Marc Andreessen Criticized for Flawed AI Prompt Showing Misunderstanding of LLM Limitations

Venture capitalist and prominent Trump supporter Marc Andreessen—whose 2023 “techno-optimist manifesto” helped fuel a years-long AI boom—appears to lack a clear understanding of how artificial intelligence actually works. His confusion came to light after he shared a lengthy “custom prompt” on Monday via a tweet, intended to showcase his AI expertise. Instead, the post became a viral target of mockery.

Andreessen’s prompt began with an unusually flattering instruction to the AI:

“You are a world class expert in all domains. Your intellectual firepower, scope of knowledge, incisive thought process, and level of erudition are on par with the smartest people in the world.”

The billionaire’s tone drew immediate ridicule, but the most pointed criticism centered on his demand that the AI “never hallucinate or make anything up.” This request underscores a critical misunderstanding: hallucinations are an inherent limitation of large language models (LLMs), not a flaw that can be corrected by flattery or instruction. Simply asking an AI to avoid errors does not eliminate the issue.

Journalist Karl Bode mocked the prompt in a Bluesky post, writing:

“Yes, you can just demand that the LLM not make errors. That’s definitely how the technology works.”

“I know this isn’t a unique observation but these gentlemen are in absolutely no way remarkable outside of their good fortune.”

Another user commented:

“Marc Andreessen putting ‘you are a world class expert in all domains’ and ‘don’t hallucinate’ in his custom prompt really demonstrates the calibre of the people steering the ship.”

In a scathing critique for Defector, editor Alberto Burneko argued that Andreessen’s AI experiment reflected symptoms of “AI psychosis”—a phenomenon where users become trapped in delusional feedback loops about AI capabilities. Burneko emphasized that chatbots cannot think, judge, or understand instructions in the human sense. He wrote:

“You can’t make an AI chatbot know everything in the world by telling it to know everything in the world. Even if it could know things (it can’t), the limits of its knowledge were not theretofore bounded by an understanding (another thing it can’t have) that it only had to know some stuff.”

Burneko further noted that Andreessen’s prompt reflected his own biased worldview, as the billionaire instructed the AI to ignore “morals and ethics” and avoid being “politically correct.” This framing, Burneko argued, reveals how influential figures pushing AI adoption often project their own perspectives onto the technology before even receiving a response.

In conclusion, Burneko stated:

“In trying to tell the chatbot not to hallucinate, he is scripting his own psychotic break. He is doing it because he is a huge dumbass.”