“Concerned” OpenAI to Consult with U.S. Government About Advanced “PhD-Level” Super AI Capable of Replacing Humans

AI is eyeing everyone’s jobs, from researchers to doctors, or at least that’s the latest report making the rounds.

Over recent months, Sam Altman and other big names in the AI world have been hyping up the idea of agentic AI. There’s growing concern that a super artificial intelligence could soon upend labor markets, maybe even in just a couple of years—or sooner, if the loudest voices have it right. The real worry is that neither governments nor society are ready to handle the massive changes this might bring.

A new report from Axios suggests this potential future may be nearer than many suspect.

Just recently, tech moguls like Mark Zuckerberg at Meta and the CEO of Salesforce have talked about cutting back on hiring, planning to replace human roles with AI. Axios now claims that OpenAI is making headway with its agentic AI project. This form of AI focuses on highly specialized tasks rather than general ones. So far, errors and hallucinations have made it less than reliable for anything significant, but that might soon change.

According to the report, Sam Altman, CEO of OpenAI, will present “PhD-level” artificial intelligence agents to the U.S. government in a private meeting on January 30, 2025. If these AI protocols advance enough to refine their own tasks—akin to what a skilled human might do—they could take on roles far more critical than they currently hold.

Axios describes a mix of excitement and apprehension among OpenAI staff, based on private conversations reported by friends. Although the report is scant on details, the full announcement from OpenAI could be just around the corner.

On the flip side, Microsoft is pushing hard to bring agentic super AI into the mainstream.

Microsoft’s partnership with OpenAI means it’s heavily involved, providing the infrastructure for these AI advancements. Despite some reported bumps in their collaborative journey, Microsoft relies on OpenAI’s language models for its own AI services.

As always, it’s tricky to tell what’s hype and what’s fact in reports about OpenAI. But if even a fraction of this proves true, major societal shifts could be coming our way in the near future.

Firms like Microsoft are sprinting to incorporate super agentic AI into everyday use, aiming to replace costly and time-consuming tasks that humans currently handle. Microsoft Azure already offers some agentic AI services, though they are still fairly basic, often geared towards customer support rather than more complex endeavors. Truth be told, today’s AI hasn’t exactly excelled in many of its assigned tasks.

For instance, Microsoft Copilot can help polish your Excel spreadsheets today, but in the future, it might be able to create detailed spreadsheets for intricate tasks at your command. Looking forward, it could build entire software suites customized for you, perhaps even alter the Windows experience or generate on-the-fly Xbox games. Imagine telling Copilot, “make me a Minecraft map that feels like a survival horror game,” and having it delivered to you. Sources indicate Microsoft is exploring how these advanced language models might totally replace the current 3D graphics pipeline. This could mean an AI rendering a game like Skyrim directly on your screen, reacting dynamically to your inputs, without needing the game installed in the traditional sense. Eliminating latency issues frequent in current cloud gaming could be on the horizon, with some video processing moved locally to give a smoother, almost natural gameplay experience.

Nevertheless, elected leaders are often tech-challenged, potentially unprepared for the sweeping changes this technology could bring. While there’s a fortune to be made, especially for investors backing the right companies, there’s also the risk of severe societal disruption if proper measures aren’t in place. These tech breakthroughs could lead to a brighter future, but they also threaten to displace the vulnerable. Our current economic model relies on human work and intervention. What happens when there’s no more work? We might find out soon enough.

And just to be clear, no AI had a hand in crafting this article.