Recent searches
Search options
A demonstration of offline and online LLM image description responses
I have one image, and two very differing descriptions of said image. The first is from #Llava13B, a local model installed on my #Mac, and the second is from #GPT. The GPT response is so huge that I'm going to have to post it as threaded replies, but it shows the vast difference between models. For this first post, the Llava13B response has been included as #AltText.
1/
@FreakyFwoof Very detailed and explanatory description there. I like it.
I’ve been noticing that I’ve been arguing with AI quite regularly lately. Yesterday it blatantly told me I was wrong in telling them what the object in the photo was. All I was trying to do was get info about whether the entire object was in the frame, and if specific elements were clearly visible, so that I knew it was fine to share.
@cassana But you must be wrong because you're only human, and AI is AI, right? Right? 
@FreakyFwoof Right! It’s that almighty superbrain that has access to the whole universe of information known as the World Wide Web of Lies. But it’ll tell you so politely: I’m sorry, but that object isn’t in the photo”, with an implied very Soto voce “you ignorant monkey”.