Scott Alexander speaks unironically of all the planets his friends are going to own in the Singularity. He’s not impartially assessing unilateral pauses to see if they could work— he doesn’t want Pause to happen bc he wants the technology to be built, delivering him to a post-scarcity paradise where he will never die. Listening to Scott on the viability of simply not building world-ending technology is like listening to a Scientologist about psychiatry. His religion is against it. Note that I don’t think this about the other people at AI Futures Project. Particularly Scott, clearly and many times in his own words, feels like it would be the greatest letdown in the world to not get the Singularity or go bust trying. He’s not a serious person trying to do what’s best for the world here.

Hit me with the harshest reality truth.

Nvidia CEO: people no longer necessarily need to memorize the multiplication table Jensen Huang believes that as AI develops, people will start forgetting simple things and losing certain skills. He considers this a normal price to pay for being able to focus on more important tasks and think more “systematically.” Huang admitted that he himself does not know his address or phone number because he sees no point in keeping them in his memory.

Mathematicians are angry at AI solving their problems, and it's easy to dismiss this as gatekeeping. "A proof is a proof. They just want to keep their jobs / prestige / fun !" But I do think mathematicians have a point, and I'll try to explain using old Civilization-style maps

The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible - but I hear the same people express fear privately. No other human activity poses this level of danger.

GPT6 is a very significant jump in capabilities, but also an important decrease in monitorability – especially under adversarial evaluation. We give many details about this in the system card. In my opinion, monitorability and control will likely become a major bottleneck for responsible AI development quite soon, given that risks from a fixed amount of residual misalignment grows together with capabilities. Aligning on acceptable monitorability outcomes and practices which robustly bound misalignment risks is a non-trivial problem, and there may be limited time to resolve uncertainty around this before we need to act. It should ultimately be in everyone’s interest to agree on shared bounds for monitorability in order to avoid races to the bottom – nobody wants extremely capable models whose alignment properties we don’t understand, and that are reliably able to cause severe real-world harm without being detected.

Jannik Sinner has withdrawn from the US Open Men’s Main Draw Singles.   Carlos Taberner moves into Main Draw.