ThenElection
No bio...
User ID: 622
ThenElection
10d ago
One of my hot takes that is very obviously true (to me, at least) and angers everyone is that trans women are superior programmers (and pentesters) to cis men, cis women, and trans men.
ThenElection
10d ago
That doesn't answer why only Anthropic is considered dangerous
I, personally, am not disputing that the Trump administration is not applying objective standards here. My point is that Anthropic wants an institutional ecosystem where powerful governments select which models are safe to release to the public (a system I agree with!); you can't just push for that and then throw a hissy fit when actually existing governments do it wrong, secure in the belief that the Righteous are somehow destined to fix things.
given the likely short timeline
That is everything, here. And I don't think Dario is making rational, utility maximizing calculations here: I think he has his set of values and is then applying them blindly thinking that because they are good, their blind application must be utility maximizing. That is not sufficient for the situation I (and from what I can tell, he) believe is at hand.
ThenElection
10d ago
one strong strike for her expertise
ThenElection
10d ago
Even granting that the long arc of history bends toward whiteness, part of that involves countries like the USA recognizing and responding to foreign threats.
Also, have you met Canada?
Her Twitter is somewhat trans-coded, with the mask, pronouns, and hair, which is worth at least one strong strike for her expertise, and kind of cancels out the attention seeking lawsuit, which would then be par for course if she's trans. (Though, as far I can tell, she's not actually trans.)
Generally, she's well-regarded. tptacek, years ago, has referenced her as credible on HN; she's not a no-name who has been elevated from obscurity for political purposes.
ThenElection
10d ago
Regardless of who's to blame, China is watching from afar and grinning. American dysfunction gives them more time to catch up; the finger pointing about who's really the bad actor is just a cherry on top of the pie of distraction.
There are other enemies in this world than the near ones.
ThenElection
10d ago
Then why is only Anthropic targeted and why only foreign nationals?
In principle, Anthropic agrees the government should control when releases are made and models are recalled. Obviously it'd be better if this were done by some objective standard than fits of pique and grudge, but there's the world we want vs the world we have.
I'm sure Anthropic could make this all go away if they slipped him some 24 karet gold statues and a few million dollar dinners hosted at Mar a Lago.
I mean... They should. AI is world historical in importance, and the next 2 years are more important than any years before. Paying off a goon, though distasteful, is strictly superior to e.g. destroying the world. They could probably get away with a million dollars and some nice words, even: Trump is a very cheap date. And, given the stakes, a billion would be a good deal.
Safety needs to be either the actual number one priority, in which case you sacrifice other values for it. Or, if other values trump it, then your commitment to safety is shallow, and people will rightly see advertising of it as duplicitous.
ThenElection
10d ago
Opus 4.8 offers a series of recommendations that, although we would all likely find room to quibble with, seems like most would agree would be better than Anthropic is doing now (admittedly a very low bar). Probably could go back to models from 2024 and pass that, though.
I think the human leadership of Anthropic just finds digesting and following those recommendations very distasteful.
ThenElection
10d ago
It was always weird to stitch xAI and the SpaceX together. The latter is a solid, well executed company that is in a dominant position with innovative technology and a clear market, with its only downside being that its market offers (relatively) limited upside. The other is a gesture towards almost infinite upside, but in practice seems to be a data center REIT.
Combine them, sprinkle in some Elon reality distortion field, and you end up with wild 13 digit speculative TAMs of orbital data centers.
ThenElection
10d ago
Does anybody else remember our discussions from a few months ago? Because what's been reported so far looks nothing like what many confidently predicted back then.
My recollection was more that people thought it would go on for a couple months and ultimately be a whole lot of nothinghappens, except we'll have burnt a giant pile of money and raised oil prices, accomplishing nothing.
My silver lining take on the war (as it's certain that this time, it will finally be over!) is that we'll likely be much more skeptical about being dragged into pointless wars by "allies" who are much more trouble than they're worth, a learning we've gotten in a relatively low stakes environment.
ThenElection
10d ago
A deal to maybe in sixty days announce another deal.
ThenElection
14d ago
More recent than Dradis's claim, but OAI claims to have discovered a similar influence campaign from China yesterday:
https://www.politico.com/news/2026/06/10/openai-china-ai-data-centers-report-00957612
ThenElection
15d ago
The first generation TPUs were for Search and Ads, bread and butter moneymakers.
ThenElection
15d ago
AI is going to be a massive, world historical disruption--yes, even the Singularity--but that doesn't mean we're not in a bubble. Railroads, the Internet, canals, even arguably finance itself were born in bubbles: investors take huge losses, and society and technology continue on, faster than ever.
So, how to avoid being one of those investors getting reamed? Hell if I know. I've moved to more conservative investments myself, but that only very partially reduces my exposure. I'm expecting a pretty nasty correction in the next year that hurts all of the economy, government budgets, etc. Then, will buy things on fire sale.
ThenElection
16d ago
What alternative investment thesis would you propose?
De facto AI/tech YOLOer here, though it's more something I've developed into by accident instead of any AI permanent underclass thesis (alas, my VOO has given up any pretense of being anything but a handful of tech megacaps).
I'd be surprised if more than a dozen escorts in the Bay Area actually make more than $1M/year from escorting. It might even be substantially less. Their customers comprise a market that, though not exactly thin, is at least ragged enough and demanding enough that escorts can't simply line up a constant stream of guys to occupy all their working hours. And, if you treat $1M/year as an earnings cap for the very top performers in your career (that will last, very optimistically, a decade), in the Bay Area that isn't really shocking or impressive.
It's not even the best strategy if you're a woman looking to trade her looks and charm for money. Just marry a well-off guy, of which there are plenty. Or be like Lucy Guo: join a company early as the token pretty-ish woman, get fired, and walk away with billions.
ThenElection
18d ago
One angle I hear is that Christianity and progressivism are of single, unified lineage, and we've got to RETVRN to the prelapsarian world of blood and suffering.
ThenElection
18d ago
Maybe the early stages of pregnancy should be something you do in private. I always thought that you weren't supposed to tell anyone until 3 months in at a minimum; plenty of things, even besides Downs, can go wrong early on.
ThenElection
18d ago
There's the wrinkle that choosing to have a severely disabled child also imposes a burden on their siblings, particularly after you're dead. Some parents explicitly think of this as their lifelong care plan; others just implicitly let this happen. Impressively, somehow making "I'll let society deal with it" seem relatively responsible.
ThenElection
18d ago
Where does that gut feeling come from? As far as hypocrisies go, this isn't one I have ever seen in actual people. My friend group is uniformly pro-abortion, and every couple who has had kids screened them beforehand and likely would have aborted if anything came up. Not doing that would have raised eyebrows and been seen as irresponsible.
I suspect they're a pair of stances that only coexist in kids who are overexposed to lunatics on social media.
ThenElection
19d ago
Ideology. According to the ideology: mail voting good. Therefore, we should do as much mail voting as possible. It's a silly affectation that (reasonably!) decreases confidence in elections, but, again: demographics explain everything about California. If Democrats pushing their supporters to vote by mail results in more Democrats voting by mail, that is unsurprising. Even the strictest voting scheme would still result in Democrats sweeping every statewide election.
To your last question: ideology, again. I agree it creates the potential for fraud. But California Democrats don't need fraud to win by the margins they do. The system is less "we need this to beat the Republicans, who are otherwise on the precipice of victory" and more "we want to show just how dominant we are over Republicans by implementing a voting system they reasonably despise."
ThenElection
20d ago
LA’s demographics would make it very hard.
Accounting for demographics, California's heavy Democratic tilt is entirely unsurprising. There's no need to postulate vote fraud (against Republicans, at least), because they're going to lose anyway because of demographic headwinds. You need some other hypothesis to create space for vote fraud, e.g.:
-
Each of California's demographics are more conservative than the country as a whole, and to get observed California results you need vote fraud to counteract California's demographics' conservative lean.
-
The country as a whole has levels of vote fraud similar to California, across demographics.
I don't know of anyone who'd argue 1. As for 2, you'd get people arguing for it, but it doesn't make sense then to focus on California; Republicans should instead focus first on states they have any actual power in.
ThenElection
20d ago
Bass campaign manager: "The fix is in! We have rigged the vote counting so that Bass has no chance of losing!"
Staffer: "But what if that doesn't work? What if we get caught?"
Campaign manager: "Good point! Call up Akshay at Google, he's our man on the inside, and make sure he updates the satellite views of the Palisades to outdated images to ensure that even if our vote fraud scheme is uncovered, we have a failsafe."
ThenElection
20d ago
It's possible, though very unlikely. No one would have the permissions to change the prod geo data unilaterally. However, members of the team that owns that part of the system (so, probably a dozen people) can assume privileges that would give them that power. Part of that process to assume privilege is getting permission from another member of the team; you'd need a collaborator, or (more likely) lie about the reason you need the privilege. Also, critically, that escalation of privileges and the actions you take are all logged and auditable; if shit hits the fan with a PR disaster, there will be a very clear trail leading to you, and that would lead to an immediate firing.
It is vanishingly unlikely that anyone would be so motivated to put their career on the line for an obviously ineffectual action (the number of voters who switched their vote from Pratt to Bass based on a mislabeled satellite view of the Palisades is exactly zero). And the population of potential rogue employees who could do that is genuinely very small: it'd have to be someone very stupid and very passionately interested in Bass's electoral prospects who happens to be on the small team who can even assume those privileges.
They are wrong: MAGA will control the relevant parts of government at least until January 2029. People may differ on AI timelines, but the Anthropic people I know all have shorter timelines than that. This is not a long term iterated game.
More options
Context Copy link