- The first adopters are much better placed to win – “do it,” says Sam Altman
- Today, AI is an intern, but soon it will be an experienced worker capable of tackling complex problems
- The next step is acted, but does that mean accelerated scientific discoveries, self-improvement or something else?
The CEO of Openai, Sam Altman, offered his honest ideas on the EVA landscape of AI, comparing previous, current and future trends to point to companies in the right direction.
“I think, do it,” said Altman in an interview with the CEO of Snowflake, Sridhar Ramaswamy, during the company Summit 2025 of the company, stressing that companies pending the next risk of IA breakthrough are left behind.
“When things change quickly, companies with the fastest adoption speed … win,” he added.
The AGE definition changes
Thinking about how his advice has changed in the past year, Altman has noted a major change in AI’s abilities in this short time.
“We have reached a real inflection point for the use of these models,” he said, but “an interesting question is what we say different next year”.
For the future, Altman said that he is considering companies adopting AI not only to automate routine processes, but to take up larger challenges.
In 2026, he predicted that Will’s companies “throw a ton of calculation” to “an extremely important problem[s]”And” models will be able to find things that people of people cannot do. “”
Altman compared today to “an intern which can work for a few hours”, but plans a near future where AI looks like an experienced or elderly worker. The founder of Openai predicts that “in certain limited cases [and] A few ways, “we could see agents capable of discovering new knowledge very soon, and use cases should only increase with additional developments.
Asked about general artificial intelligence (AGA), Altman explained that humans are very good as adjusting their expectations, which questions the static definitions of the term. By transferring his mind for five years, he said that most people could have considered today’s catpt as act, but our expectations today have strengthened.
“The thing that counts is the rate of progress you see from one year to the next in the past five years should continue for at least the next five, probably far beyond,” he added, suggesting that we are far from the summit.
His favorite objective for AGA would be to increase considerably, perhaps “quadruple”, a scientific discovery, but he also recognizes that others see acted as a system capable of self-improvement, potentially even consciousness.
In conclusion, Altman was asked a hypothetical question – what would he do if he had 1,000x more calculation?
“I would ask him to work very hard on the research on AI, to find out how to build much better models, then to ask for this much better model what we should do with all the calculation,” he said.
Offering us an overview of the not too distant future of Openai, Altman revealed that “the models in the next year, two years, will be quite breathtaking … There is a lot of progress to come”.
