<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>AI on The Signal — Blog by François Fournier</title><link>https://blog.francoisfournier.eu/tags/ai/</link><description>Recent content in AI on The Signal — Blog by François Fournier</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sun, 16 Aug 2026 10:00:00 +0200</lastBuildDate><atom:link href="https://blog.francoisfournier.eu/tags/ai/index.xml" rel="self" type="application/rss+xml"/><item><title>Six of eight AI projects should have been a Python script</title><link>https://blog.francoisfournier.eu/articles/ai-projects-that-should-have-been-a-script/</link><pubDate>Sun, 16 Aug 2026 10:00:00 +0200</pubDate><guid>https://blog.francoisfournier.eu/articles/ai-projects-that-should-have-been-a-script/</guid><description>&lt;p>Startup pitch: &amp;ldquo;We use AI to automate X.&amp;rdquo; Investors: &amp;ldquo;Wonderful. Here is €2M.&amp;rdquo;&lt;/p>
&lt;p>Six months later: an AI budget of €120,000 — data scientists plus compute — and a model at 78% accuracy. Benchmark of a 200-line Python script: 74%.&lt;/p>
&lt;p>Difference: four points. Cost: €120,000.&lt;/p>
&lt;p>(If you think your AI project is different — it might be. The odds are against you.)&lt;/p>
&lt;hr>
&lt;h2 id="what-eight-audits-showed">What eight audits showed&lt;/h2>
&lt;p>Project one, support ticket classification. Solution sold: fine-tuned LLM. Solution that works: regex plus ten if/else rules. Saved: €60,000.&lt;/p></description></item></channel></rss>