‹ BackHN Continuity

Thread

Vote on which of Hacker News' challenges for AI have been met

202 points · 271 comments · stabbles

  1. Retr0id · · focus · HN ↗
    Heh, there&#x27;s one of mine: <a href="https:&#x2F;&#x2F;stoppels.ch&#x2F;goalposts&#x2F;?c=39727943" rel="nofollow">https:&#x2F;&#x2F;stoppels.ch&#x2F;goalposts&#x2F;?c=39727943

    &quot;GPT-4 looks at original ASCII art of a foot, not copied from the web, and says it is a foot.&quot;

    The vote is currently 64% yes, 18% no.

    Just now I asked Opus 5.5 to generate an ASCII art foot, and it did a passable job. It&#x27;s not great, but it&#x27;s a foot. Then I pasted it into ChatGPT (whatever they&#x27;re serving to the free tier by default, which seems to be 5.6 Luna), and it said it was a &quot;train&#x2F;locomotive&quot;: <a href="https:&#x2F;&#x2F;chatgpt.com&#x2F;share&#x2F;6abeaa39-cc80-83ed-851f-29370db08964" rel="nofollow">https:&#x2F;&#x2F;chatgpt.com&#x2F;share&#x2F;6abeaa39-cc80-83ed-851f-29370db089...

    Maybe it&#x27;s Opus&#x27;s fault for drawing a bad foot but I think it&#x27;s fair to say LLMs are still pretty bad at ASCII art (without additional tool calling etc).

    1. asdfasgasdgasdg · · focus · HN ↗
      Opus 5.5 was able to parse and understand an ASCII art foot when I pasted one in.
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.