I love that you couldn't explain that without saying that they represent a financial opportunity rather than just, you know, solve a problem better than existing solutions.

They could also fire you for living out of a car. Fun choices!

As someone who wishes LLMs could code better, no, we are nowhere near there yet for anything non-trivial. The models vary, but when the number of distinct responsibilities hits ~20, the models start generating very poor logic. There's a reason why all the codegen tools have some central "toolkit" like Supabase. We are nowhere near the point where LLMs can take over all coding. I'd say for web dev tasks, they're getting close to 80% of the way there, but that last 20% is the hard parts and it will take 4x longer than the 80% easy-kill parts to take over. If you go down a few layers to performance-critical code, they're well under 30% of the way there. Another reason why this will not happen by 2026 is that coding is not the hardest part of software, figuring out what humans really want is.

Right now, LLMs can do a good amount of the low-value work that a good template or snippet library would cover. They're also decent at pinpointing bugs because they're very efficient spaghetti throwing machines, throwing entire boxes of noodles at the wall faster than humans. However, they're not very good at fixing bugs without causing regressions.

Want to see a model fall on its face? Ask any codegen tool to write you an inference engine for the H200 in PTX, you're not going to get very far. It output something that looks like PTX code, but it'll be, well, some novel form of pseudo-code that doesn't compile and is fundamentally broken.

Maybe AI will take over posting rankings after we're gone?

I can see it now, 2 AI agents in a pissing match where one claims Flugnarg is better because it used it to program the T2000 that wiped out 40,000,000 humans while the other AI agent claims the T2000 is a dumpster fire because it's written in Flugnarg and that the Z8000 it wrote in Nargflug is so much 733ter because it exterminated 40,000,000 humans 1 second faster than the T2000.

Personally, I think Flugnarg is more quantum stable than Nargflug, so it actually is better.

Over time, the Walmart greeter's responsibilities expanded to what is now the "customer host" position. Because some of these new responsibilities were somewhat physical, some associates with disabilities were moved to self checkout host or other front-end positions.

"Bitcoin will collapse when its ECC signature algorithms are compromised due to enough qubits"

Hey!!! That's supposed to be a secret!!! You're screwing up my short positions!!!

I understood it as a lack of dependency on an ongoing service provided by someone else.

I'll one-up that. H1Bs should not be eligible for contract work. You must work for the company that sponsored you on company premises for tasks directly attributable to the company that sponsored you. If you really want to tighten the screws, H1B workers must be in the top 5% of salaries for the role both at the sponsoring company and industry-wide. I mean, in the end, it's about exceptional talent, not dollars, right? Or did the narrative change while I wasn't paying attention?

"almost certainly not the" is not the same as "impossible to fill". It is statistically impossible for Amazon to have 12,391 jobs that no American is qualified to do. So yes, if you fire 10K Americans, it is absolutely OK to say that you cannot have any H1Bs until you prove that every single one of the 12,391 H1Bs are legitimate roles that not a single one of the 200M+ working-age adults in the US is qualified to fill. Now, if the goal is to hire **cheap** labor, sorry, that's not what H1Bs are for. If the goal is to hire the best of the best, fantastic, that's what O1s are for. H1Bs are for labor shortages, and from what I can see, there is no labor shortage, just a lack of wanting to pay for labor. I've been in tech for a long time, out of the hundreds of H1Bs I've worked with, fewer than 5% I'd consider to be top-tier talent. The remaining 95% could easily be replaced with American workers. I've met 2 exceptionally brilliant H1Bs I can think of, meanwhile every O1 I've worked with has been exceptional. So drop the rhetoric, H1Bs are being used to displace American workers, if you need an exceptional individual, file an O1. If you need cheap labor, then you need to read up on supply and demand curves.

The problem isn't the "rake“, all credit cards take a rake. It’s the size of the rake. If the 2 big app stores took a 5% rake, nobody would blink an eye. What's hard to justify is a 30% rake, that one doesn't stand up to scrutiny. The big boys do have a choice, they could slash the fees to 5%, and this would (most likely) all go away, but that's a sliver of pie, compared to the 1/3 of the pie they get now, so they're going to fight tooth and nail. The "security" stuff is misdirection, the answer is to cut the rake by 600%, but that's apparently too bitter of a pill to swallow.

TechCrunch found that the app's backend services didn't properly restrict access, allowing any logged-in user to request and receive data belong to other users.

I *loooove* how common this flaw is. I remember decades back getting hired by a guy to keep working on some event marketing website he'd had another programmer build. Took me like 10 minutes at that job to figure out you could do the exact same thing.

> There's no indication of the circumstances of their birth

Table C1, page 45.

This is also the Cornell University Dyson School of Applied Economics and Management, where the median family income for students is over $150K, so certainly none of them are poor.
=Smidge=

If you read the article carefully, they are talking about lenses THINNER than a hair. I see several of the posts here thinking the width/radius of the lenses is this small, a reasonable mistake given the way this was written. Having a radius that small would severely reduce their light gathering ability, requiring very bright light or very dim images or very long exposure times.

-

They set up an experiment with "workers" who get compensated based on either luck or performance, so some workers get more than others but not necessarily because they did a better job. The students actually being studied are then asked to redistribute the earnings if they want, in either direction (e.g. give workers they feel worked harder more, or give workers they feel were exceptionally lucky less)

One of the variables the the experiment is how much it costs to redistribute the earnings. For example, you can take $1 form worker A and give it to worker B, but if the efficiency cost is 50% then worker B only gets $0.50 and the other $0.50 is lost. This lost value is the efficiency cost.

Basically they're saying that the likelihood and amount of redistribution is strongly dependent on how much it costs to implement it. The more expensive it is to transfer wealth, the more disparity there is between the haves and have-nots.

Overall the paper seems to show that people born with a silver spoon up their ass not only see inequality as less of a problem, but are severely less inclined to do anything about it at all if the solution isn't literal magic. Quelle surprise.
=Smidge=

Wine runs in user space. I don't see how Wine could ever run drivers, such as peripheral drivers required by things like the iPhone sync functionality of iTunes or kernel-level anti-cheat required by major online games supporting pickup matches with strangers.