To receive industry-leading AI updates and exclusive content, sign up for our daily and weekly newsletters. Learn more
OpenAI CEO Sam Altman said: Website Today, I’ll go into more detail about his vision for an AI-powered future, as he calls it (the blog post title is “The age of intelligence.“
Specifically, Altman argues that “deep learning works” and can generalize to different domains and hard problem sets based on training data, allowing it to “solve hard problems” like “fixing the climate, founding space colonies, discovering all of physics.” He states:
“Exactly. Humanity has discovered an algorithm that can truly learn any data distribution (or the underlying ‘rules’ that produce any data distribution). With astonishing accuracy, and the more computation and data we have available, the better our ability to help people solve hard problems. No matter how much time we spend thinking about this, I don’t think we can truly grasp how important it is.”“
In a provocative statement that has already been picked up by many AI industry insiders and observers in the X debate, Altman also spoke about the notion of superintelligence, or “Much smarter than humans,Previous OpenAI statements said this goal could be achieved within “a few thousand days.”
“This may be the most important thing that has ever happened in history. Superintelligence could happen within a few thousand days! It may take longer, but I am confident we will get there.”
1,000 days is roughly 2.7 years, much less than the five years most experts estimate.
Many AI researchers, particularly those at OpenAI, are pursuing superintelligence, a lesser version of which is usually called artificial general intelligence (AGI). A new startup from former OpenAI chief scientist and co-founder Ilya Sutskever is also focusing on safe superintelligence.
While AI models are starting to perform well in “IQ tests” and knowledge benchmark tests, they are not yet better than humans. So far, most use cases of generative AI have been as assistants to assist human workers in completing tasks, rather than as computer programs that are much smarter than the average human.
AI Expert for Everyone
But Altman believes this use case for AI assistants and agents will become widespread within a few years.
“There are many details that still need to be worked out, but it would be a mistake to get distracted by any particular challenge,” Altman said. “Deep learning is working, and the remaining problems will be solved. There’s a lot to say about what happens next, but what’s most important is that AI will get even better as it scales, and it will bring meaningful improvements to people’s lives around the world.”
He further added that AI will enable everyone to achieve a lot in the near future as each person will have their own personal AI team of virtual experts in different fields and children will get personal tutoring in every subject.
It’s no surprise that Altman is an AI maximalist, given that he runs one of the leading AI companies: OpenAI recently released o1, the most powerful AI model to date, which can reason without much human direction.
Altman noted that there are several obstacles to widespread use of AI, including the need to make computing cheap and the availability of advanced chips, and further suggested that if we don’t build the infrastructure to support AI development, “AI will become a very limited resource, and wars will be fought over it, and it will mostly be a tool of the rich.”
Not entirely positive
But Altman isn’t completely optimistic about the potential of AI. He points out that AI also has a downside, saying:
“It’s not an entirely positive story, but the upsides are so great that we owe it to ourselves and to the future to think about how to navigate the risks that are in front of us.”
Altman talks about people losing their jobs to AI. What he said beforenodded briefly. One of the biggest fears People outside the tech world bubble.
For Altman, work will change under AI for better or worse, but people will never run out of things to do.
Altman’s declaration won’t come as a surprise to anyone who’s followed the growth of OpenAI and generative AI over the past few years, but the timing of his musings led some to believe this was a way for OpenAI to secure its next round of funding. According to the report: The bid is for $6 billion to $6.5 billion, giving the company a valuation of $150 billion.
However, it is interesting that Altman chose to post this message on his personal website rather than the official OpenAI website, suggesting that he views this as his own opinion rather than official company policy.
