OpenAI celebrated ChatGPT’s second anniversary in a big way a few weeks ago. The company held a massive “12 Days” press event, with each day bringing a new AI feature to various products. ChatGPT was a big focus and received various interesting upgrades in December, including a next-gen reasoning model (o3) that should see a wider release in the near future.
However, OpenAI did not announce an upgrade for the main model most people use. That’s GPT-4o, which supports Advanced Voice Mode and multimodality. The model was launched in May, so it shouldn’t feel like we need something newer and more powerful.
However, the AI industry is moving at break-neck speeds compared to the rest of the tech landscape, and we’ve come to expect innovations to drop at a faster rate. Google is doing the same thing; everyone is.
It’s clear that OpenAI doesn’t have GPT-5 ready for a teaser announcement. Reports say the firm is struggling with training the AI and is running behind a self-imposed schedule.
Tech. Entertainment. Science. Your inbox.
Sign up for the most interesting tech & entertainment news out there.
Email: SIGN UP
By signing up, I agree to the Terms of Use and have reviewed the Privacy Notice.
Rather than detailing any of that, Sam Altman is doing what you’d probably expect from the CEO of a for-profit company. He’s teasing the near (or more distant) future of AI by talking about AGI (artificial general intelligence) and the next step after that, AI superintelligence.
Sam Altman penned a blog post on his website titled Relections, where he mused on the brief history of ChatGPT and where OpenAI is heading. AGI came up right away:
As we get closer to AGI, it feels like an important time to look at the progress of our company. There is still so much to understand, still so much we don’t know, and it’s still so early. But we know a lot more than we did when we started.
Towards the end of the blog, Altman got back to AGI and superintelligence, saying OpenAI is confident about creating AGI. “We are now confident we know how to build AGI as we have traditionally understood it,” the exec said. However, AGI won’t necessarily come this year. Instead, AI agents might be the killer new feature of AI like ChatGPT:
We believe that, in 2025, we may see the first AI agents “join the workforce” and materially change the output of companies. We continue to believe that iteratively putting great tools in the hands of people leads to great, broadly-distributed outcomes.
According to reports, AI agents aren’t necessarily tied to the GPT-5 upgrade. These are different development projects at OpenAI.
As for how OpenAI understands AGI, TechCrunch points out two different explanations.
First, we have the company’s own definition for the term: “highly autonomous systems that outperform humans at most economically valuable work.” Then there’s a joint OpenAI-Microsoft definition of AGI: AI that can generate at least $100 billion in profits. When that happens, Microsoft will no longer have access to OpenAI tech. Microsoft is one of the biggest investors in OpenAI.
Back to Altman’s blog, he touches on the step after AGI, superintelligence. That’s AI so advanced it will exceed what humans are capable of:
We are beginning to turn our aim beyond that, to superintelligence in the true sense of the word. We love our current products, but we are here for the glorious future. With superintelligence, we can do anything else. Superintelligent tools could massively accelerate scientific discovery and innovation well beyond what we are capable of doing on our own, and in turn massively increase abundance and prosperity.
The CEO acknowledged that superintelligence sounds like sci-fi right now, but it could happen in a few years. “We’re pretty confident that in the next few years, everyone will see what we see, and that the need to act with great care, while still maximizing broad benefit and empowerment, is so important.”
Altman also said his blog post was inspired by a recent interview with Bloomberg, where he addressed some of the same topics, including AGI and superintelligence.
All of that sounds exciting, yes. It’s also sci-fi. And if you’re familiar with OpenAI bleeding AI safety engineers whose jobs were to rein AI in, you’ll also worry a bit about a superintelligence ChatGPT being aligned with humanity’s best interests. After all, many people, including AI experts, worry that AI could destroy humanity.
I’d be more interested in what comes next from ChatGPT than Altman’s musings. That includes upgrades for GPT-5, a release timeline, and safety features OpenAI is developing for the models that will precede AGI and superintelligence.
The post OpenAI CEO Sam Altman talks AGI, AI workers, and AI superintelligence appeared first on BGR.
