Even if Moore's law was still in effect and the computer resources required stayed the same and compute stayed as efficient per watt (neither is true), it would just halve compute costs every 18 months. You're able to read about people hitting $4000 costs/month on the $200 plan upthread. That's 8 years until it's cost effective.

Are they really ready to burn money for 8 years?

Uber operated at a loss for 9 years. They're now a profitable, market-winning business.

Amazon operated at a loss for 9 years, and barely turned a profit for over a decade longer than that. They're now one of the greatest businesses of all time.

Spotify operated at a loss for 17 years until becoming profitable. Tesla operated at a loss for 17 years before turning a profit. Palantir operated at a loss for 20 years before turning a profit.

And this was before the real age of Big Tech. Google has more cashflows they can burn than any of these companies ever raised, combined.

Uber is profitable now, on a yearly basis. But they still presumably have a long way to go until they're actually profitable in the sense that they've made a profit over the lifetime of their existence.

Amazon pivoted to a totally different business that was profitable from the get-go.

Google became profitable within a couple of years of starting and went into the black on total lifetime spend almost immediately. As was normal, back then.

Those aren’t good comparisons.

Uber operated at a loss to destroy competition and raised prices after they did that.

Amazon (the retailer) did the same and leveraged their position to enter new more lucrative markets.

Dunno about Spotify, but Tesla and palantir both secured lucrative contracts and subsidies.

Anthropic is against companies with deeper pockets and can’t spend to destroy competition, their current business model can only survive if they reduce costs or raise prices. Something’s got to give

They are good comparisons. All startups go against incumbents/competitors with deeper pockets.

Re: Anthropic specifically, I tend to agree, hence why I'm saying the deeper pockets (eg. Google, Amazon, etc) are perfectly positioned to win here. However, big companies have a way of consistently missing the future due to internal incentive issues. Google is deathly afraid of cannibalizing their existing businesses.

Plus, there's many investors with deep pockets who would love to get in on Anthropic's next round if their technical lead proves to be durable over time (like 6 months in AI terms).

This fight is still early innings.

It’s true startups go against deeper pockets, but I stand by my analysis since Uber / Amazon / Tesla (to a degree) were early tech companies going against old companies and not competing with others doing the exact same thing. They operated at a loss to defeat the old guard. Today that model doesn’t work well, and Anthropic are against deeper pockets that are doing nearly the exact same thing as them. If they were the only innovative company with huge outside investment against entrenched and unwilling to innovate older companies like Uber and Amazon then I’d agree there was a bigger chance.

And I like Anthropic, I want them to be successful, but they just can’t operate at a loss like this for long, they have to make some tough calls, and trying to cut corners behind the scenes is not good for long term trust

Haven't they already cannibalized search? It really sucks now.

Google search results "sucking" probably is an indication that they are squeezing money out of it well. Just because you don't like the results you are getting doesn't mean the average user isn't still using Google a ton and generating $$$ for Goog

Could this just be survivorship bias? How many companies burned money until they died? This isn't some hot take. I'm kinda interested. Surely more companies failed with this model than survived.

Anecdotally have worked for a company in the past that did just that and eventually went bankrupt. Know of many many more just in my city, for what it’s worth.

As long as Anthropic is better than their competitors they’ll continue to get my business.

Well, those companies were all successful, it's a bit of survivorship bias to only consider those. How many companies operated at a loss for years and eventually went out of business?

Those decades of 0 interest rate is now gone though. And if only 1 competitor will survive well like in all above sectors then sure it might feel fine now but almost a trillion dollars of private investment is waiting to sink.

I think people also expect models to be optimised over time. For example, the 5x drop in cost of o3 was probably due to some optimisation on OpenAI's end (although I'm sure they had business reasons for dropping the price as well).

Small models have also been improving steadily in ability, so it is feasible that a task that needs Claude Opus today could be done by Sonnet in a year's time. This trend of model "efficiency" will add on top of compute getting cheaper.

Although, that efficiency would probably be quickly eaten up by increased appetites for higher performance, bigger, models.

Every subscription service loses money on its heavy users. What matters is the average. Lots of people go for the higher plan because they need it once, then never downgrade even if their regular usage doesn't justify it. And then there are all the people paying for months where they don't use it at all

Among private users willing to pay $200/month average usage is likely very high, but if Anthropic can convince companies to buy plans for entire departments average usage will be much lower on those.

Another issue is that $4000 costs is assuming the regular API is offered at cost, which is unlikely to be true.

> They've just done the work to tailor it specifically for proper tool using during coding. Once other models catch up, they will not be able to be so stingy on limits.

I don't subscribe to the $100 a month plan, I am paying API usage pricing. Accordingly I have learned how to be much more careful with Claude Code than I think other users are. The first day I used it, Claude got stuck in a loop trying to fix a problem using the same 2 incorrect solutions again and again and burnt through $30 of API credits before I realized things were very wrong and I stopped it.

Ever since then I've been getting away with $3-$5 of usage per day, and accomplishing a lot.

Anthropic needs to find a way to incentivize developers to better use Claude Code, because when it goes off the rails, it really goes off the rails.

> The first day I used it, Claude got stuck in a loop trying to fix a problem using the same 2 incorrect solutions again and again and burnt through $30 of API credits before I realized things were very wrong and I stopped it.

The worse it performs, the more you pay. That’s a hell of a business model. Will users tolerate that for long?

> The worse it performs, the more you pay. That’s a hell of a business model. Will users tolerate that for long?

I mean, AWS seems to be doing fine with that business model.

The thing is, all the models are not that 'smart'. None of them is AGI.

Currently it's much more important to manage context, split tasks, retry when needed, not getting stuck in an infinite loop, expose the right tools (but not too many), ...

Prices for yesterday's frontier models will fall but there will always be the next big model. similar to how game graphics get ever better but ever more demanding at the bleeding edge.

Yes but games also look an awful lot better (fidelity wise) than not so many years ago.

The problem with models is that they create lots of junk content.

Industries can often get away with polluting when they're small, but once they reach planet scale salting the earth behind you is not as reliable of a tactic.

Claude is decent for sure, but if you are using these models for 'smarts', that is a whole separate problem. I also think honestly people are sleeping on Mistral's medium 3 and devstral medium. I know it isn't 'smart' either (none of them are), but for mundane tasks need valid code output, it is extremely good for the price.

I use o3 to brainstorm research problems, and it's been pretty useful. Especially the deep research feature.

As a sounding board for things you are already well familiar with, I agree, and have experienced the same, and that can be useful. It's also a much better experience than say using Google to do the same, or just a rubber ducky.

The NLP these models can do is definitely impressive, but they aren't 'thinking'. I find myself easily falling into the habit of filtering a lot of what the model returns and picking out the good parts which is useful and relatively easy for subjects I know well. But for a topic that I am not as familiar with, that filtering (identifying and dismissing) I do is much less finessed, and a lot of care needs to be taken to not just accept what is being presented. You can still interrogate each idea presented by the LLM to ensure you aren't being led astray, and that is still useful for discovering things, like traditional search, but once you mix agents into this, things can go off the rails far too quickly than I am comfortable with.

Will the other models really catch up, though? To me it seems like Anthropic's lead in programming has increased over the past year. Isn't it possible that over time, some models just become fundamentally better at some things than other models?

> Will the other models really catch up, though? To me it seems like Anthropic's lead in programming has increased over the past year.

To me it seems that they are all converging over time. So what if one is "better"[1]? The others will soon be there too.

[1] My experience with them is that Claude (Opus) is only slightly better than the competition at coding; it might do a better job on 2 out of every 5 tasks than the competition, but those people using the competition still get those two tasks done anyway.

I've used Augment before moving to Claude, they are pretty similar, often can't tell the difference. I don't think there is that much difference in the dev focused llms.

I mean, not based on anything we’ve seen so far in the DL space. The algorithms are public, the compute is fungible: the only differentiator is data. But deepseek demonstrates that it’s somewhat easy to siphon data off other models so… yeah unclear where the moat is.

I think you're right, but you're also ignoring the effects of monopolies and/or collusion. There's a absolutely a chance prices don't come down due to broader anti-competitive plays.