4 comments

  • sigmar1 hour ago
    &gt;Evaluators validate each tag individually — for example, protein, preparation, or health, individually rather than judging the item as a whole.<p>Am I reading this right that the jury is multiple LLMs each iterating through each tag and voting on each? Why wouldn&#x27;t you tune one LLM to be really competent at a single tag? Like a single &quot;spicy evaluator LLM&quot; or &quot;protein evaluator LLM&quot;?
    • vector_spaces1 hour ago
      I am guessing there are so many possible tags that this wouldn&#x27;t be pragmatic
  • vector_spaces1 hour ago
    I am sorry to be harsh but I find it amateurish that they would use an AI generated hero image for this and presumably fabricated LLM output -- fabricated by an AI image generator no less<p>Whenever I create an image like this for the purpose of a demo, I make certain that it demonstrates either real input&#x2F;output or at least is exemplary of real input&#x2F;output because the whole point is to instill confidence in the tool. Sure, if the raw outputs aren&#x27;t clean&#x2F;comprehensible enough for presenting to stakeholders or others, fine, clean them up to make them comprehensible or add explainers, but there shouldn&#x27;t be any need to fabricate the inputs.<p>I feel obligated to respond to the hypothetical &quot;But they don&#x27;t want to tie it to a particular restaurant or brand&quot; -- you don&#x27;t have to! Doordash has taken generic food photos for this exact purpose.
  • TeeWEE1 hour ago
    Basically it’s AI on top of AI for metadata extraction.<p>There are a lot of claims in the article but not a lot of hard data. In the end they still don’t know if the data is correct.<p>Good luck with your glutes allergy.<p>The weird thing for me is the prompt optimization loop? Why not fine tune the model instead of AI generating the prompt?
    • aix129 minutes ago
      Also, &quot;healthy&quot; as a boolean flag is, franky, a bit of a joke.
    • IanCal48 minutes ago
      &gt; The weird thing for me is the prompt optimization loop? Why not fine tune the model instead of AI generating the prompt?<p>Why is it weird to optimise the prompt? Whether you optimise the model is a separate issue.<p>If you use any closed models you can’t fine tune them, which is another reason for most but here they also fine tuned models.
  • hansmayer2 minutes ago
    [dead]