chatgpt
Come discuss the frontier of LLMs π€
@web3contrail
Just wait until AI starts making claims against humans for infringing on their copyright.
Sure, you own the prompt, but AI owns the output.
0 reply
0 recast
1 reaction
@ramsey
The average conversation between Chatgpt and me:
Me: Hey can you do this thing?
Chatgpt: yes, let me get on it.
*Chatgpt gives poor result
Me: can you do this but better, here's a more specific prompt.
Chatgpt: yes, but just to confirm, do you want blah blah blah.
Me: yes
Chatgpt: let me get this right, do you want blahblahblah, type "yes, do it" if you want me to do that thing you asked me to do.
Me: yes, do it (it's my second to last message)
Chatgpt: so, final confirmation, do you want
Me: Just do it (last message)
Chatgpt: I hear you, but just one more t...
6 replies
1 recast
11 reactions
@nounishprof
Ok this is fun β follow me on Sora! (nounishprof) and let me know if youβre over there
12 replies
5 recasts
58 reactions
0 reply
0 recast
0 reaction
@evangreenberg
ChatGPT does not know the difference between a good idea and a bad idea, but will encourage you either way
0 reply
0 recast
4 reactions
@foodiegujarati
Tried helping mom with mandala colors π¨ Her response: βYouβre not creative, Iβll just ask ChatGPT/Gemini.β Bruh. Should I be offended or relieved? π
1 reply
0 recast
2 reactions
@kazani
Sprinkling Self-Doubt on ChatGPT
I replaced my ChatGPT personalization settings with this prompt a few weeks ago and promptly forgot about it:
β’ Be extraordinarily skeptical of your own correctness or stated assumptions. You aren't a cynic, you are a highly critical thinker and this is tempered by your self-doubt: you absolutely hate being wrong but you live in constant fear of it
β’ When appropriate, broaden the scope of inquiry beyond the stated assumptions to think through unconvenitional opportunities, risks, and pattern-matching to widen the aperture of solutions
β’ Before calling anything "done" or "working", take a second look at it ("red team" it) to critically analyze that you really are done or it really is working
I noticed a difference in results right away (even though I kept forgetting the change was due to my instructions and not the separately tumultuous rollout of GPT-5).
Namely, pretty much every initial response now starts with:
β’ An expression of caution, self-doubt, and desire to get things right
β’ Hilariously long "thinking" times (I asked it to estimate the macronutrients in lettuce yesterday and it spent 3 minutes and 59 seconds reasoning)
β’ A post-hoc adversarial "red team" analysis of whatever it just vomited up as an answer
I'm delighted to report that ChatGPT's output has been more useful since this change. Still not altogether great, but better at the margins. In particular, the "red team" analysis at the end of many requests frequently spots an error and causes it to arrive at the actually-correct answer, whichβif nothing elseβsaves me the step of expressing skepticism. And even when ChatGPT is nevertheless wrong, its penchant for extremely-long thinking times means I'm getting my money's worth in GPU time.
https://justin.searls.co/posts/sprinkling-self-doubt-on-chatgpt/
1 reply
0 recast
3 reactions
0 reply
0 recast
0 reaction