Or in an image. Don't wake up the leviathan if you can't go beyond navel-gazing.
To clarify, I don't mean sensible here as "good," but in an Hobbesian sense. Once you translate a domain into fear, you invite the sovereign to claim it. Once it does, it speaks its own language, decides in its own name, and ceases to be yours to manage.
The war machine bit reminds me of the predictions in the late 20s and early 30s by German fascist (non-Nazi) intellectuals, that Germany will fail in the event of war with the Soviet Union, because its reigning ideology was incapable of producing a formidable enough war machine.
In reality, they instituted labour control, switched the focus to mobilization (turning the state into a war machine) which "improved" employment numbers. Other than that they focused on advertising deficit tricks, visible projects like the autobahn, and statistical manipulation to show a miracle.
Reading a book and got reminded that it's funny and disturbing that one aspect of NSDAP propaganda has largely survived to modern times and even has wide acceptability: That Hitler came in during an economic crisis, turned things around, and bought about an economic miracle.
Erdős was ahead of his time. He was really focused on creating a dataset for building and testing new AI tools. He should be called the forgotten godfather of AI and put on a TIME cover alongside some other "architects of AI" who don't deserve to be there.
Unlike for Twitter, my habit of overusing bluesky, and using it more informally, means "there's always a tweet" sort of thing can actually be true.
I do appreciate the fact that the power hour is done by 730 AM local time. Rarely need to check my phone after 830 AM.
The above imagery is what people often invoke. But if they are so smart, surely they would know that none of what they talk of as "solutions" are viable (economically, for one; who will foot the bill for 100s of billions in investments if you are going to pause?). So what are they trying to achieve?
FYI: I'm a foreign national on a visa and wouldn't be allowed to use these models. But Anthropic's general messaging is too idiotic, not just for themselves, but very broadly (beyond AI). I refuse to grant they are these sincere and concerned autistic persons, who are expressing genuine concern.