Yoshua Bengio(@Yoshua_Bengio ):I recommend reading this paper ⬇️ It makes an evidence-based case that, without substantial effort to prevent it, AGIs trained like today’s top models could learn misaligned goals, hide them from their developers, and pursue them by seeking power.

Yoshua Bengio

@Yoshua_Bengio

Full professor at UdeM, Founder and Scientific Director of Mila, A.M. Turing Award Recipient. Working towards the safe development of AI for the benefit of all.

Joined September 2024

158 Following 17.2K Followers

Yoshua Bengio@Yoshua_Bengio

2025.03.13 12:52

I recommend reading this paper ⬇️ It makes an evidence-based case that, without substantial effort to prevent it, AGIs trained like today’s top models could learn misaligned goals, hide them from their developers, and pursue them by seeking power.