its &lt;a href="https://llm.datasette.io/en/stable/index.html"&gt;&lt;code&gt;llm&lt;/code&gt;&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20240422073247/https://llm.datasette.io/en/stable/index.html" title="2024-04-22 archived copy of https://llm.datasette.io/en/stable/index.html" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
 program.
In today&amp;rsquo;s post, I write about my first real success with an LLM.
I shall update my subtitle to &amp;ldquo;He who prompted okay once&amp;rdquo;.&lt;/p&gt;</description></item><item><title>On Domain Events, English, and ubiquitous language</title><link>https://steadymonkey.eu/blog/on-domain-events-english-and-ubiquitous-language/</link><pubDate>Wed, 17 Apr 2024 12:31:37 +0000</pubDate><guid>https://steadymonkey.eu/blog/on-domain-events-english-and-ubiquitous-language/</guid><description>&lt;p&gt;I attended a Python meetup last week, hosted by &lt;a href="https://jobrad.org/"&gt;JobRad&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20240415165625/https://jobrad.org/" title="2024-04-15 archived copy of https://jobrad.org/" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
.
The main presentation was a showcase of an &amp;ldquo;Event System&amp;rdquo;, or how they used messages to share customer data between their ERP (&lt;em&gt;Enterprise Resource Planning&lt;/em&gt; , the big database of their processes, inventory, and all other business activities) and their CRM (&lt;em&gt;Customer Relationship Management&lt;/em&gt; , the main piece of software required by customer support with sales orders and support tickets).&lt;/p&gt;</description></item><item><title>I got in the Framework Mainboard Developer Program</title><link>https://steadymonkey.eu/blog/i-got-in-the-framework-mainboard-developer-program/</link><pubDate>Sat, 21 May 2022 13:58:09 +0000</pubDate><guid>https://steadymonkey.eu/blog/i-got-in-the-framework-mainboard-developer-program/</guid><description>&lt;p&gt;Two weeks ago, I received an e-mail telling me I had been accepted in &lt;em&gt;the Framework Mainboard Developer Program&lt;/em&gt;.
&lt;a href="https://frame.work/"&gt;Framework&lt;/a&gt; 



&lt;a href="https://web.archive.org/web/20220521154919/https://frame.work/" title="2022-05-21 archived copy of https://frame.work/" class="archive-link" target="_blank"&gt;(archived)&lt;/a&gt;
 has a very interesting approach to sustainability in electronics that reminds me of Fairphone, though I don&amp;rsquo;t want to compare them directly.
I worked at Fairphone for more than three years at the time of Fairphone 2, and the modular aspect of the phone was amazing.
I have been following Framework since they launched a crowdfunding campaign for their first generation modular laptop, it&amp;rsquo;s a really positive outcome that they are still there and made it to the second generation.&lt;/p&gt;</description></item></channel></rss>