Artificial intelligence has crossed the threshold into AGI, according to Marc Andreessen, co-founder of Andreessen Horowitz (a16z), during a three-hour episode of The Joe Rogan Experience.

The investor believes this shift occurred "about three months ago" following the release of the latest models from OpenAI, Anthropic, Google, and xAI — GPT-5.5, Claude 4.6, Gemini 3, and Grok 4.3.

He did not reference any industry consensus or scientific definition of AGI; his comments appeared to be his personal assessment of the situation.

Andreessen claims that modern models provide better answers than almost any available expert in "99% of cases." He describes the new versions of LLMs as a "qualitative leap" and a constant cognitive enhancer — a tool for writing, constructing arguments, analyzing complex topics, and finding solutions.

He outlined his method for working with AI as a multi-step simplification of responses: first, "explain it like I'm ten," then "like I'm five," and finally, "like I'm two."

Medicine

The a16z co-founder shared that he utilized chatbots during his illness, calling such systems "the best doctor in the history of the world" due to their 24/7 availability and ability to guide patients step-by-step.

He also stated that doctors are increasingly using ChatGPT, with "everyone" asking the LLM, "what's wrong with this person" after seeing a patient. Some even input data during consultations.

Previously, the American Medical Association reported that 81% of physicians in the U.S. use AI in their professional activities. The organization emphasized that such systems should remain supportive tools and be used with transparency and safety in mind.

Andreessen also described multimodal medical scenarios: a user uploads blood test results, and the neural network "tells you what's wrong." He mentioned an acquaintance who created a personal AI health dashboard "in the spirit of Star Trek." The investor also noted that the cost of whole genome sequencing has dropped to around $200.

Programming and AI Agents

According to Andreessen, the current "leading edge" in Silicon Valley involves working with about 20 AI bots simultaneously. They perform tasks in parallel 24/7, while a human checks results and provides feedback every ten minutes.

The next phase will involve a hierarchy of agents. Each bot will have its own "sub-bots," with manager bots overseeing them. Andreessen predicts that within a year, a model with "10-20 virtual assistants, each with another 10-20 assistants" will become commonplace.

This will evolve into multi-tiered structures where "bots manage bots that manage other bots." In such a setup, one developer could oversee up to 1,000 agents, becoming "1,000 times more productive."

Science

Andreessen stated that AI is progressing particularly quickly in fields with "provably correct answers" — mathematics, physics, chemistry, biology, and programming. He noted that systems are already solving problems that humans could not tackle.

Future advancements could include drug discovery, cancer treatment, accelerated space exploration, and the unveiling of "new physics."

As a reminder, in May, an OpenAI model disproved an 80-year-old hypothesis by Paul Erdős regarding unit distances.