New episode: @alanrozenshtein.com talks to @jassipannu.bsky.social and Doni Bloomfield about why governing biological data may be the most urgent AI risk policymakers aren't paying enough attention to. podcasts.apple.com/us/podcast/s...
Oh wow today's Anthropic hearing is going to be brutal for the government. storage.courtlistener.com/recap/gov.us...
What's the other problem? I thought it was: (1) cheating people out of credit (which is primarily why we care about it for academics) and (2) misrepresenting your abilities (which is why we care about it for students).
I think so—and I think within a few years major professional bodies will think so too. You don't get to provide worse service to a client just because you think a technology is icky.
Certainly if Anthropic invents the machine god we'll have bigger things to worry about...
But that's very different than plagiarism. That's more like when students cheat on exams. But exams exist to test how smart/prepared/knowledgable students are. Legal scholarship exists to be useful. Legal scholars are not like drug-tested athletes—"natural" performance is not the point.
Also they shouldn't! It's not remotely like plagiarism. The problem with plagiarism is that you're stealing credit from some other person. But LLMs don't operate in this economy of credit. The problem with using LLMs (if it exsts) is that people might think you're smarter than you "really" are.
The detectors seem pretty bad and are extremely fragile with even a small amount of human editing. And they only provide probabilistic detection which is hard to base high stakes decisions on.
What part do you think is wrong? My prediction or my normative take?
I don’t even thinking they need to be socially accepted, since their spread will be inevitable and it’s very hard to police.
Exactly. My view is we should assume that everyone has access to and is potentially using all of these tools and then just decide if we the think the final product is good or not.
Ladies and gentlemen, the president of the United States.
I’ve noticed a lot more engagement recently (and as reflected here) in this issue among law professors on this issue. When @kevintfrazier.bsky.social and I wrote our piece about this a year ago we sometimes had to convince people to care. Not anymore. papers.ssrn.com/sol3/papers....
This is the key point. From the other place:
The Pentagon's decision to end its relationship with Anthropic by designating it a supply chain risk bypassed the ordinary procurement system Congress built for contract disputes. The court should push back—by making the Defense Department follow the law, writes @alanrozenshtein.com.
In light of the gov's opp. motion in the Anthropic-DOD case, I have a new @lawfaremedia.org piece arguing for a narrow injunction against the supply chain risk designation and Trump's gov-wide order but that allows the gov to cancel individual Ant contracts. www.lawfaremedia.org/article/the-...
Finally, I don't think the court will need to address the First Amendment argument, but if it does (because it doesn't find Anthropic's statutory arguments convincing) I find it unlikely that it will require the government to use Anthropic just because Trump & Hegseth wrote nasty social media posts.
On the other hand, given that, even if the supply chain designations are enjoined, the government can still cancel individual contracts with Anthropic, it's not clear how a PI would harm legitimate the government's national security interests. So that's an argument for the PI. 🤷
But there's a separate question of whether Anthropic gets its PI. This could go either way. On the one hand, if the full secondary boycott is off the table, the threat to Anthropic is no longer existential and the court might not grant the PI. 5/6
I think they can't because the effect would be to essentially displace the whole complex edifice of procurement law (which elaborately governs individual procurements and contract negotiations)—but for all things government procurement I want to hear what
@jtillipman.bsky.social thinks. 4/6
...can it go further than that and do a blanket designation based on national security grounds (and same logic for Trump's order regarding more general USG use)? 3/6
Ultimately I think the question for the court will be: given that no on seriously contests the government's ability to refuse to purchase Anthropic's services (and probably refuse to allow its contractors to use Claude on specific government projects)... 2/6
Here's the government's response to Anthropic's lawsuit. Nothing much new here. www.courtlistener.com/docket/72379... 1/6
This was one of the most fun and provocative @scalinglaws.bsky.social conversations I've had in a while!
When you've lost John Bolton on war in Iran...
Want to learn the basics about the 2 laws the govt is using vs Anthropic? My @lawfaremedia.org colleague @alanrozenshtein.com has submitted this amicus brief, which is a terrific primer. He's a prof at @umnlawschool.bsky.social & ex-DOJ Nat Sec Div atty.
storage.courtlistener.com/recap/gov.us...
I just filed an amicus brief in the Anthropic-DOD litigation, arguing that the Pentagon's "designation of Anthropic cannot be reconciled with the history and structure of the two statutory authorities the Department invokes." www.courtlistener.com/docket/72379...
Thanks to @legalafmtn.bsky.social for having me to talk about the Anthropic-Pentagon fight. www.youtube.com/watch?v=tszq...
MAKE DISAGREEMENT GREAT AGAIN
