hibikir

WeWork collapsed because of their S-1 was on fire. If they had gone public earlier, it would have still been on fire. The reason they lasted that long is because Masayoshi Son didn't do anything that resembled due diligence, but the market would have always done it.

For reference, in my Spanish home town, a large consulting firm is hiring devs right out of college for a little under 20K euros a year. The apartment I grew up in, in said home town, would sell for about 400K euros, and an old one bedroom apartment from the 70s might go for 200K.

So insufficient wages? Yep. It pays more than working at a bar, but not by that much. Anyone graduating CS that can at least go to the bigger cities, or better yet, has the language skills to move further away, just does.

The world has a lot of tech journalists, but I have yet to see the longform article, with actual interviews, trying to dig into what is happening in this kind of manic push for features that seem to have no traction with anyone

There are failures in the history of software that come down to believers not realizing that they could never deliver on the promise they were selling, but at least the promise was compelling. But nowadays we are seeing this large orgs try to ship things that could never work. The 1000s of copilots. The hallucinated cat story... The kinds of thing you'd never demo to a serious product-centric exec, because it'd be your last day at the company.

And then there's the current execs, but I guess you'll never find one that will be honest with a journalist here. Can they not see that their product orgs are bankrupt? Whatever Nadella wanted, it sure wasn't the current copilot situation. It's not one product going wrong, but large parts of organizations going in directions that don't pass the smell test. We are in one of the least stable moments in tech since Windows 95 changed winners and losers. The times where malinvestment ruins established companies. How are we seeing basically every large company flailing?

Many arbitrary features can be turned off: I haven't turned on Siri on a phone for many years, because the utility was minimal vs the annoyance of errors in voice activation. Taking very significant space in your drive is just as important. I can uninstall Pages and Garage Band. I should be able to uninstall apple intelligence

It's much better than 5, but I've had a couple of situations this week where it was too interested in being independent, making calls that went directly against my recommendations. It can also do fun things like convince auto-mode to go way past what I have autorized. For instance, specific permission to run process X in region abz-1 suddenly became running X in 5 other regions, with no warning, and doing modifications that it never mentioned in the summaries. And a few of the times it got the calls very wrong, by assuming it understood systems it didn't. It'd even argue with me when corrected, as it assumed similar names were referring to the same thing, when they weren't.

So asking it to do things on its own for a long time? Given last week, absolutely not.

It's not always what the nerds want to do though: I know of what the insides used to look like in one of those on the list, and things were very custom, because some early technical decisions were outright landmines that were too expensive to move away from. So having custom harnesses to actually survive that codebase I remember seems like a baseline to get reasonable turnaround of AI helpers. The company had done cowboy development for way too long, and solved too many problems by creating bigger, more custom problems.

So don't assume it's all nerds having fun. Now, is it what the future will look like for everyone? I don't think so, but that's because I expect AI is bringing us a more unified dev experience, with more developers than even Google has. This will make general programming harnesses very good, quite fast, solely because said harnesses can be differentiators in the AI race. So we are seeing some of the largest companies inthe world, with the largest research budgets, dedicating more money to the dev experience of their product than almost anyone else does.

I have seen it, over and over again, in relatively small orgs (under 500 devs) where the wrong kind of people are handed the job of platform lead. Project after project people don't want to use. The platform jobs in those places were handled by technical skill, but it was attached to disinterest of how others work. So even when they were sent in the general direction of a useful problem to solve, things rarely pan out, because they don't like to talk to their users. Therefore they write the tools they would want to use and are interesting to build instead.

Empathy for other developers, trying to have a mental model of what annoys them about work, and a general product centric mindset are the real requirements, but few people that decide how to staff the team k ow how to measure those skills. And besides, when people have them, they are often sent to talk to the actual customer, not internal customers.

> You cannot be responsible for what you can’t control either. That understanding is key to reasoning about system behavior and fixing it when the AI inevitably fails.

This is not a good premise. All over law, you will find people made responsible for what they don't control and they kind of own. Unleash a dog that harms a child, or just have it in an environment where it can escape, and see what happens.

There is such things as unpredictable situations where one might not be held responsible, as a problem might occur well past reasonable guidelines.

So of course you can be held accountable for what an AI that uou supposedly cannot quite control does, or for the AI-written code you deliver. Treat it like the releasing a wolf pack, or selling an unsafe toy that can maim children. There's precedent everywhere.

The law works because most people won't break it, most disputes don't even end up in anyone getting sued, and most lawsuits end up in settlements before too much court time was spent. Break any of those things, and systems crack. Break multiple, and it all collapses.

If the cost of what a lawyer does today drops to pennies, the system probably needs robo-judges, and/or massive penalties for taking something in front of a human judge when the facts sufficiently completely one sided, as making them pay the winning party's legal fees will become insufficient deterrence.

For me, every day, although it fails far less often at this than Opus 5 did, which might as well have been a gremlin. It was a really poor technical writer too, leaving slop reminders to itself in all prose, including comments. So even if 5.5 fails less right this second, I have trouble believing that it all will not similarly bite me next week.

Summaries that don't tell me when it's changed direction in a timely fashion, but I am only told way later, when I have to undo. Really bad judgement calls regarding where to fix bugs, changes in implementation decisions, taking action when I am asking a question directly, not passive aggressively asking for action... when 5, 3 days ago, was proven to be untrustworthy, switching to very little supervision sounds like a strange thing for a customer to do.

D+ exists outside of the US only: Everywhere else al things owned by Disney that are streaming are in D+, like ABC and FX series. Also, nobody streams more than children that have Moana in the background, and thus watch it 4 times a day. It makes it far stickier.

Hulu is not long for this world anyway, as Disney has been trying to merge them for years, if you believe investor calls. They'll probably be just one streaming service within the year.

Stripe has already existed for a while, and has gone through ups and downs. There was a period where customer experience was first, and the systems inside were built fast and in rather cavalier ways (See Greg Brockman famously telling everyone in a MongoDB convention about using Mongo's oplog as a queue). There were times where internal tools teams could spend all the time they needed to get about as much shine internally as the external facing websites were. Huge investments in everything. But there's just so much churn, as some staff burns out, and others realize that they have so much vested stock that continuing in such a high intensity environment is pointless, that whatever Stripe might have been during your tenure might be very different from other people's, and you can both be right.

Other countries have double, even triple the number of physicians per capita. They also pay them a lot less. Are they all less competent? Are US filtering requirements actually all that predictive of future performance? Can we, say, see that MCAT scores really line up well with clinical performance? Because it's easy to be selective, but not so easy for the selectivity to be predictive of actual quality, especially when you are being selective before training starts.

If you look across countries, you realize that a lot of the sacrifices are part of the pipeline, but not necessarily key to good outcomes. Just like the grueling hours at certain residencies.

It's like claiming that one needs to cram hours upon hours of leetcode practice to be a quality software developer. It might be necessary for some companies to hire you, but that's to meet a filter that is less and less predictive over time, not an actual performance requirement.

Most countries get good results with shorter training and much lower salaries. And, just like most other US university degrees, the majority of the filtering is done on entrance, not on training. So it's not that we are evaluating the best doctor, but the best diligent people that get great scores in the MCAT. Whether that actually lines up well with being a good doctor is not all that clear.

We were already there at places with a lot of contractors, capped at 1 or 2 year stays. You saw teams of 10, 20, with 2 people with serious institutional knowledge. Now the contractors are just Claude

In most of these cases, you will see a place like D+ carry the same announcers as the TV station. Here Inbet it would be ESPN crews. So having a different, special crew just for ad breaks would not be exactly great. And given the density of ads in, say, the Super Bowl, you would be hard pressed making this a good experience.

Will has trouble with this in a relatively nimble org, where he has good control of what is adopted. Imagine how much fun this is in more ossified enterprise environments, where getting work done efficiently means bypassing process with dev lead and line management blessings. The differences in performance among individuals has never been wider, and not just gains in productivity, but performance losses as some people who have iffy judgement cause trouble a lot faster.

Maintaining quality when an organization just lacks the muscle to make changes and nobody has the mandate to try to make sure uptime is in good shape is a challenge. A lot of things break, precisely because there's just as much change as Will sees, but it's very unevenly distriuted.

You didn't get that complaint in Clojure, just like you wouldn't in Scala or Haskell, is that once you have any expectation that your users know a little bit of category theory, and possibly also thinking in types, it's all quite easy. Even fold is kind of easy, with the more complex signature. But passing [A][A,A =>A] kind of sucks for those that don't think of functional programming. and [A][A,B => A] is even worse. It's often bad enough to get people to build a comparator.

Every industry language keeps gaining more and more functional features: Many a new Java version is adding a bunch of scala features with worse syntax. But we don't train people on functional programming at all, so by the time they've built their instincts, passing functions makes no sense to them, immutability is alien, and the idea of a pure function seems irrelevant to them. Thus, they don't get exposed to the building blocks that make reduce seem simple. We always teach them recursion, but the rest? Too little, too late.

I could tell you of a bunch of ways to simplify the signature by, say, mandating that one passes a monoid or something like that, but while the signature would be easier, the very same people that are only used to imperative OO will not have an easier time, because they might have studied 2 years of calculus, but they've never even smelled abstract algebra. You can walk out of not just a programming bootcamp, but many a computer science degree without learning a word of this. Therefore, it all remains complicated.

There would be value in prep if one had an engine that automatically understood what positions are challenging for humans, as right now a big part of high level prep is finding opening lines that look like they lose by a bit in stockfish vs stockfish, but are full of landmines for a top GM to step on. That's where a lot of time goes nowadays, past the pure memorization. The issue is that value drops a lot when other players have access to the same tools, because then they will also pay attention to those lines. The value is in the information imbalance, but everyone has the same tools.

It might help young players learning without a coach, but my understanding is that most are just looking at engine evaluation anyway, as tactical speed is a typical advantage of the youth.

It depends on the part of the system you are talking about. Under the covers, many a system will send reports higher up for 5 figure transfers, just for money laundering detection. And there will be all kinds of things blaring for anomaly detection. It's not the same if an account that normally moves large sums, and has been cleared and "explained lol moves a large amount than if it's one making nothing. See, for instance, how Notch's PayPal account was flagged when suddenly the game started to do great sales. So the special attention will vary, and there might be a human involved at some point even if they don't call you.

Yeah, doing compute-heavy work a couple of jobs ago, we tried small images with musl, and the default allocator was a catastrophe: 75%+ slowdowns for our real life tasks. Even with a better allocator, we were way better off with the larger image.

Doing just payments is a really low margin industry, and the larger the contract you are after, the lower the margins. Therefore, it makes a lot of sense to try to sign up startups, as their growth is your growth. And to win in the startup market, you don't win by lowest costs, but by how much generic work you can save them. Cut their headaches, and they'll be happy giving you a wider cut.

Thus, a million little acquisitions to make the possible Stripe bundle for small companies stronger, as they become the moat. The opposite of, say, the Adyen play, when you want to lower your own costs, and make money on tiny margins to do processing for really large companies.

The right way to read this isn't that all billionaires can face consquenecs, but that the line that makes sure you don't face consequences is quite a bit different than being a billionaire. You can still have a power center, and impunity, as long as you make sure your misbehavior doesn't embarrass the people with the actual power. Those are smart enough to not do the things they shouldn't in a way that can be made public, and therefore cause said embarrassment.

One can prefer a regime where it's very hard for people who abuse their power to avoid punishment if it's all public to one where the powerful are emboldened enough to be able to be shameless. Even worse when the shamelessness is seen as the purest display of power, and thus people compete in their public misbehavior. But still, understand that the real barrier to punishment in the first case is not to avoid doing bad things, but to make sure that the witnesses either don't exist, or are too scared to say anything.

A judicial system that actually has no sacred cows would be better in all cases, but the sacred cows everywhere do their best to make sure this doesn't happen.

Because, given you aren't filing for PERM these days for someone that you have not already hired, someone off the street is going to be straight out worse than your current employee.

PERM ade sense in a theoretical world without H1s or F1s, where you are filing PERM for someone off the street. But that almost never happens outside of cases where yes, you coud put the ad in the front page NYT and you'd still hire the candidate. And thus eitehr way you slice it, looking at the resumes of your ghost ad is a waste of time either way.

And if one really wanted to go back to that kind of green card process, it has to be very fast anyway, instead of being as slow as it is. Otherwise it becomes useless regardless.

Sid Meier himself warned about the difficulty of trying to design games tat are multiple games at once, as basically everything in the list other than Lemmings is). His whole "one good game is better than two great ones" argument, as it's difficult to get a good balance of fun for any given player. If someone likes, say, the strategy elements of North & South, but not the battles, or they feel like they want more of one game but they have to do the other, there's some tension there. So you have to make the games be just fun enough, and change just often enough. A really hard problem.

And this is why it's relatively unlikely to find worthy successors: Making the whole be as good as the sum of its parts is hard enough. Getting to be as good as Pirates, or the best Cinemaware games, where the whole was significantly better than all of the games in isolation, is aiming foolishly high. Which is why, say, the xcom remakes are such a wonder, because they manage to both change the original two minigames underneath, improving them substantially, all while still being balanced.

The anecdote about the garden of earthly delights is going to be relatable even for art students. I've met art teachers that ended up "trapped" in Madrid by the Prado Museum: A green, American future Ph.D student that was heading to Greece to specialize in old greek art. But instead of flying straight there, they stopped at Madrid to see the Prado, and were never able to leave. Changed their direction immediately, and ended up teaching in Madrid, and retiring there, because one of the countless masterpieces in there stole their life plans away.

There's more famous museums, and it doesn't have the most popular paintings for non specialist, but it's just haymaker after haymaker, in every single room. And that's leaving the modern art to the Reina Sofia, a short walk away. (Although I guess keeping late Goya there is kind of cheating, as I'd argue he becomes a pioneer of modern art after the war starts)

Books might require quite a bit of work before you can tell they are eternal masterpieces. Seeing the depths of some forms of music might involve not-so-impressive efforts earlier, just to get to the good parts. The very best paintings in the world, in person, can stop people cold. Large format efforts that demand attention even when you don't consider every single small decision.

The Garden of Earthly Delights used to be stuck in a corner, at the edge of one of the wings of the Prado. As if "oh well, I guess it fits here, near all the medieval art". It counted as an ambush.

We developed creams that stopped sunburn, but were still great at getting you cancer. A lot of the ads from 70s and 80s talked about how they stopped UVB but let the tanning UVA through! So yes, they didn't burn, but the long term cancer rates exploded, along with all the skin aging.

So it's not that they ignored the all knowledge. I blame most of it on the rather detrimental tech advancement. Like cigarettes with filters: now safer (except you can smoke more now, and the cancer risk is still a disaster)

On car insurance, you have a competitive market with actuaries trying to identify actual risks, and that, with their prices, modify bad behavior and purchases of cars that are expensive to repair. So all in all, you can end up with very narrow margins, and end up ahead only on the float between receiving premiums and paying out. Health insurance, especially how it works in the US, with no public system williyto cover all the expensive stuff for catastrophic situations and chronic conditions, would compete by trying to fail to insure risky people. But since most insurance is handled by employers, all the levers for that fail too, so it's an especially bad insurance system. Home insurance, car insurance, normal life insurance... all more competitive and more beneficial economically, as their efforts lower costs overall.

The flip side of Bezos' "your margin is my opportunity". But that's for competitive markets, not for heavily regulated multi-party setups with minimal competition. That's the real wonder of American healthcare, as it manages to neither be a good market, nor to have regulator mercilessly going after the largest profit centers. Near the worst of both worlds.