<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Biology on IT News</title>
    <link>https://it-news.uk/tags/biology/</link>
    <description>Recent content in Biology on IT News</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Thu, 02 Jul 2026 00:00:00 +0800</lastBuildDate>
    <atom:link href="https://it-news.uk/tags/biology/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>OpenAI&#39;s GeneBench-Pro doesn&#39;t let AI models cheat on biology — and most of them fail</title>
      <link>https://it-news.uk/posts/openai-genebench-pro-biology-benchmark/</link>
      <pubDate>Thu, 02 Jul 2026 00:00:00 +0800</pubDate>
      <guid>https://it-news.uk/posts/openai-genebench-pro-biology-benchmark/</guid>
      <description>&lt;p&gt;OpenAI released a new benchmark Tuesday called GeneBench-Pro, and it is not the kind of test you can cram for. Instead of asking an AI model to recite facts or follow a fixed procedure, it drops the model into a messy, incomplete dataset and asks it to figure out what to do.&lt;/p&gt;&#xA;&lt;p&gt;The idea is simple: real biology research is not a multiple-choice exam. A scientist staring at a genome sequence does not have a clean prompt and four options. They have noise, gaps, and conflicting signals. GeneBench-Pro tries to measure whether AI can handle that reality.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
