It's a European guy coping on HN.
That's OK.
They have no idea my sub-region of California produces the entire GDP of their country.
647 karma · joined November 30, 2020
That's OK.
They have no idea my sub-region of California produces the entire GDP of their country.
> You don’t even need a visa to leave
Either unable to critically read or extraordinarily pedantic. Classic HN.
There is simply no visa that allows skilled labor to come to the US, work a temporary job for a multinational that's paying them in their home country, and leave.
The closest thing apparently is the "B-1 in lieu of H-1B"[1] and guess what? Another commenter posted this FT article that accuses them of abusing this B-1 visa [2].
Traveling for work is a huge pain in the ass, doubly so for this sort of temporary work assignment, and triply so if it's to the USA.
I've always been told to use a tourist or family reunification visa. For example, if China cracked down on this, they could easily put 10,000+ Americans in jail for a similar "visa abuse". They obviously would not bite the hand of foreign direct investment like this though.
I think it's informative to interpret the law - especially in the age of Congressional gridlock - as 300 years of terrible legacy code papered over the Herculean efforts of ops teams (the government bureaucracy). When that ops team starts arbitrarily treating the oceans of gray zones to their whims, to reward friends and punish enemies, you start the long trek to serfdom...
[1] https://www.wsmimmigration.com/us-immigration/temporary-work...
[2] https://www.ft.com/content/c677b9aa-2e89-4feb-a56f-f3c8452b3...
Because he's Saquon Barkley. Other than your lead + maybe 1 or 2 others, everyone else is pretty interchangable. At least Saquon is interesting.
> the core insight: predict in representation space, not pixels
We've been doing this since 2014? Not only that, others have been doing it at a similar scale. e.g. Nvidia's world foundation models (although those are generative).
> zero-shot generalization (aka the money shot)
This is easily beaten by flow-matching imitation learning models like what Pi has.
> accidentally solved robotics
They're doing 65% success on very simple tasks.
The research is good. This article however misses a lot of other work in the literature. I would recommend you don't read it as an authoritative source.
Hmm
The prevailing wisdom is that American dominates tech simply because American companies are better. While that's true, I argue that a huge portion of this comes instead from the monopolistic winner-take-all dynamics of tech not found in other industries. Despite the American car manufacturers being significantly worse, they still exist. The European tech industry is significantly worse, and they're almost nowhere to be found in tech. The Chinese tech industry is worse in many ways as well, yet shielded by the Great Firewall, they've found themselves with cloud hyperscalers, frontier AI models, global consumer businesses, etc.
Once upon a time, this mattered less, but as American values diverge from the rest of the liberal democracies, the the effect American tech has on our political discourse and economies are becoming a matter of national security.
This is from the perspective of someone who works in the American tech industry but is not American, and has spent a fair amount of time in China.
As an opinion piece, I'm not sure my arguments are entirely correct, but I think they're worth thinking about.
Just that it would happen.
Every time something like this comes out, his supporters mass flag the article no matter how relevant it is to tech. Browsing /active is the only way to see some of these.
In their defense, if every atrocity and grift this administration committed were allowed free and fair voting on HN, it would completely clog the front page.
Daily reminder to use /active instead of /
Basically the solve rate was much lower without the use of a Bluetooth sensor, and they did a bunch of other things that made the result less impressive. Still a long way to go here.
Reproducing ML Robotics papers requires the exact robot/environment/objects/etc -> people fudge their numbers and have strawman implementation of benchmarks.
LLMs are so expensive to train + the datasets are non-public -> Meta trained on the test set for Llama4 (and we wouldn't have known if not for some forum leak).
In some way it's no different than startups or salesmen overpromising - it's just lying for personal gain. The truth usually wins in the end though.
Can you really claim that nationalism existed among the majority of people when those people were illiterate, had never traveled more than 10 miles outside their place of birth, and had never met someone who they could place outside their concept of their nation?
What does Waymo get out of this? It seems to me like if the value proposition in self driving is there (privacy, consistency, etc), people would be willing to download another app? Especially once word of mouth goes through.
Plus currently the Waymo app has neat features (when the car is on its way to you, it shows you stuff like when it's stopped at a red light or whatever) that Uber probably can't replicate due to API access issues.
I think the point is that it's not this simple: "Even more so when the current economic prosperity of the colonizing countries is still very much derived from their past."
Poor countries that were colonized are definitely held back by it. The colonial powers left extremely extractive institutions that focused on funneling power and profits into small groups of people. When they left, those institutions and that culture stayed.
But isn't it difficult to say that Spain or Portugal's "wealth" today derives from colonial holdings years and years back? What did they extract that lasts today? In those 400 years, did they not spend most of their extracted silver and gold on things that are trivialities today (porcelain, tea, etc)? Or waste it buying arms during civil wars and dictatorships? I mean there was mass starvation during Franco's regime.
It seems to me likelier that their wealth comes from the things they have in common with prosperous countries that never were colonial powers (Taiwan, South Korea, etc). Educated populace, functioning capital markets, property rights, rule of law, etc.
It's also hard to paint a gloomy picture since there are probably 10s of millions of Americans who would move to NYC in a heartbeat if an exodus of families really happened and brought with it a corresponding drop in rents.
West Virginia's problems are West Virginia's problems and have no effect on anyone living a perfectly wonderful life in California.
> When a founder goes to the bank to get a mortgage, it's difficult to explain to someone at, say, Bank of America that yes, their savings and income don't look very impressive, but they do have a lot of equity in a promising company – which, by the way, isn't yet profitable, but it will be! (Maybe.) A Big Banker doesn't look at that story and see a potential high earner. They just see someone who is incredibly cash poor and risky.
> "Boo hoo," you might think. "Pity the poor venture capitalist who can't afford to buy into their fund." But this person did not come from a wealthy or privileged background. They were relatively young. They didn't have family who could advance them the cash.
They offer loans and financial products to employees, founders, and VCs (new to me), that are cash poor but have equity in presumably valuable companies. A relationship that is supposedly hard to forge with other banks.
Their revenue is probably more on the order of 5-10b - the vast majority of payments volume is from large customers which negotiate much better rates that 2.9%.
I know they're just barely not profitable, so rev of ~8b, and total OpEx (salaries + AWS + cost of sales, etc) being approximately 8b sounds right to me.
At 8b profit, with normal tech multiples, they would be closer to a 200b company!