This is an interesting perspective, but I think the fallout from sticking to his guns here is probably greater than the public blowback he would receive from serving the DoD. Without this specific sticking point, the public would know that Anthropic was serving the DoD, but not what specifically the model was being used for, and it would be difficult to prove it wasn't something relatively innocuous.
How do you figure? As of this moment, it doesn't seem any clearer to me that Anthropic is going to win than it was a week ago. Prediction markets show DoD backing down as the most likely outcome, but I don't think they're out of the woods by any means.
Anthropic has been pushing for commonsense AI regulation. Our current administration has refused to regulate AI and attempted to prevent state regulation.
"The government doesn't have control of this technology" is an odd way to think about "the government can't force a company to apply this technology dangerously."
Some companies have good intent. Public benefit corporations are a thing. They aren't really relevant, because unscrupulous companies outcompete them.
Your assertion that pro-consumer companies would outcompete unscrupulous ones depends on consumers and regulators holding them accountable. So why are you arguing against being suspicious of companies?
Obviously the best strategy for companies is to appear to be pro-consumer, but "cheat" (meaning price fixing but also things like advertising and buying up competitors) as much as possible. In that context, "all companies are anti-consumer" is a decent shorthand for "you should assume every company is anti-consumer because the regulatory environment favors it, even if there are exceptions."
I think you're strawmanning a bit. They're not saying poor people are tax cheats, just that tax cheats tend to be poor. This makes sense for the same reasons other types of crime are also associated with poverty. This is not to say that wealthy people do not also evade taxes, but they do so in ways that are harder to catch and prosecute. You're implying that going after poor people is some sort of classist discrimination but I think it's far more likely that there are good reasons for it.
Or just that there are more poor people. Say 10% of all people are tax cheats, evenly across income. The top 1% who are the rich is much smaller than the bottom 50% who are the poor. So in absolute numbers there will be far more poor tax cheats than wealthy. Even if 100% of the wealthy are tax cheats, that still ends up being fewer wealthy tax cheats than poor tax cheats. Anything involving absolute numbers of audits is going to be skewed to show more happening to the poor, because there are so many more poor people than rich people.
Last I checked it's way closer to 28% than 48% of people that have earned income of at least $1 (thus EITC) and total income less than $25k -- which fall under the bucket of 48% of audits were for those with EITC and income under 25k. They are definitely disproportionately going after the poorest workers.
am i missing something, or is the statistic that is used to pinpoint someone as poor, is the same statistic that is gamed here? Namely the amount of income that a person declares to the IRS?
It may push a minority of users who really care about open source to Linux phones. I expect the majority of users will grumble but cave and re-adopt mainstream Android or Apple.
I assumed the "thinking" output from Gemini was the result of a smaller model summarizing because it contains no actual reasoning. Perhaps they did this to prevent competitors training off it?
It increases. In Germany an ministry wanted to hide an document about glyphosat, this caused many people to request access to it that an judge ordered to make it public.
I would guess It depends on how thoroughly the organization doing the censorship can exert control over information.
For example, I’ve been pretty impressed by the extent to which the Chinese government is able to influence public discourse within its borders when they want to. But China is also home to over 90% of the world’s Chinese speakers, and it has its own domestic social media industry with very few users from outside the country, the Great Firewall, less of a culture of anti-authoritarianism, etc.
I’m not sure how feasible it would be for the US to get to a comparable position. The US is nowhere close to being a supermajority of the world’s English speakers, and it might be hard for the government to impose an isolationist policy on the country’s tech industry without inciting a revolt by its tech oligarchs.
The GFW is real and many people don’t bother trying to jump it anymore. Mostly people in China don’t really care what the rest of the world is thinking, they got 1.4 billion in their own world, it’s big enough. The USA has only 342 million people in comparison, and we have a whole country to the north of us that barely talks with a different accent. Heck, half of our movie stars are Canadian or Australian (it feels like it anyways).
reply