Your example of failed US nation-building is postwar Germany??

Kalashnikov worked for decades to build better rifles. Regret is an old man's game.

Even if these AI engineers do regret what they've built, I fear it's too late. A couple years ago, when news broke about Project Lavender, the Google engineers who worked on it protested and got fired. Google's still building weaponized AI systems.

> And one thing that really stands out is that there are really not that many shortcuts.

Completely agree, but the one exception that really stands out to me is canning. Napoleon offered a large cash prize for a cheap and effective method of preserving food for use as army rations. The method used (canning) doesn't require any special equipment (glass jars or tin cans are nice, but not necessary). It would have theoretically been possible to discover this thousands of years earlier. What would have happened if Hannibal or Pyrrhus or Cyrus offered the reward instead?

One problem with canning is actually the material for cans. Without glass, you'll have to use ceramic jugs. They will be more difficult to sterilize and seal properly.

Definitely not impossible, but probably not discoverable without the knowledge of germ theory.

My guy, there were double-digit number of people who had his level of access to both foreign surveillance and domestic surveillance. They were going to find him out, so he did away with pretense and put his name on it.

Why do you think that an LLM wouldn't have biases?

You're both partially right, and that highlights the difference between nonviolent and violent resistance. You are incorrect in saying that a resistance is always trying to disincentivize the counterparty. Even in your example, the NVA didn't overrun their counterparty (the US military); they convinced enough of the US voting public (which is very much a separate entity from the US military) that "Peace with honor" was a viable, preferable option.

The original judgement held that the driver was 2/3 responsible, Tesla 1/3 responsible, which seems reasonable. The $243 million wasn't for causing the accident, but was a punitive amount for doing things that looked an awful lot like lying to the court and withholding evidence.

This makes a lot of sense and makes the verdict seem reasonable, thanks for providing the context.

The precise wording is

"investigate, block during the pendency of an investigation, regulate, direct and compel, nullify, void, prevent or prohibit..."

>Isn’t it up to the parents to limit social media use?

That was the argument tobacco companies used, and the courts ruled against them, which is why it matters quite a bit what Mosseri thinks.

I agree with you about student loan forgiveness, but disagree with your assessment of universities as a system. The humanities (which are highly profitable for the university) are suffering an existential crisis because their funding keeps getting cut, while STEM programs (which have low or even negative ROI after lab/equipment/resource costs) keep getting expanded. There's obviously more nuance to that, but at a high level, universities are prioritizing majors which have higher job placement at the expense of their own bottom line.

Is this actually true?

I think about my lab classes and typically these were separate credit hours. More money to the university. Outside chemistry, we were using outdated equipment that wasn't getting refreshed/bought new.

If what you are saying is true, Humanities needs to cut costs of their credit hours or double down. For entertainment degrees (art, music), its prohibitively expensive to casually obtain the education. For business/politics/psychology, there is at least some sort of return on investment to be expected.

"Labs" as in the physical facilities, not the class credits. Computer labs are one example - my school had a Linux lab that, uh, wasn't much used outside of CS/IT/maybe EE. There's also lab equipment for research professors - a nice oscilloscope can cost more than a new car, and students won't ever actually get to use that. Not saying it's wrong - I really appreciated that Linux lab - but it's definitely not equality between departments.

For your second point, I've never heard of a university that charges different credit rates for different majors (outside of special stuff like paying for certain PE classes), but that seems like it would be a university policy, not a department policy.

How about clawbacks from the universities.

For all loans? What if a person takes a loan to get into a medical program but can't find a job after and defaults the loan.

Is the university responsible? The arguments seem to be based on majors that have a poor job market.

Seems massively unconstitutional to retroactively impose penalties, especially if the previous behavior didn't violate any laws. Not to mention the reputational risk on the trustworthiness of US as a place to do business and the risk of abuse from future administrations. You might cheer that universities are getting their just desserts for scamming students, but your political adversaries might use it to claw back funding from universities for being too "woke" or whatever.

> Seems massively unconstitutional to retroactively impose penalties

For existing agreements, of course.

Going forward, for those who are the real beneficiaries of the loans, they should have a skin in the game.

Why aren't universities standing behind their product and offering financing without the unusual non-dischargeable nature of the loans? Do they not stand behind their products?