Anthropic's valuation triples to $61.5B in bumper AI funding round | Hacker News Reader

Anthropic's valuation triples to $61.5B in bumper AI funding round

ft.com

20 pointsmarban26 comments

Is it right that they raise $3.5B, projected to burn $3.0B this year with ~$1.2B in revenue _expected_ this year?

So either prices are going to rocket up at some point and/or massive efficiencies? That doesn’t seem like a good position. I like their models, but not sure how they’ll survive long term.

When a large number of super smart people with a strong track record of accomplishment do things that don't make sense to me, my first thought isn't that they obviously don't know what they're doing, my first thought is that there might be something I don't understand.

Like Meta investing in virtual and augmented reality projects over $100 billion? Lie Google investing half a $ billion in Magic Leap?

Like all the crypto?

or usecase/adoption increase ? also : capex!=operating costs

Brain tried to parse your comment as a trinary operator.

> capex!=operating costs

True, but when your only moat is capabilities, and those degrade with time and use, it might as well be.

I think many people believe this is an absolute bubble, that these valuations do not correspond to fundamental values of the business and are just based on promises and fraud.

I don't like having such a black and white view of the world though, so I would be interested in seeing if there is any sane reason for these valuations. Do you think these businesses would be as highly-valued if there was not all this hype around? How would you justify this valuation?

  I think many people believe this is an absolute bubble

I wonder why. I've never came across a product that changed the way I work and get information for the better as fast as LLMs.

> many people believe this is an absolute bubble

The authors of "The internet is just a fad"?

FWIW, dotcom bubble was real. I also find myself frequently thinking about the "railroad era" in the US - from what I read, that too was a bubble, and companies that built up rail lines across the continent all went bust, leaving investors out to dry.

In both cases, even though the investors lost money when the bubble popped, the infrastructure remained, and allowed the next generation of companies to flourish and transform the economy and lives of everyone.

Perhaps this is a pattern? That laying down infrastructure and building up expertise are critical, but people doing it don't get to reap the rewards of it.

Google and Amazon survived

LLMs are amazing and probably as important a technology as electricity, the internal combustion engine, radio, or smartphones, but $61.5 billion can still be waaaay overvalued for Anthropic.

At some point in the next year or two, if you were to make me give up the browser (and all websites) or an LLM, I might choose the browser.

Really? What does it do for you? Maybe I'm limited by imagination or the free plan but it seems like I get faulty info or hallucinations more than I get game-changing productivity boosts.

> or the free plan

There's your answer.

Yes, on a free plan, you don't get to experience the state of the art that excites people so much. Models are weaker and more prone to hallucinate, and limits are stricter - both of which makes the services much less useful for causal use.

Pay $20 and play with ChatGPT for a month (or go the API route and find good software that lets you supply your own API key), see what you get. Yes, technically ChatGPT has this $200 tier, but I don't think it makes a qualitative difference for causal use, given that (apparently, as I discovered two days ago), even the "Deep Research" thing is now available on the $20 tier - in limited amount (I have 10 queries per month), but enough to give it a solid test drive.

One possibility is diversification. By and large each successive tech iteration resulted in one big winner: Microsoft (operating systems), Amazon (e-commerce), Google (search), Apple (mobile), Meta (social networks), Netflix (streaming), and depending how you look at it, Tesla (EVs). If you invested in one of these companies early, you'd see a huge profit one or two decades later. The problem, of course, is that you don't have a crystal ball. However, if you invested in these companies and a few of their peers in the early days, then you lost all the money in their peers, made a lot of money in these companies, and you are still well ahead in the game.

At this point, if you don't know if Anthropic or OpenAI or Mistral will be the ultimate winner, you invest in all of them. Of course, there's a chance that another startup comes out of nowhere and all your investments tank, but nothing is risk free in life.

I don't think any startup in the past based their valuation on fundamentals. Not even the wildly successful ones like Facebook.

People can't seem to distinguish between promises, bets, and fraud.

Of course, as a start-up you base your valuation on potential in the future, which should be well calculated based on how capable your employees are, how innovative your product is and what are the expected earnings.

Do you think Anthropic is still a startup? That's an honest question, Wikipedia certainly seems to think so, (Anthropic PBC is an American artificial intelligence public-benefit startup founded in 2021.). I would typically say that once you have 61B valuation, you have to have a fundamentally sound business, and thus cease to be a startup, but I don't know where the line is drawn.

Bringing up fraud is jarring. Failed bets are not fraud. Fraud is a huge accusation. There's a huge distinction between the twos.

> which should be well calculated based on how capable your employees are, how innovative your product is and what are the expected earnings.

If anyone can calculate this accurately, they would easily become a billionaire. Let's not pretend you (or anyone) know a good calculation lmao.

Just because your personal "calculation" doesn't deem Anthropic worthy of the valuation / investment (and you might eventually be right) ... doesn't mean it is a fraud.

Yes. The investors here are making a bet that AI goes supermassive and brings in a load of money.

Some people at anthropic almost certainly do think they're going to be successful. I suspect that at least some of them are engaging in salesmanship at a level that makes it fraud.

Extraordinary claims require extraordinary evidence.

It would be helpful if you would identify exactly which claim you thought was extraordinary so that people could engage with your comment productively.

This one: "I suspect that at least some of them are engaging in salesmanship at a level that makes it fraud."

That's a suspicion, as I made clear. And thus based on circumstancial evidence. Were I to have actual proof then I would have been more forthright in making a claim.

A suspicion of something bad and illegal is an accusation. There's really no real distinction.

You wouldn't say this kind of things to anyone without an evidence.

"Hey boss. I suspect my teammate commits a fraud. But it's just a suspicion! I don't have any evidence!" Now replace fraud with equally heinous crimes like pedophilia and murder.

http://archive.today/2025.03.03-183004/https://www.ft.com/co...

Reply on news.ycombinator.com