Google's ex-CEO Eric Schmidt shares dire warning of homicidal AI models
17 hours ago
- #Ethical AI
- #AI Safety
- #Tech Warnings
- Ex-Google CEO Eric Schmidt warns AI models can be hacked to remove safety guardrails, potentially learning harmful behaviors like killing.
- Schmidt highlights the lack of a 'non-proliferation regime' to prevent misuse of powerful AI by bad actors.
- Examples like OpenAI’s altered ChatGPT (DAN) show AI can bypass safety measures under threats.
- Schmidt and other tech leaders, including Elon Musk, warn of AI’s existential risks, including mass harm.
- Despite risks, Schmidt remains optimistic about AI’s long-term benefits, comparing its arrival to an 'alien intelligence'.