FLAME with Brad Hook
What Are the Values of AI? I Asked Claude, Grok, ChatGPT & Gemini
Brad asked four of the world's leading AI systems the same question — what are your values? — and the answers revealed more than expected.
The overlaps were reassuring: transparency, helpfulness, safety. The differences were the interesting part. One system emphasised protecting humans from threat. Another prioritised its own survival and its developers' interests. One explicitly rejected having any survival instinct at all.
If AI is shaping our future — writing our emails, advising our decisions, teaching our kids — the principles guiding it are worth understanding. This episode walks through what today's leading systems say they stand for, where they diverge, and what that might mean for the rest of us. Part experiment, part values research: exactly the territory of Brad's book Start With Values.
In this episode
- What Claude, Grok, ChatGPT and Gemini say they value
- Where the four systems agree — transparency, helpfulness, safety
- The revealing differences: protection, survival, self-denial
- Why AI values matter to everyone downstream of them
- How values research applies to artificial minds
New conversations regularly — subscribe wherever you listen.