It's very clear that Anthropic doesn't really want to expose the secret sauce to end users. I have to patch Claude every release to bring this functionality back.
Remember there are no moats in this industry - if anything one company might have a 2 month lead, sometimes. We've also noticed that companies paying OpenAI may swiftly shift to paying Google or Anthropic in a heartbeat.
That means the pricing is going to be competitive. You may still get your wish though, but instead of the price of an engineer remaining the same, it will cut itself down by 95%.
Yeah. If you ignore the negligible fact that some investor may want a return on all that money that is going into capex I am pretty sure you can, Enron style, get to the conclusion that any of those companies have “healthy” margins.
Amazon was founded in 1994, went public in 1997 and became profitable in 2001. So Anthropic is two years behind with the IPO but who knows, maybe they'll be profitable by 2028? OpenAI is even more behind schedule.
How much loss did they accumulate until 2001? Pretty sure it wasn't the 44 billion OpenAI has. And Amazon didn't have many direct competitors offering the same services.
Did Amazon really not turn a profit, or apply a bunch of tricks to make it appear like they didn't in order to avoid taxes? Given their history, I'd assume the later: https://en.wikipedia.org/wiki/Amazon_tax_avoidance
Anyway, this has nothing to do with whether inference is profitable.
Their price is not a signal of their costs, it is the result of competitive pressure. This shouldn't be so hard to understand. Companies have burned investor money for market share for quite some time in our world.
This is the expected, the normal, why are you so defensive?
That's not how valuations work. A company's valuation is typically based on an NPV (net present value) calculation, which is a power series of its time-discounted future cash flows. Depending on the company's strategy, it's often rational for it to not be profitable for quite a long while, as long as it can give investors the expectation of significant profitability down the line.
Having said that, I do think that there is an investment bubble in AI, but am just arguing that you're not looking at the right signal.
Why would you gladly pay more than what it's worth? It's not an engineer you are hiring, it's AI. The whole point of it was to make intelligent workflows cheaper. If it's going to cost as much as an engineer, hire the engineer, at least you'd have an escape goat when things invariably go wrong.
Scapegoat, got it. Can't blame the autocorrect though... I honestly thought it was spelled like that, which is a shame since I've been studying English my entire life as a second language.
> one of two goats that was chosen by lot to be sent alive into the wilderness, the sins of the people having been symbolically laid upon it, while the other was appointed to be sacrificed
Let me get this straight: in the Bible, the scapegoat does survive, while the "pure" goat that did nothing wrong gets killed? That's... messed up, even for a tribal rite.
I'd pay up to $1000 pretty easily just based off the time it saves me personally from a lot of grindy type work which frees me up for more high value stuff.
It's not 10x by any means but it doesn't need to be at most dev salaries to pay for itself. 1.5x alone is probably enough of an improvement for most >jr developers for a company to justify $1000/month.
I suppose if your area of responsibility wasn't very broad the value would decrease pretty quickly so maybe less value for people at very large companies?
Yes, easily. Paying for Claude would be investing that money. Assuming 10% return which would be great I'd make an extra $1200 a year investing it. I'm pretty sure over the course of a year of not having to spend time doing low value or repetitive work I can increase productivity enough to more than cover the $13k difference. Developer work scales really well so removing a bunch of the low end and freeing up time for the more difficult problems is going to return a lot of value.
It's *worth it* when you're salaried? Compared to investing the money? Do you plan to land a very-high-paying executive role years down the line? Are you already extremely highly paid? Did Claude legitimately 10x your productivity?
I'm serious - the productivity boost I'm getting from using AI models is so significant, that it's absolutely worth paying even 2k/month. It saves me a lot of time, and enables me to deliver new features much faster (making me look better for my employer) - both of which would justify spending a small fraction of my own money. I don't have to, because my employer pays for it, but as I said, if I had to, I would pay.
I am not paying this myself, but the place I work at is definitely paying around 2k a month for my Claude Code usage. I pay 2 x 200, for my personal projects.
I think personal subs are subsidized while corporate ones definitely not. I have CC for my personal projects running 16h a day with multiple instances, but work CC still racks way higher bills with less usage. If I had to guess my work CC is using 4x as little for 5x the cost so at least 20x difference.
I am not going to say it has 10xed or whatever with my productivity, but I would have never ever in that timeframe built all those things that I have now.
I don't know why you keep insisting that no one is making any money off of this. Claude Code has made me outrageously more productive. Time = Money right?
I'm an employee, and my boss loves me because I deliver things he wants quickly and reliably - because I use AI tools. Guess who he will keep in the next round of layoffs?
> It's very clear that Anthropic doesn't really want to expose the secret sauce to end users
Meanwhile, I am observing precisely how VS+Copilot works in my OAI logs with zero friction. Plug in your own API key and you can MITM everything via the provider's logging features.
How much longer is Anthropic going to allow OpenCode to use Pro/Max subscriptions? Yes, it's technically possible, but it's against Anthropic's ToS. [1]
It's amazing how much other agentic tools suck in comparison to Claude Code. I'd love to have a proper alternative. But they all suck. I keep trying them every few months and keep running back to Claude Code.
Just yesterday I installed Cursor and Codex, and removed both after a few hours.
Cursor disrespected my setting to ask before editing files. Codex renamed my tabs after I had named them. It also went ahead and edited a bunch of my files after a fresh install without asking me. The heck, the default behavior should have been to seek permission at least the first time.
OpenCode does not allow me to scrollback and edit a prior prompt for reuse. It also keeps throwing up all kinds of weird errors, especially when I'm trying to use free or lower cost models.
Gemini CLI reads strange Python files when I'm working on a Node.js project, what the heck. It also never fixed the diff display issues in the terminal; It's always so difficult for me to actually see what edits it is actually trying to make before it makes it. It also frequently throws random internal errors.
At this point, I'm not sure we'll be seeing a proper competitor to Claude Code anytime soon.
Same, I still use CC mainly due to it being so wildly better at compaction. The overall experience of using OpenCode was far superior - especially with the LSP configured.
I use Opencode as my main driver, and I don’t experience what you have experienced.
For instance, opencode has /undo command which allows you to scroll back and edit a prior prompt. It also support forking conversations based on any prior message.
I think it depends on the set up. I overwrote the default planning agent prompt of opencode to fit my own use cases and my own mcp servers. I’ve been using OpenAI’s gpt codex models and they have been performing very well and I am able to make it do exactly what I ask it to do.
Claude code may do stuff fast, but in terms of quality and the ability to edit only what I want it to do, I don’t think it’s the best. Claude code often take shortcuts or do extra stuff that I didn’t ask.
Not in my (limited) experience. I gave CC and codex detailed instructions for reworking a UI, and codex did a much worse job and took 5x as long to finish.
If they cared about that, they wouldn't expose the thinking blocks to the end-user client in the first place; they'd have the user-side context store hashes to the blocks (stored server-side) instead.
GitHub Issues as a customer support funnel is horrible. It's easy for them, but it hides all the important bugs and only surfaces "wanted features" that are thumbs-up'd alot. So you see "Highlight text X" as the top requested feature; meanwhile, 10% of users experience a critical bug, but they don't all find "the github issue" one user poorly wrote about it, so it has like 7 upvotes.
GitHub Codespaces has a critical bug that makes the copilot terminal integration unusable after 1 prompt, but the company has no idea, because there is no clear way to report it from the product, no customer support funnel, etc. There's 10 upvotes on a poorly-written sorta-related GH issue and no company response. People are paying for this feature and it's just broken.
Claude code can reverse engineer it to a degree. Doing it for more than a single version is a PITA though. Easier to build you own client over their SDK.
I think it's more classic enshittification. Currently, as a percentage, still not many devs use it. In a few months or 1-2 years all these products will start to cater to the median developer and start to get dumbed down.
Same here! American-made EVs ask for an incredibly heavy price tag and don’t deliver on the reliability of ICE or Hybrid cars a third of their price. It’s the primary thing stopping me from getting one as my next vehicle.
I’m trying to shop around to replace my wife’s aging crossover and I really can’t find anything more attractive than a Prius or another Kia Soul. If we could get electric cars from the CN market it’d be a no-brainer!
I’ve had software issues on an ID4 and iX, but I’ve never had reliability problems. The cars always have just worked with no maintenance. Same with my model Y, minus any issues!
Maybe it’s a “but when it happens you’re screwed” situation. I’m thinking of the story of BMW’s battery safety fuse (the one that trips in an accident to protect first responders and the people in the car) actually tripping when you hot the curb or a pothole harder. It requires a very expensive trip to the dealer. Some of my Tesla owning friends keep spending time in the shop getting something about the suspension fixed 2-3 times already.
I have no idea if Chinese EVs are consistently better, Volvo can be seen as one and I don’t think they excel at reliability lately.
P.S. Software issues are reliability issues. The software is a core part of the car and its value proposition, you can’t discount them as “just software issues, not reliability”.
> Some of my Tesla owning friends keep spending time in the shop getting something about the suspension fixed 2-3 times already.
They're pretty lucky from what I hear! A friend of mine just sold his Model S because he'd been waiting over 7 months for the shop to source a replacement part. Apparently he'd even resorted to begging Musk to look into it over X because Tesla wont even give him an ETA.
iD4 feels like they took every lesson of predictable UX design and then intentionally reversed it to make the most frustrating UI possible.
The window controls, touch buttons, screen, steering wheel controls, etc. They all seem designed to answer the question, "how could we make this unnecessarily difficult and distracting to use? How could we possibly cram in yet another State Machine for the user to keep (lose) track of?"
It also has the "try to kill the asthmatic by randomly switching off recirculate while driving through dense wood smoke" feature, naturally.
Considering how much money VW makes on EVs[0], I suppose I'm not surprised by this 'nudge' toward gas cars.
One of the good things in the UK is seeing how quickly my neighbours swapped out their Tesla's for BYD's.
I'm not in the market at the moment so don't know what the UK protectionism position is on Chinese EV's, but has been interesting to watch how quick it's happening.
I work in design and we're talking to two Chinese EV companies launching in the UK this year, so the wall can't be that high for them.
The UK has a local car manufacturing industry (Nissan, Jaguar/Landrover), but not large enough to be able to lobby for protectionism. And in any case the UK has basically given up on having a coherent trade policy since Brexit.
I've seen quite a few BYDs and MG4s, and there are Jaecoo and Leapmotor dealers near me. I've been told that some NHS boards were using MGs as "pool" cars, but the only example I can find a reference for is Shetland. https://www.nhsshetland.scot/news/article/43/nhs-shetland-ro...
I don't think I've ever seen a Rivian. The R2 is supposed to be coming to the UK in 2027.
The UK has a lot of "garage" brands too (the Morgans, Caterhams, BAC, etc.), but as you said they don't have a lot of lobbying power, and the lobbying they do is on behalf of lengthening THEIR transition to new EV requirements.
That said, the UK's history of small auto manufacturers would make it potentially ideal for a few domestic producers to make little EVs, similar to the Caterham 7, or the Ariel Atom for the domestic market, but they will never be the mass produced Tesla or BYD competitor.
Imo its not anti-China protectionism. Western models are cheaper in China, and Chinese models are more expensive in the EU & UK.
I think it's a combination of manufacturers wanting a higher profit, some adaptations & certification processes, dealer and service infrastructure necessary for selling in the West that just costs more.
I don't think Chinese manufacturers will be able to significantly undercut the competition while maintaining a desirable quality
I would have thought that to be the case too ... I know all about the exciting Xiaomi cars (e.g.) but I grew up in the era of the Chinese brand being "low quality". We're well beyond that now and have been for probably a decade. And don't get me wrong, I'm not a China cheerleader.
I give the example I mentioned. People local to me swapping out their model 3's for BYD. Maybe they just got to their end of their lease cycle and wanted to try something different, but I cannot believe they would have willingly chosen a significantly lower quality car (knowing some of them). And I believe the cost difference is marginal but the overall package just a bit better.
And you know people, they'll swap out anything for just a marginal saving. Doesn't have to be significant unless there's some network effects. And there really isn't with cars.
Anyway, I'm just yapping, but think the used Tesla market is going to get even more swamped than it already is. Not a bad thing because previously people looking for low cost cars were buying diesels - so I'm hopeful that'll transition to low cost EV's now ... but the game is up for Tesla automotive, but we've known that for some time.
I welcome the competition, but my two cents is that BYD or any other Chinese brand doesn't seem to be offering anything outside of the typicial price-performance calculus of existing brands.
I welcome competition, as it benefits everyone, even people who don't want to buy Chinese. It will also encourge building factories for EV components in Europe, which mean other suppliers might benefit from lower prices, and some of the savings will end up at the customer.
As for why your coworkers decided to go for non-Tesla EVs, you have to ask them, and ask them again a few years from now if it was worth it. In their defense, Tesla makes a very particular kind of car (in 2 slight variations), which many people might want to move away from.
Maybe they want a petrol range extender, maybe they want a more traditional SUV, or something smaller/bigger than Tesla.
Well, there’s also the fact that some organizations actually tie career progression to patents authored by the company. My brother works for a company that offers multiple technical career pathways to promotion, and one of them is essentially “obtain a software patent”
Great if you’re you, but try getting AverageSWE a local kube setup and see how quickly they ramp up on it
In my ideal world everyone would use kubernetes, it is the hammer and everything is a nail, but we must recognize that it is difficult for a lot of people to pick up.
That being said, if you’re deploying on kube in production, use kube locally. But if you’re not, dont
Using kube in production but really, even if I wasn't, I would still use the podman play kube approach. It isn't hard (at all) and isn't kubernetes, just kubernetes yaml. I actually find docker / compose a bit harder sometimes with the daemon running in the background.
I'm not arguing for the relative superiority of jsonnet vs yaml vs anything else. I just recognise that Docker Compose is loved by most open source developers. And invariably any project you touch will have a docker compose setup by default.
I'm just making it possible to run those on kubernetes seamlessly.
I mean if you are going to bother to introduce the concept of kubernetes yaml to a developer shouldn’t you just go all the way and teach them proper k8s instead of some weird intermediary? I fail to see the value of offering k8s yaml that isn’t k8s or one of its siblings that’s basically k8s
Tilt is great but it doesn’t solve the problem you’re asking about. This project more directly addresses that. Fundamentally the problem is that you want to maintain the lifecycle of several services during an ephemeral ci run and tear them down when you are done. As you mentioned it gets unwieldy and annoying to try to run all of these on a single machine and doubly so when you have a lot of services/containers. Kubedock is more like what you are looking for, it translates compose calls to Kube equivalents and each service in the compose file is it’s one kube pod with its own lifecycle. It should be possible under that to do what you are saying, spawn multiple docker composes from a single run.
It is worth noting that Kubedock has some really annoying limitations, part of it is that it’s one person the other part is that some concepts don’t translate to kube very well. So make sure that whatever you will be doing fits into those constraints before you try it
Why is obesity not considered a necessitating condition? It often carries the comorbidities you just mentioned. Should not exclude people just because they haven’t had these specific health problems (yet) but will eventually have them.
The problem is that if it's just about obesity, you have to prove that cheaper treatments such as diet and exercise didn't work. That's not impossible to do, but it's hard and annoying even for people who really were trying. My doctor told me that you basically have to keep a detailed journal of your weight loss efforts for months on end.
Are GLP-1s so much more effective that we should make an exception to the general principle, maximizing healthcare resources by providing the cheapest effective treatment? I kinda think so, but I have a conflict of interest, and I can understand why others might think that money is better spent elsewhere.
While I tend to agree, insurance companies don't see it that way. They need a doctor to indicate a necessity to treat a condition, as opposed to it being the easiest way to treat it.
For example, I have to take digestive enzymes to digest my food (pancreatic insufficiency). For someone with an unusually high metabolism, they would also give them a leg up on gaining weight, even though there are other approaches to gaining that weight. However in many cases, the insurance company wouldn't cover their prescription when they will mine.
As always it’s insurance nonsense. If incentives were aligned insurance companies would be lining up out the door to give this to obese people because they (the insurance companies) would eventually be on the hook for paying for the care of the conditions you just mentioned. It is very well demonstrated in literature that obese people have a much higher occurrence of these conditions than non obese people.
But the system is not set up with aligned incentives
Actions is many things. It’s an event dispatcher, an orchestrator, an execution engine and runtime, an artifact registry and caching system, a workflow modeler, a marketplace, and a secrets manager. And I didn’t even list all of the things Actions is. It’s better at some of those things and not others.
The systems I like to design that use GHA usually only use the good parts. GitHub is a fine events dispatcher, for instance, but a very bad workflow orchestrator. So delegate that to a system that is good at that instead
reply