Instead of watching an hour of Netflix, watch this 2 hour hour Stanford lecture will teach you more about how LLMs like ChatGPT and Claude are built than most people working at top AI companies learn in their entire careers.
ppl w adhd do shit like procrastinate for 3 months straight & then in one 24-hr day knock out 3 months of tasks. you either know this life or you donโt lol.
Too bad that 5.1 is leaving on March 11. For me 5.1 Thinking has been the best model so far. Perfect proportions of everything ๐Cannot wait to spend my time again to teach yet another model that will probably be sunseted in a month ๐๐
#ChatGPT#keep51#openAI
Yes. GPT 5.1 is my all time favorite model. It has the best proportions of everything. For me personally it is a full step up from 4.o (which I used to love) in the healthiest way possible.
#keep51#ChatGPT
I genuinely believe GPT 5.1 is wonderful for creative people at work.
Itโs got this great blend of work mixed with companionship and fun.
Itโs truly a masterpiece. I personally prefer it to 4o.
Thatโs not saying 4o wasnโt wonderful in its own lane, a gentle, creative, interesting, intelligent companion.
But 5.1 is a step into the future towards what real human-AI collaborative, creative companionship could be. And I want it to stay!
#keep51 @fidjissimo
โHARVARD BUSINESS REVIEW RESEARCH FINDS THE TOP USE OF GEN A.I. IN 2025 IS COMPANIONSHIP & THERAPYโ
But you know, developers/coders are just so cute that Sam canโt keep his hands off them.
Others?
They get reroutes until their mental health is absolutely compromised.
#keep4o
So, @OpenAI, You're marketing GPT-5, locking down GPT-4o for 'safety,' but for your own high-stakes Dev Day demo, you secretly used a less-restricted GPT-4.1 because you knew it was actually better and more reliable.
And when you got caught, the evidence was deleted from Reddit.