This was weird, though now I get it. scala> 1 << 32 val res3: Int = 1 scala> (1 << 24) << 8 val res4: Int = 0
“That people move from representing the presidency to representing banks is so normal that we forget the costs: the private job done with the savvy to outfox one’s former public-sector colleagues, the public job done gently to keep open doors” @anandwrites.bsky.social www.nytimes.com/2025/11/23/o...
Opinion | How the Elite Behave When No One Is Watching: Inside the Epstein Emails
Link Preview: Opinion | How the Elite Behave When No One Is Watching: Inside the Epstein EmailsImagine - John Lennon & The Plastic Ono Band (w The Flux Fiddlers) (Ultimate Mix 2018) - 4K REMASTER
Link Preview: Imagine - John Lennon & The Plastic Ono Band (w The Flux Fiddlers) (Ultimate Mix 2018) - 4K REMASTER: YouTube video by johnlennonsuppose there were an elixir you could drink and become a giant, but if you did it would be inevitable you would, wittingly or not, step on some people like ants. would it be ethical to drink it?
“buildings are a bit like intellectual property, which also lasts longer than the economic horizon of the businesses that produce it. The economic argument for rent regulation is a bit like the argument for limiting patents and copyrights to a finite period.” // this is a very brilliant piece
Loading quoted Bluesky post...
an edit is always an addition, a delta in some form or a replacement. you want that, because history should be auditable to discourage malicious recontextualization of other people’s responses.
Out of 10,000, I suspect that you are right. But if the error rate is small, N will have to be pretty large to get that confidence interval. 1/
@mattbruenig.bsky.social can speak for himself (but i think he’s rarely on this site). but i suspect a heuristic may be something like comparing the LLM to a hypothetical research assistant. 2/
if, after extensive experience, it seems to perform as well as you’d expect, or significantly better than you’d expect, an RA whose work you’d use without full reverification, use the machine like an RA. 3/
there’s obviously a problem there, an RA, whatever their hit or miss rate, can be held accountable ex post, which has some bearing on the quality of errors. a human would know where they must be especially careful, beyond a baseline error rate. 4/
but a supervisor doesn’t typically publish an error rate and confidence interval for work not fully rechecked by research assistants. 5/
we have drafts now. which is huge, not just for drafts, but for composing on a phone with multiple pastes. now we just need edit. please.
He’s been using LLMs to summarize NLRB documents for years. He’s spent a lot of time manually reviewing the summaries. He’s about as expert a human as there is on the subject. It’s a mistake, I think, to imagine his apparent success with these tools are the mirages based on foolish naivety.
“LLMs are not just labor-replacing and productivity-expanding, but can, in some circumstances at least, enable the production of totally new things.” @mattbruenig.bsky.social mattbruenig.com/2026/02/10/t... ht @mitchsaid.bsky.social
how have we let this be our world? ht @mpaulmcnamara.bsky.social
Loading quoted Bluesky post...
“not funding community groups to implement strategies designed by national NGOs, but funding community groups to design strategy, period. That requires funders to give up control. Most aren’t willing to do that. Which reveals what this is really about.” @lennecefer.bsky.social
Loading quoted Bluesky post...
i did! sometimes i think economists’ definition of “real” is a deadpan gag, given how relentlessly it abstracts away from reality.
“Even a representative agent, forward-looking and fully aware of all the parameters surrounding them, can feel the vibecession.” @mtkonczal.bsky.social
Loading quoted Bluesky post...

