How does one justify this valuation in a world DeepSeek and Qwen are free?
Most people going, “these valuation are insane” because these AI companies are not profitable and doesn’t seem to be a clear path to profitability.
That's what OpenAI head said in an intreview
To recreate all Microsoft’s meaningful profit-generating offerings would cost a very small fraction of $3T. In SaaS space startups are competing with MS products for much cheaper. In operating systems, commercial Linux (RHEL, SteamOS, etc) is competing for much cheaper. In gaming, Microsoft has contributed much to game cost inflation, meaning others compete for cheaper. In hardware others compete for cheaper. In home office, even Google and Apple likely spend less to compete. Maybe in web services Azure leads in costs and has some moat, but I’m only saying this because I don’t know it well.
This is not to say MS is not an impressive company, but $3T is not its real value. From the stock market side, PE of 32-ish suggests it’s meaningfully over-valued too.
So I know you kind of meant that question in a half joking way, but it is a good question. You can’t explain $3T with real value, it has to be speculative or bubble-ish.
Comparing how much money it would take to build a competitive portfolio of products to compete with Microsoft is like comparing how long it would take to build a Twitter or Facebook clone. Sure, it wouldn't take that long or that much money, but product is the easy part.
We can also ask whether Microsoft's market is free, from a macroeconomic perspective. I would say it is not. There is still a very significant consumer lock-in from Microsoft's dominant position. Some would even argue that this lock-in is monopolistic as Microsoft created many markets it inhabits (operating systems, home publishing, small business software, PC gaming, and similar). I don't mean to criticize here - monopolistic is a very loaded word politically. Objectively, It's natural that businesses are monopolistic for some time when they create markets. But not forever.
My point is that Microsoft's monopolistic position is not sustainable unless it provides services that are either significantly differentiated or cheaper than the competition. As such, it should not command its market cap. Microsoft understands this, as its strategy is to establish itself as a monopoly in emerging markets (like AI) by using "embrace and extend," a business tactic they are known for. But I don't believe monopolies in moat-less markets are easy to sustain. This, once again, brings the market cap into question.
Perhaps I am missing something; this is just my opinion, anyway. But I am trying to look at it objectively from a corporate finance perspective. It's easy to get lost in marketing speak. Microsoft does a lot of "upward management" aimed at its public shareholders, so it's important to look at the fundamental facts. Microsoft is an impressive company, and it has many things to brag about, but some matter more in its actual business economics, and some matter less.
I don't believe infrastructure matters as much as the monopolistic lock-in they have (for now) and the competitive edge they don't. Perhaps they will find a more sustainable business model as it's clear from how their products and offerings have been developing that they are urgently trying to.
This is the important point. A SaaS company with a competitive offering isn't participating in the economy. Participation requires distribution. Distribution consists of on the ground sales, channel partners, post sales support, marketing, education, a developer ecosystem, trust, security and compliance buy in, workforce training, and a host of other components that a new company doesn't have. The vast majority of startups with great products perish because they don't have this and there's few shortcuts.
Monopolistic practices are really neither here nor there. You can see distribution win over product quality in every industry. Out of product, brand, and distribution, product contributes the least to market cap.
No or government in the world currently functions without Microsoft...
Costs a bit of money, but dealing with Google Docs, Synology, Dropbox, email and web hosting providers took so much effort. Setting up Sharepoint etc. isn't trivial either but at least it's a standard setup that you'll always find someone who can manage it.
Word, Excel and Powerpoint are the standard tools everyone uses to get their job done.
Until lots of free AI models close the gap again. It is simply not good enough for such valuations like this for Anthropic.
Any other valuation higher than 70B for Anthropic is really overvalued and such higher valuations are always defended by those who invested in it already.
I realize that on benchmarks it appears R1 is very close to Sonnet 3.7, but in practice I use Sonnet 3.7 much more.
I'm out of the loop. To whom are you referring?
It really seems impossible to justify these evaluations from any kind of economic perspective.
Sonnet with thinking is at 64.9% (the top), but it's pricey - about double the cost without thinking.
A model that is 50% less likely to try to prove false theorems tho is much more useful than than one that happily tries.
I'm not sure that HN folks appreciate this, but making commercial software is way easier than math and science (all the evidence you need, except the experience of understand QFT or advanced mathematics) is found in HN itself.
No matter who you are, the rest of the world combined will always be bigger. That is why Linux won the cloud, and why Windows server simply could not compete.