"[Microsoft's] platform power didn’t just come from controlling applications on top of Windows, but the OEM ecosystem underneath. If OpenAI builds AI for everyone, then they are positioned to extract margin from companies up-and-down the stack — even Nvidia. "
Ah yes, the ChromeOS strategy. How'd that work out for Google?
Building a platform is good, a way to make quite a bit of money. It's worked really well for Google and Apple on phones (as Ben notes). But there's a reason it didn't happen for Google on PCs. Find it hard to believe it will for OpenAI. They don't (and can not) control the underlying hardware.
Whenever I see a post at #1 with not comment, I know it's been artificially pumped to the top. So many people upvoted but not a single comment yet. Let there be some comments! lol.
OpenAI is a geopolitically important play besides being a tech startup so it gets pumped in funding and in PR, to show that we're still leading the world. But that premise is largely hallucinated.
Platforms usually deliver significant value that is hard to replicate. OpenAI doesn't have any such thing. It's trivially replaced, and there's many competitors already. OpenAI is ahead of the curve, but they don't seem to have any particular way to do sticky capture. Migrating to a different LLM is an afternoon's work at most, not nearly the complexity of porting an app between OS' or creating a robust hardware driver model.
Openai is an artificially and ridiculously inflated balloon. It has nothing except initial market capture with hype. But yes they will keep whipping investors and keep burning money.
Can someone elucidate us as to how so many platforms (ChatGPT, Gemini, Claude, etc etc) all sprung up so quickly? How did the engineering teams immediately know how to go about doing this kind of tech with LLMs and DNNs and whatnot?
I wonder if OpenAI's app platform is going to be more like Windows (most economic value goes to user and app partners) or like Facebook (most economic value goes to facebook, app makers get screwed. I mean Microsoft acted badly towards a lot of partners, but it was a true platform.
OpenAI doesn't have a moat unfortunately. One URL replacement away and you can switch most models in minutes. I have personally done this many times over the last year and a half.
It only takes labs to produce better and better models, and the race to bottom on token costs.
I think platform monopolies are a thing of the past, where, when most of the world was asleep, a few silicon valley garage companies took over the green field and locked-in a huge customer base irreversibly, thus colonizing the world. The world is now much more awake and connected, ruling out any concentration of dominance. You can't have a repeat of colonial times.
I am not sure about this. They definitely created a brand new service and data flows that didn’t exist before and they have the majority of the mind share, however it’s already commoditized. The next two to three years will show how the chips fall. I can see that it’s tough or almost impossible for apple to get a share in this but google is right there to take the consumer side. For enterprise again we have to wait and see how gcp and AWS do.
The value is not in the llm but vertical integration and providing value. OpenAI has identified this and is doing is vertical integration in a hurry. If the revenue sustains it will be because of that. For consumer space again, nvidia is better positioned with their chips and SoCs but OpenAI is not a sure thing yet. By that I don’t mean they are going to fall apart, they will continue to make a large amount fmloney but whether it’s their world or not is still up in the air.
I'm on the verge of unsubscribing from Stratechery. The last month has been a bunch of fawning over Meta, YouTube, and constant talk about and fawning over OpenAI and whatever latest models are coming out. It's kind of tiring and boring. I swear I heard them talk about some YouTube influencers event like five times across their different shows and across time. Like, I do not care at all.
The current AI wave has been compared (by sama) to electricity and sometimes transistors. AI is just going to be in all the products. The trillion dollar question is: Do you care what kind of electricity you are using? So, will you care what kind of AI you are using.
In the last few interviews with him I have listened to he has said that what he wants is "your ai" that knows you, everywhere that you are. So his game is "Switching Costs" based on your own data. So he's making a device, etc etc.
Switching costs are a terrific moat in many circumstances and requires a 10x product (or whatever) to get you to cross over. Claude Code was easily a 5x product for me, but I do think GPT5 is doing a better job on just "remembering personal details" and it's compelling.
I do not think that apps inside chatgpt matters to me at all and I think it will go the way of all the other "super app" ambitions openai has.
Stratechery has always been shallow, but these overt advertisements are disturbing:
- Sneaking in how someone went from a Sora skeptic to a purported creator within a week.
- Calling the result the "future of creation".
- Titling the advertisement "It’s OpenAI’s World, We’re Just Living in It".
What they are doing here is to pitch Sora to attention deficit teenagers in order to have yet another way to make the hair of the favorite content creator red. As if that didn't already exist.
The victory lap from Sam Altman and all the money being raised makes people forget the following:
- Open source LLM models at most 12 months behind ChatGPT/Gemini;
- Gemini from Google is just as good, also much cheaper. For both Google and the users, as they make their own TPU;
- Coding. OpenAI has nothing like Sonnet 4.5
They look like they invested billions to do research for competitors, which have already taken most of their lunch.
Now with the Sora 2 App, they are just burning more and more cash, so people watch those generated videos in Tiktok and Youtube.
I find it hilarious all the big talk. I hope I get proven wrong, but they seem to be getting wrecked by competitors.
I don't like Ed Zitron automatic dismissal of everything AI and the constant profanities in his writing are getting old, and it's usually not very well structured, but that said... I like the perspective he has about the money involved.
OpenAI needs 1 trillion dollars in next four years just to keep existing. That's more than all currently available capital, together, from all private equity. If you take all private equity together, it still won't be enough for OpenAI in next four years.
It's just staggering amount of money that the world needs to bet on OpenAI.
The question is if the company can add more value to the models than someone else. I still see a lot of gaps in the ecosystem, eg evaluation/testing systems and integrations beyond chat interface and active control to get good results, not to mention other types of models that deal with 3d world or temporal data. There is an opportunity for an outsider to come and grab parts of the pie whilst the biggest are competing.
They do look like trying to grab the market with tooling but if you can use their tools (oss) and switch the models then where is the moat?
My feeling is that the tech industry has been in, "hot water" since at least 2018 and has been using private equity and bullshit hype-trains to garner interest in new technologies in lieu of the public getting hip to the fact that mostly computers are spying on you, making you mentally ill, and stealing your data in exchange for, "being able to participate." As pointed out by others OpenAI and the rest of the AI ecosystem will need a financial miracle to stay afloat and offer their products for a competitive price.
There's so much of what, "AI" is becoming that just seems like a massive psy-op to breathe one last breath of life into what is the skeleton of the old Silicon Valley. Innovation is possible but if the future really is liberal authoritarianism/oligarchy there's no room in the contrived market for, "innovative products that greatly improve human life."
I mean we transitioned from the product to the brand, with no reason to. Soon it will just flow back into a product. I don't care if a Samsung fridge has gemini GPT or openai GPT, as long as it works
I don't understand the link between the title of the article, and its content. If I summarize their three points:
1. Corporate strategy of OpenAI is becoming a monopole
2. OpenAI is investing in infrastructure because they think they'll have lots of users in the future
3. Making videos on Sora is fun, and people are gonna post more of these.
How does that substantiate "we live in OpenAI's world"? Am I missing something?
21 comments
[ 3.0 ms ] story [ 64.3 ms ] threadAh yes, the ChromeOS strategy. How'd that work out for Google?
Building a platform is good, a way to make quite a bit of money. It's worked really well for Google and Apple on phones (as Ben notes). But there's a reason it didn't happen for Google on PCs. Find it hard to believe it will for OpenAI. They don't (and can not) control the underlying hardware.
OpenAI is a geopolitically important play besides being a tech startup so it gets pumped in funding and in PR, to show that we're still leading the world. But that premise is largely hallucinated.
It only takes labs to produce better and better models, and the race to bottom on token costs.
The value is not in the llm but vertical integration and providing value. OpenAI has identified this and is doing is vertical integration in a hurry. If the revenue sustains it will be because of that. For consumer space again, nvidia is better positioned with their chips and SoCs but OpenAI is not a sure thing yet. By that I don’t mean they are going to fall apart, they will continue to make a large amount fmloney but whether it’s their world or not is still up in the air.
In the last few interviews with him I have listened to he has said that what he wants is "your ai" that knows you, everywhere that you are. So his game is "Switching Costs" based on your own data. So he's making a device, etc etc.
Switching costs are a terrific moat in many circumstances and requires a 10x product (or whatever) to get you to cross over. Claude Code was easily a 5x product for me, but I do think GPT5 is doing a better job on just "remembering personal details" and it's compelling.
I do not think that apps inside chatgpt matters to me at all and I think it will go the way of all the other "super app" ambitions openai has.
- Sneaking in how someone went from a Sora skeptic to a purported creator within a week.
- Calling the result the "future of creation".
- Titling the advertisement "It’s OpenAI’s World, We’re Just Living in It".
What they are doing here is to pitch Sora to attention deficit teenagers in order to have yet another way to make the hair of the favorite content creator red. As if that didn't already exist.
- Open source LLM models at most 12 months behind ChatGPT/Gemini; - Gemini from Google is just as good, also much cheaper. For both Google and the users, as they make their own TPU; - Coding. OpenAI has nothing like Sonnet 4.5
They look like they invested billions to do research for competitors, which have already taken most of their lunch.
Now with the Sora 2 App, they are just burning more and more cash, so people watch those generated videos in Tiktok and Youtube.
I find it hilarious all the big talk. I hope I get proven wrong, but they seem to be getting wrecked by competitors.
https://www.wheresyoured.at/the-case-against-generative-ai/
OpenAI needs 1 trillion dollars in next four years just to keep existing. That's more than all currently available capital, together, from all private equity. If you take all private equity together, it still won't be enough for OpenAI in next four years.
It's just staggering amount of money that the world needs to bet on OpenAI.
They do look like trying to grab the market with tooling but if you can use their tools (oss) and switch the models then where is the moat?
Pre-AI :: Examples :: Timeline
---------------------------------------------------------------------
NHI Non-human Intelligence :: e.g. dolphins, apes, crows etc. :: millions of years
HI Human Intelligence :: e.g. Einstein, Trump, Confuscius, Homer :: thousands of years
Post-AI
---------------------------------------------------------------------
AI Artificial Intelligence :: ChatGTP, Gemini, many others :: countable months
AGI :: Artificial General Intelligence ; not there yet; :: zero
AIApHI AI Assisted/Approved/Audited Human Intelligence :: See AI :: countable months
HIApAI Human Intelligence Assisted/Approved/Audited AI :: The Future? :: zero
I have mentioned no individuals here to avoid legal action. My point on AI is ... wait and see. Chill.
There's so much of what, "AI" is becoming that just seems like a massive psy-op to breathe one last breath of life into what is the skeleton of the old Silicon Valley. Innovation is possible but if the future really is liberal authoritarianism/oligarchy there's no room in the contrived market for, "innovative products that greatly improve human life."
There's hope in: https://worrydream.com/
1. Corporate strategy of OpenAI is becoming a monopole 2. OpenAI is investing in infrastructure because they think they'll have lots of users in the future 3. Making videos on Sora is fun, and people are gonna post more of these.
How does that substantiate "we live in OpenAI's world"? Am I missing something?