Why GPT-5 Could Hurt OpenAI's Future
🔮 This time it feels different for OpenAI.
Hey Everyone,
Covering OpenAI is pretty amusing lately, and I hope you are getting something out of my coverage.
Today I really want to focus on GPT-5. GPT-5 in many ways could be the make or break moment for OpenAI’s ability to lead the world in the coming months and years.
The pressure is on for OpenAI with GPT-5, with Bard (Gemini Pro) overcoming every other GPT-4 version and it's now sitting at the 2nd spot in the LLM race on the somewhat ridiculous leaderboards that engineers like to watch, OpenAI is only as good as its last dominant model.
So what about GPT-5?
GPT-4 Turbo’s dominance likely won’t last the assault of Google Gemini Ultra, and we’ll know soon enough. Either way with GPT-5 likely a few months out, the debate and speculation is increasing around it and its new capabilities. GPT-5’s launch and the position OpenAI is in also deserves some of our attention.
Sam Altman in an interview with Axios about a week ago said that he believes future AI products will need to allow "quite a lot of individual customization" and "that's going to make a lot of people uncomfortable.” Yes training LLMs on our personal data texts is not a very nice thought privacy wise.
Sam Altman’s interview with Bill Gates a few months back but aired recently gives more specific info on GPT-5 than we have gotten in a while.
There’s speculation on X that GPT-5 doesn’t come out until Q2, 2024 which begins in April and ends in July, 2024. That makes sense since Sam Altman has gone on his usual charm offensive about it, roughly six months before its anticipated release.
What’s fascinating is in 2024, Llama-3 and Mistral might be as impressive as Google Gemini Ultra in their actual impact on Generative AI’s overall development and enterprise adoption. We mostly know what to expect with GPT-5 and its impact might be considerably less because of everything else that will be going on. GPT-5 won’t ever have the hype that GPT-3.5 or GPT-4 had because., they felt new.
That’s not to say that GPT-5 won’t be considerably better than GPT-4 Turbo. However the law of diminishing returns will be felt with GPT-5 in my estimation in a significant way.
The hype around OpenAI got out of hand in recent years, and I think GPT-5 is that moment it all comes down to Earth. With that being said, I could be totally wrong in my analysis and speculation. Still sometimes it’s enjoyable to speculate on matters like GPT-5 and the viability of OpenAI.
The sad part is a lot of the articles I read about GPT-5 this week, appear to be ChatGPT generated work. What does that tell you?
I can’t help but notice since the arrival of ChatGPT, the internet in the last six months has reduced in quality by the widest margin ever recorded. This is directly thanks to Sam Altman and OpenAI’s innovation. So how might GPT-5 change the future of civilization?
It’s a big question. How advanced can it possibly be?
Keep reading with a 7-day free trial
Subscribe to AI Supremacy to keep reading this post and get 7 days of free access to the full post archives.