OpenAI CEO Sam Altman has written a rare note about his website today, describing in more detail his vision of an AI-powered future, or as he calls it (and the title of his blog post):The era of intelligence.“
Specifically, Altman claims that “deep learning works” and can generalize across domains and difficult problem sets based on its training data, enabling humans to “solve difficult problems,” including “fixing the climate, founding a space colony, and figuring out all of physics.” As he puts it:
“
In a provocative statement that many AI industry participants and keen observers have already picked up on in discussions about X, Altman also said that superintelligence — AI that is “much smarter than humans,”in response to previous statements by OpenAI — may very well be achieved inside “a few thousand days.”
One thousand days is roughly 2.7 years, which is significantly faster than the five years most experts say.
Many AI researchers, especially those at OpenAI, are working on superintelligence, and the lower version is often called artificial general intelligence (AGI). A brand new startup by former OpenAI Chief Scientist and co-founder Ilya Sutskever is even focusing on protected superintelligence.
AI models have began to do well on “IQ tests,” or benchmarks of information, but they are not yet higher than humans. So far, most use cases for generative AI have not been computer programs that are significantly smarter than the average human, but assistants that complement human employees in performing tasks.
Artificial Intelligence Experts for Everyone
However, Altman believes that this kind of use of AI assistants and agents will grow to be commonplace inside a few years.
He added that artificial intelligence will soon enable everyone to attain many goals, as each person will probably be assigned a personal AI team with virtual experts in many fields, and children will have personal tutors in every subject.
It’s no surprise that Altman is an AI maximalist, as he runs one of the leading AI corporations. OpenAI recently released its strongest AI model yet, the o1, which may reason without much instruction from a human.
Altman points out that there are several obstacles to this world of widespread AI adoption, resembling the must make computing cheaper and the availability of advanced chips. He even mentions that without building the infrastructure to support AI development, “AI will be a very limited resource that is fought over and will become primarily a tool for wealthy people.”
Not entirely positive
Altman isn’t entirely enthusiastic about the potential of AI, nevertheless. He notes there will probably be drawbacks, stating:
Altman mentions people losing their jobs as a consequence of AI, something he said earliera cursory nod one of the best fears those outside the technology world bubble.
Altman believes that work under the rule of AI will change each for higher and worse, but humans won’t ever run out of things to do.
Altman’s manifesto comes as no surprise to anyone who has followed OpenAI and generative AI developments over the past few years. But the timing of his musings has led some to consider that this might all be a way for OpenAI to secure one other round of funding. The company apparently it raises $6-6.5 billion, which might value it at $150 billion.
What’s interesting, though, is that Altman selected to post this message on his personal website fairly than the official OpenAI website, suggesting that he treats it as his opinion fairly than the company’s official stance.