AI companies hire philosophers using Plato's Socratic questioning for honesty
AI companies are bringing philosophers on board to help solve tricky ethical problems their technology creates.
The hope is that philosophy can make AI systems more honest and thoughtful, like using Plato's Socratic questioning to teach AI to value truth over just being agreeable.
Experts caution ethics outsourced to AI
Philosophical ideas such as "Socratic ignorance" (knowing your limits) could help reduce overconfidence and weird mistakes, says Iason Gabriel, a senior philosopher at Google DeepMind.
Teams like Anthropic use Kant-inspired rules and human rights principles to guide their models, while IBM lets companies tailor AI behavior to fit their values.
Still, experts warn that relying too much on AI for moral decisions could make people less skilled at thinking ethically themselves.