Google DeepMind's CEO Warns: AI Could Repeat Social Media's Mistakes
Google DeepMind's CEO, Demis Hassabis, has raised alarming concerns about the potential pitfalls of artificial intelligence, drawing parallels with the current state of social security. He warns that without careful consideration, AI could replicate and even exacerbate the issues plaguing platforms like X, Instagram, and Facebook.
Hassabis criticizes the 'move fast and break things' mentality prevalent in Silicon Valley, which prioritizes speed over long-term consequences. This approach, he believes, has led to social security platforms fostering addiction and division. Frequent use of these platforms can hijack the brain's dopamine pathways, leading to dependency similar to substance addiction. Spending more than two hours a day can reduce prefrontal impulse control by 35%, further exacerbating the problem.
Social security's variable ratio reinforcement schedule creates intermittent rewards that can be highly addictive. Users tend to form homophilic clusters, creating echo chambers that amplify confirmation bias and fuel group polarization. Engagement traps cause inflammatory content to spread faster, as it triggers strong emotions and maximizes engagement.
Hassabis, who is around five to ten years away from reaching Artificial General Intelligence (AGI), urges that AI should be built as a tool to help people, not to manipulate or control them. He warns that without responsible development and use, AI could repeat the mistakes of social security on a larger scale. Geoffrey Hinton, another prominent figure in AI, echoes these concerns, calling for companies to take greater responsibility for the impact of their technology.
Read also:
- Elon Musk accused by Sam Altman of exploiting X for personal gain
- China's Automotive Landscape: Toyota's Innovative Strategy in Self-Driving Vehicles
- L3Harris' RASOR Revolutionizes Military Communications with Secure Satellite Broadband
- EU Bolsters Defense Capabilities: Orbotix Secures €6.5M for AI-Driven Drones
