OpenAI has put 80 to 90 percent of its research effort behind whatever comes after GPT-7.
Boris Power, head of applied research, recently said that 80% to 90% of the company’s research is going toward GPT-7, GPT-8, and even further ahead, because “most of the value is there.”
As for small, in-cycle upgrades like 5.1 to 5.2, he says internally they’re seen as “deliberate short-term bets.”
He also mentioned that today’s biggest challenge isn’t that the models aren’t strong enough—it’s that most users don’t know what AI can do. GPT-6 is more like a colleague who can directly take on tasks.
Betting on the next generation also means compute spending will keep rising, and the stories about chips and power haven’t been told yet. Guess what year GPT-7 will be released?
Give me a year. #AI
Boris Power, head of applied research, recently said that 80% to 90% of the company’s research is going toward GPT-7, GPT-8, and even further ahead, because “most of the value is there.”
As for small, in-cycle upgrades like 5.1 to 5.2, he says internally they’re seen as “deliberate short-term bets.”
He also mentioned that today’s biggest challenge isn’t that the models aren’t strong enough—it’s that most users don’t know what AI can do. GPT-6 is more like a colleague who can directly take on tasks.
Betting on the next generation also means compute spending will keep rising, and the stories about chips and power haven’t been told yet. Guess what year GPT-7 will be released?
Give me a year. #AI