Ph.D.-level AI agents could be here soon

A top AI company is preparing to release a Ph.D.-level agent, according to a new Axios report. Dr. Jack R-AI-yan, reporting for duty🫡.

It’s probably OpenAI. CEO Sam Altman is attending a closed-door meeting with US government officials on January 30 to demonstrate new tech he believes will showcase the economic power of AI, per Axios and the NY Times.

What are AI agents?

Artificial intelligence tools that can perform multi-step tasks with minimal human supervision, such as booking travel. Anthropic already launched an agent; Microsoft has unveiled AI tools to automate tasks like sending emails and managing documents; and Google is working on its own AI agent, according to The Information ($).

The potential breakthrough: An AI agent that can perform tasks to accomplish an objective at a level similar to a Ph.D.-possessing human. In a recent Joe Rogan Experience episode, Meta CEO Mark Zuckerberg said he foresees AI agents replacing mid-level software engineers at his company sometime this year.

But…The AI hype train may be outrunning its tracks. Data constraints remain, and models are still prone to hallucinations. Investors and companies also benefit from feeding into hype to juice valuations and secure funding.

📰 In other AI news: OpenAI last week released an Economic Blueprint detailing how artificial intelligence can bring about an era of “re-industrialism” in the US.