Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Schmidt says declining institutional trust means people trust friends, and AI can manufacture fake friends to befriend people and push false causes. He cites a friend who believed data centers kill cattle because social media promoted a farmer’s baseless claim; he found no police or veterinary reports.

Related Insights

The ability to label a deepfake as 'fake' doesn't solve the problem. The greater danger is 'frequency bias,' where repeated exposure to a false message forms a strong mental association, making the idea stick even when it's consciously rejected as untrue.

Large language models present all information, including falsehoods, with a consistently confident and articulate style. This fluency builds a quiet trust in the user, making them more susceptible to believing dangerously wrong information, particularly on high-stakes topics like health and politics.

Beyond generating fake content, AI exacerbates public skepticism towards all information, even from established sources. This erodes the common factual basis on which society operates, making it harder for democracies to function as people can't even agree on the basic building blocks of information.

Longitudinal metaphor research shows users have shifted from conceiving AI as neutral tools toward anthropomorphic entities like assistants and teachers. In psychological evaluations, this perceived humanization elevated perceived warmth alongside competence. This perceived warmth can induce misplaced trust, leading even seasoned deception and deepfake researchers to overlook hallucinated citations in professional work products.

The problem with AI has evolved beyond 'garbage in, garbage out.' Today's systems can rapidly ingest misinformation from public sources and present it as fact, creating a feedback loop. This means bad information is not only used for poor decisions but is actively amplified and distributed faster than ever before.

A man faking an NFL career allegedly used Google's AI-generated summaries as "proof" of his claims. This reveals a new vulnerability where bad actors exploit AI's tendency to confidently present unverified information, effectively laundering falsehoods into what appears to be authoritative, third-party validation.

Social media thrives on the psychological reward of posting for human validation. As AI bots become indistinguishable from real users, this feedback loop breaks, undermining the fundamental incentive to post and threatening the entire social media model which is predicated on authentic human receipt.

The proliferation of AI agents will erode trust in mainstream social media, rendering it 'dead' for authentic connection. This will drive users toward smaller, intimate spaces where humanity is verifiable. A 'gradient of trust' may emerge, where social graphs are weighted by provable, real-world geofenced interactions, creating a new standard for online identity.

The system of cheap labor, AI drafting, and fake accounts is topic-agnostic. It was built for commercial purposes but can be easily repurposed for malicious intent. The machine doesn't care if it's amplifying a product launch or state-sponsored disinformation; it just works.

The online frenzy over Netanyahu's supposed death was fueled by trivial details like a crease in his palm looking like a sixth finger. In an age where AI makes the public doubt reality, even easily debunked visual artifacts can spiral into massive conspiracy theories.