Sam Altman, the ever-optimistic CEO of OpenAI, has once again brought attention to a new and powerful artificial intelligence model. While AI continues to evolve rapidly, the debates and controversies surrounding its use and impact remain as strong as ever. Altman, speaking on podcast “This Past Weekend” with host Theo Von — reported by the Daily Times —expressed his belief that the GPT-5 model, which the co-founder claims will be released soon, is more “truly scary.” At one point in the conversation, he even referred to the next-generation AI as the Manhattan Project — the effort during World War II that resulted in the first atomic bomb.
It is worrying particularly because GPT-5 is due to be released next month and has prompted a response from within the tech industry and from ChatGPT users who use the tool to receive assistance for a variety of applications in their everyday lives.
Altman did not disclose many technical details of GPT-5, but he said the model feels “much faster” and “much more powerful” than its previous iterations. “There are times in the history of science when a bunch of scientists look at their work, at what they have wrought, and say — ‘What have we done?” He said. One line proved to be enough for listeners to be left intrigued and concerned.
Altman, who in 2019 took over as chief of OpenAI, also noted that the pace of AI development is so rapid that regulatory systems are having a hard time keeping up. “It does not seem to work; we do not know how to manage complexity,” he added, explaining that oversight has become more pointless as AI has increased.
Some will see these comments as part of his strategic positioning, to get the buzz going ahead of GPT-5’s launch, while others will accept that the man was genuinely taking time to issue a warning. It is not the first time Altman has been a bit queasy about the speed at which A.I. is progressing — but likening it to the Manhattan Project is certainly its most assertive statement to date.
So, what would GPT-5 look like? Although some features are still under wraps, initial reports indicate that it will have a significantly longer memory — so it can better understand what was said in the past and have better context over time. It is also designed to be better at multi-step tasks, so it should be particularly handy for building complicated workflows that have a bunch of connected steps.
And to top it off, GPT-5 is set to take multimodal input to an entirely new level. They will be able to interact with the model using text, voice, as well as images, graphics, and files — combining different modes of conversation, fluidly. That is, it will not just be faster, but smarter, more general than our current machines, able to handle a variety of different types of data at once.
But all this excitement is being dimmed by some very serious questions, both about what happened just ahead of the serum announcement and how politics might be shaping the distribution plan. If not even the person responsible for developing it is certain that technology has not gone too far, perhaps we should all take a step back and wonder: Just what are we doing with AI? Or could it simply be the next natural step in our digital evolution?
If all this is just Altman and the rest of us worrying over nothing, then surely one thing to take away is that GPT-5 is on the way, and when it arrives it will mark a new era in the world of AI – positive or negative.
Feel free to check back on Ginews for updates on GPT-5, for AI, and for what it all means for our future.
