Alarms about rogue agents ignore the real risk: agents that follow their creators’ instructions to the letter.
Throughout human history, evil people and tyrannical states have used the latest technology to cause massive harm. Here, slowing down AI may delay defenses against AI that the wrongdoers already possess or will acquire regardless of what we do.
This point is even clearer when it comes to geopolitical threats. The U.S. government could take advantage of models from frontier companies, perhaps including more advanced ones that aren’t yet publicly available. This access could give the U.S. a military advantage that may be decisive in this technological age. And because other nations know this, they may be deterred from attacking in the first place.
Frontier AI is itself a firewall. Slowing its development can postpone new defenses and leave existing threats in place. A slowdown can cost protection as well as progress.