Hacker News
Anthropic reportedly prices its IPO at $83 per share, valuing at $1.5T
barbariangrunge
|next
[-]
ncd
|next
|previous
[-]
barbariangrunge
|root
|parent
|next
[-]
ie: for the price, I'd rather invest in a smaller lab that doesn't need to conquer the world to justify it's share price
donmcronald
|root
|parent
|next
[-]
So the big thing IMO is that normal people don’t understand the risk of getting dependent on an AI company that has all your “memories” and that remembers every conversation perfectly. Imagine 20 years of context locked behind a subscription.
6thbit
|root
|parent
|next
|previous
[-]
They also market themselves as a public benefit company and safety driven, which are not priorities for growth maximization. Their main page says "Anthropic is a public benefit corporation dedicated to securing (AI's) benefits and mitigating its risks."
Then there's cost pressure from the competition, like oAI's cost per task being quite cheaper with only marginally less intelligence than ant's models.
So there doesn't seem to be a clear path to long term upside. But short to mid term and before OAI's own IPO it's likely to attract more money than it should.
debrisontrack
|root
|parent
|next
|previous
[-]
CamperBob2
|root
|parent
|previous
[-]
Success at pulling off what he claims they are trying to pull off -- owning all aspects of AI regulation, basically -- will indeed position them for growth, albeit at everyone else's expense.
CrimsonCape
|root
|parent
|next
[-]
So you get to see a public model but have no concept of how much better the internal model is than the public model. But more critically, how much the internal models push beyond each prior generation. My guess is that eventually they will internally see diminishing feedback at some point, suggesting they have hit the cap of the limits of the current frontier architecture.
That feels like the logical pivot point.
My guess is that the keep-up game is only going to last throughout the IPOs, and then the big names will settle (or collude...) on dropping new models every 6 months or a year. The only thing I can imagine keeping that from happening is the deepseek competition.
NBJack
|root
|parent
|next
|previous
[-]
estearum
|root
|parent
|previous
[-]
> China is keeping messaging tight. That is why optimism for AI is so high in China.
Nah, China is positive on AI because China's living memory is waves of technological advancement pulling a billion people out of destitute poverty. Meanwhile, America's living memory is technology destroying the fabric of our society and reptilian CEOs dancing on the shreds.
> Anthropic's business model, models itself after the very thing it claims to fear.
This feedback loop is itself what Anthropic (and other AI doomers) have been drawing attention to for decades now. It's hilarious seeing dummies like Kevin "realize" this and think that it's some big secret that their big brain investigations uncovered and not literally the entire problem.
The question is, if you believe this feedback loop exists or is inevitable (today we have empirical validation), then what to do about it? Anthropic's answer is "be first and be better," which self-referentially is the problematic dynamic to begin with, but merely identifying it as a problematic dynamic does not eliminate it or create a way to disrupt it by simply opting out. That's why, in fact, it's a problem at all.
6thbit
|next
|previous
[-]
Today Anthropic, PBC priced its initial public offering of its Class A common stock at $83.00 per share. The offering consists of 980,000,000 shares of Class A common stock, 956,351,354 of which are being sold by Anthropic and 23,648,646 of which are being sold by certain selling stockholders. Anthropic will not receive any proceeds from the sale of shares by the selling stockholders.
https://www.anthropic.com/news/anthropic-prices-initial-publ...
duderific
|root
|parent
[-]
OkayPhysicist
|root
|parent
|next
[-]
Sohcahtoa82
|next
|previous
[-]
The current modern market doesn't actually care about fundamentals when it comes to certain hype stocks. It's all vibes, memes, and cult following.
People will defend buying Tesla stock despite car sales slipping and their P/E being an absolutely ridiculous 422.
So the question is...can Anthropic get people to buy their stock and ignore fundamentals? If so, the number goes up.
cookiengineer
|next
|previous
[-]
I don't understand this, to be honest. Claude is such a stupid model compared to DeepSeek v4.1 or GLM 5.3 or even Qwen3.8 flash next, especially in cyber security related tasks.
I get their marketing strategy, but it's just that. And they don't deliver on their promise and haven't done so since November anymore. They had the lead with Opus 4.6 but it got gradually so much worse that pretty much all people working in cyber that I know of all switched to different models or started to selfhost.
They (including me) were happy customers back in January, though.