6 ms·

The are valued at $170 Billion. Not quite the same as, but in same order of magnitude as OpenAI - while having only a single digit percent fraction of active us

The are valued at $170 Billion. Not quite the same as, but in same order of magnitude as OpenAI - while having only a single digit percent fraction of active users. They probably need to prepare for the eventual user data sellout, as it is becoming increasingly more obvious that none of the big players has a real and persistent tech leadership anymore. But millions and millions of users sharing their deepest thoughts and personal problems is gonna be worth infinitely more than all the average bot bullshit written on social media. That's also why Zuck is so incredibly desperate to get into the game. It's not about owning AI. It's about owning the world's thoughts and attention.

Imustaskforhelp 1y ago
The last time my brother and I were discussing about anthropic, they were worth 90B$, and that was a month ago, he asked chatgpt in the middle of the conversation, either it was a sneaky sabotage from gpt or my memory is fuzzy but I thought that 90b$ was really underrated for anthropic given the scaleAi deal or windsurf/cursor deals.
- sigmoid10 1y ago
  >I thought that 90b$ was really underrated for anthropic That was true when the tech leadership was an open question and it seemed like any one of the big players could make a breakthrough at any moment that would propel them to the top. Nowadays it has pattered out and the market is all about sustainable user growth. In that sense Anthropic is pretty overvalued, at least if you think that OpenAI's valuation is legit. And if you think OpenAI is overvalued, then Anthropic would be a no-go zone as an investor.
  - Imustaskforhelp 1y ago
    Note that it was before kimi k2 (I think) and as such back when anthropic was truly the best in class back then at coding and there wasn't any competition and every day on Hackernews would be filled about someone writin something about claude code. And the underrated comparison was more towards the fact that I couldn't believe scaleAi's questionable accquisition by facebook and I still remember the conversation me and my brother were having which was, why doesn't facebook pay 2x, 3x the price of anthropic but buy anthropic instead of scaleAI itself well I think the answer my brother told was that meta could buy it but anthropic is just not selling it
goalieca 1y ago
Companies all seem to turn against their users whenever they have revenue/earnings trouble.
- jsheard 1y ago
  Considering every AI company is hemorrhaging money with no end in sight, that doesn't bode well, does it?
- diggan 1y ago
  It seems to me like some fundamental/core technologies/services just shouldn't be run by for-profit entities, and if come across one doing that, you need to carefully choose if you want to start being beholden to such entity. As the years go by, I'm finding myself being able to rely on those less and less, because every time I do, I eventually get disappointed by them working against their user base.
  - bigfishrunning 1y ago
    Except LLMs aren't a fundamental or core technology, they're an amusing party trick with some really enthusiastic marketers. We don't need them.
    - komali2 1y ago
      Under the current system we apparently do since Chatgpt is now by far and away the busiest psychiatrist in world history. I don't think we should be so quick to dismiss the holes LLMs are fulfilling as unnecessary. The only thing "necessary" is food water and shelter by some measures.
      - psychoslave 1y ago
        Do we have statistics on this? What is the healing rate compared to human psychotherapy if so?
    - diggan 1y ago
echelon 1y ago
> while having only a single digit percent fraction of active users. That doesn't matter when their revenue per user is as high as it is. They're at $5B ARR and rapidly growing.
- Eggpants 1y ago
  And yet rapidly still no where close to running a profit. Time to push the "Its a bargain at $500/month!!!!" narrative. Once they admitted they are going to have to take money from folks who chop up journalists that made them feel sad, they proved the current pre token LLM based business model doesn't work. They haven't pulled the ads lever yet but the writing is on the wall. Which means sadly only business with other revenue streams like M$, the Google, or Amazon can really afford it long term. I'm was rooting for Anthropic but it doesn't look good.
- mapontosevenths 1y ago
  The "killer app" isn't here yet. Wait until smart glasses or watches with AI overtake cellphones as the primary method of human interaction with computers and most websites are replaced with API's that only AI's really ever use.
  - echelon 1y ago
    > most websites are replaced This is already happening. > Wait until smart glasses or watches with AI overtake cellphones Smartphones are crystalized perfection. It's such a peak design. The size, form factor, sensors, input/output modalities, and generalization are perfect. The reason companies are trying to supplant it is that they need to get out from under Google and Apple's control. It's not that anything is wrong with the smartphone. VR has a long way to go in terms of hardware problems. XR/AR is ridiculous. It's creepy, unstylish, and the utility is highly questionable. Nobody is going to want to be a walking ad.
    - mapontosevenths 1y ago
      >Smartphones are crystalized perfection. Time will tell, but to me they feel like desktops did 20 years ago. The process of enshitification has turned simple tasks complicated and everyone wants a different, privacy destroying, frustrating to use "app", each of which has a slightly different UI paradigm, a mandatory subscription I've forgotten to cancel for two years straight, and a confusing name to remember. I now have something like 90 apps installed on my iphone, and I can only remember what something like 40 of them do. My damn cat box has an app, and instead of naming it something sensible like "Shitbox 2000" they named it "Whisker". Was it "Foober Eats that had taco bell, or Instafart, maybe it was Dine-N-Dash? Where's the back button on this thing and why is it different from every other app? Is this an ad or content, does it even matter anymore? Why do I need another login, what happened to SSO? Why won't my password vault work for this one app? Did I register for this one with my google account or apple? Who took my pills? Stay off my lawn!" When the day comes that I can just tell my device what to do, and let it get it done I'll be very happy to dump that cognitive load onto someone/something else.
conradev 1y ago
It’s worth noting that companies at this scale are usually the ones purchasing user data, not selling it.
- sigmoid10 1y ago
  Only if they are in the ad selling business.
mrcwinn 1y ago
There is far more money to be made building atop this data than selling this data. Your opening statement seems to disagree with your closing statement.
- bethekidyouwant 1y ago
  is there money to be made? I thought they were all losing money…
  - vntok 1y ago
    Investments are all about "losing" money first to make money later, it's not a paradox.
- sigmoid10 1y ago
  All of this boils down to selling to advertisers. There is no real difference between doing it yourself or having someone else in the chain. Doing it yourself may be more profitable - if you can scale. But that seems to be off the table here.
paradite 1y ago
Claude Sonnet 4 is the best coding model. Period. Nothing else comes close. Anthropic probably has 80% of AI coding model market share. That's a trillion dollar market.
- Barrin92 1y ago
  >That's a trillion dollar market not if you have to constantly expend enormous sums to stay ahead of your competition or otherwise you lose your edge. It's not the best coding model because they got some mystical treasure in their basement. It's so rapidly becoming a commodity that at some point Microsoft or Google will just offer just as good a model for free and like search they'll just start milking people with ads. That's likely one of the reasons for the shifting privacy stances, not just for training but because monetization of the product itself is probably looking pretty dim in the long run.
tsunamifury 1y ago
The data is no where near valuable enough without a new high value surface to use it, and so far chat is not it. Merely selling data is extremely low value compared to also having the surface monopoly to monetize it in a very high engagement and decisioning space. I feel like you don’t understand the fundamental mechanics of the ad world. Ultimately, the big 4 own such immense decisions surface area it may be a while before any AI model company can create a product the get there.
Romario77 1y ago
Anthropic enterprise share is pretty significant - on order of 30%. I think at this time it's pretty significant. I am expecting AI companies to start using ads, it's inevitable as they need to make money at some point and $20 a month won't do it. For ads the number of users is the main thing - the more users you have the bigger the market and more money you could earn. Google desperately needs to be in this space, that's why they are throwing a ton of money on AI.
- sigmoid10 1y ago
  30% of ~4% is very little when you think about these valuations.

Imustaskforhelp 1y ago

The last time my brother and I were discussing about anthropic, they were worth 90B$, and that was a month ago, he asked chatgpt in the middle of the conversation, either it was a sneaky sabotage from gpt or my memory is fuzzy but I thought that 90b$ was really underrated for anthropic given the scaleAi deal or windsurf/cursor deals.

sigmoid10 1y ago
>I thought that 90b$ was really underrated for anthropic That was true when the tech leadership was an open question and it seemed like any one of the big players could make a breakthrough at any moment that would propel them to the top. Nowadays it has pattered out and the market is all about sustainable user growth. In that sense Anthropic is pretty overvalued, at least if you think that OpenAI's valuation is legit. And if you think OpenAI is overvalued, then Anthropic would be a no-go zone as an investor.
- Imustaskforhelp 1y ago
  Note that it was before kimi k2 (I think) and as such back when anthropic was truly the best in class back then at coding and there wasn't any competition and every day on Hackernews would be filled about someone writin something about claude code. And the underrated comparison was more towards the fact that I couldn't believe scaleAi's questionable accquisition by facebook and I still remember the conversation me and my brother were having which was, why doesn't facebook pay 2x, 3x the price of anthropic but buy anthropic instead of scaleAI itself well I think the answer my brother told was that meta could buy it but anthropic is just not selling it

goalieca 1y ago

Companies all seem to turn against their users whenever they have revenue/earnings trouble.

jsheard 1y ago
Considering every AI company is hemorrhaging money with no end in sight, that doesn't bode well, does it?
diggan 1y ago
It seems to me like some fundamental/core technologies/services just shouldn't be run by for-profit entities, and if come across one doing that, you need to carefully choose if you want to start being beholden to such entity. As the years go by, I'm finding myself being able to rely on those less and less, because every time I do, I eventually get disappointed by them working against their user base.
- bigfishrunning 1y ago
  Except LLMs aren't a fundamental or core technology, they're an amusing party trick with some really enthusiastic marketers. We don't need them.
  - komali2 1y ago
    Under the current system we apparently do since Chatgpt is now by far and away the busiest psychiatrist in world history. I don't think we should be so quick to dismiss the holes LLMs are fulfilling as unnecessary. The only thing "necessary" is food water and shelter by some measures.
    - psychoslave 1y ago
      Do we have statistics on this? What is the healing rate compared to human psychotherapy if so?
  - diggan 1y ago

echelon 1y ago

> while having only a single digit percent fraction of active users. That doesn't matter when their revenue per user is as high as it is. They're at $5B ARR and rapidly growing.

Eggpants 1y ago
And yet rapidly still no where close to running a profit. Time to push the "Its a bargain at $500/month!!!!" narrative. Once they admitted they are going to have to take money from folks who chop up journalists that made them feel sad, they proved the current pre token LLM based business model doesn't work. They haven't pulled the ads lever yet but the writing is on the wall. Which means sadly only business with other revenue streams like M$, the Google, or Amazon can really afford it long term. I'm was rooting for Anthropic but it doesn't look good.
mapontosevenths 1y ago
The "killer app" isn't here yet. Wait until smart glasses or watches with AI overtake cellphones as the primary method of human interaction with computers and most websites are replaced with API's that only AI's really ever use.
- echelon 1y ago
  > most websites are replaced This is already happening. > Wait until smart glasses or watches with AI overtake cellphones Smartphones are crystalized perfection. It's such a peak design. The size, form factor, sensors, input/output modalities, and generalization are perfect. The reason companies are trying to supplant it is that they need to get out from under Google and Apple's control. It's not that anything is wrong with the smartphone. VR has a long way to go in terms of hardware problems. XR/AR is ridiculous. It's creepy, unstylish, and the utility is highly questionable. Nobody is going to want to be a walking ad.
  - mapontosevenths 1y ago
    >Smartphones are crystalized perfection. Time will tell, but to me they feel like desktops did 20 years ago. The process of enshitification has turned simple tasks complicated and everyone wants a different, privacy destroying, frustrating to use "app", each of which has a slightly different UI paradigm, a mandatory subscription I've forgotten to cancel for two years straight, and a confusing name to remember. I now have something like 90 apps installed on my iphone, and I can only remember what something like 40 of them do. My damn cat box has an app, and instead of naming it something sensible like "Shitbox 2000" they named it "Whisker". Was it "Foober Eats that had taco bell, or Instafart, maybe it was Dine-N-Dash? Where's the back button on this thing and why is it different from every other app? Is this an ad or content, does it even matter anymore? Why do I need another login, what happened to SSO? Why won't my password vault work for this one app? Did I register for this one with my google account or apple? Who took my pills? Stay off my lawn!" When the day comes that I can just tell my device what to do, and let it get it done I'll be very happy to dump that cognitive load onto someone/something else.

conradev 1y ago

It’s worth noting that companies at this scale are usually the ones purchasing user data, not selling it.

sigmoid10 1y ago
Only if they are in the ad selling business.

mrcwinn 1y ago

There is far more money to be made building atop this data than selling this data. Your opening statement seems to disagree with your closing statement.

bethekidyouwant 1y ago
is there money to be made? I thought they were all losing money…
- vntok 1y ago
  Investments are all about "losing" money first to make money later, it's not a paradox.
sigmoid10 1y ago
All of this boils down to selling to advertisers. There is no real difference between doing it yourself or having someone else in the chain. Doing it yourself may be more profitable - if you can scale. But that seems to be off the table here.

paradite 1y ago

Claude Sonnet 4 is the best coding model. Period. Nothing else comes close. Anthropic probably has 80% of AI coding model market share. That's a trillion dollar market.

Barrin92 1y ago
>That's a trillion dollar market not if you have to constantly expend enormous sums to stay ahead of your competition or otherwise you lose your edge. It's not the best coding model because they got some mystical treasure in their basement. It's so rapidly becoming a commodity that at some point Microsoft or Google will just offer just as good a model for free and like search they'll just start milking people with ads. That's likely one of the reasons for the shifting privacy stances, not just for training but because monetization of the product itself is probably looking pretty dim in the long run.

tsunamifury 1y ago

The data is no where near valuable enough without a new high value surface to use it, and so far chat is not it. Merely selling data is extremely low value compared to also having the surface monopoly to monetize it in a very high engagement and decisioning space. I feel like you don’t understand the fundamental mechanics of the ad world. Ultimately, the big 4 own such immense decisions surface area it may be a while before any AI model company can create a product the get there.

Romario77 1y ago

Anthropic enterprise share is pretty significant - on order of 30%. I think at this time it's pretty significant. I am expecting AI companies to start using ads, it's inevitable as they need to make money at some point and $20 a month won't do it. For ads the number of users is the main thing - the more users you have the bigger the market and more money you could earn. Google desperately needs to be in this space, that's why they are throwing a ton of money on AI.

sigmoid10 1y ago
30% of ~4% is very little when you think about these valuations.