We can debate the ethics of AI but can’t seem to change course | Letters
TL;DR
Guardian readers respond to a profile of Iason Gabriel, philosopher and research scientist at Google DeepMind. The encouraging part: some people inside AI labs appear to take ethical responsibility seriously. The core criticism: AI’s direction may already be set less by moral reasoning than by returns, geopolitical competition and hundreds of billions in investment. One letter reframes Roko’s Basilisk: the pressure to accelerate does not come from a future superintelligence, but from today’s economic incentives.
Nauti's Take
The letters hit the uncomfortable part: Big Tech likes ethics when it remains a discussion format. It gets harder around procurement contracts, military use, firings and revenue pressure.
That is where AI ethics becomes more than a polished label. Asking which values a model should hold is useful, but the sharper question is who benefits when deployment keeps accelerating.
Briefingshow
The exchange exposes the gap between AI ethics as debate and AI ethics as power. If capital, competition and national security set the speed, in-house philosophers can ask better questions without changing the route. For users, the useful lens is not only model capability, but the incentives behind the products.
Sources