https://pkradiator.github.io/tags/ai/Recent content in AI on Radiator's ThoughtsHugoen-usFri, 10 Jul 2026 18:57:27 +0000https://pkradiator.github.io/posts/predictions-ai/Fri, 10 Jul 2026 18:57:27 +0000https://pkradiator.github.io/posts/predictions-ai/<p>For the past few weeks I&rsquo;ve been thinking about transformers<sup id="fnref:1"><a href="#fn:1" class="footnote-ref" role="doc-noteref">1</a></sup> and there limitations (also propelled by my job which has me working closely with mlsys). Before I give you my take on the current state of AI let me give a quick disclaimer.</p> <blockquote> <p><strong>Disclaimer:</strong> While I know the architecture and workings of neural networks in general, and transformers in particular, I am no <a href="https://en.wikipedia.org/wiki/Mechanistic_interpretability">Mechanistic Interpretability</a> expert or even versed enough ,to my liking ofcourse, ahmm (Imposter), to be making this claims.</p>