Comment by mbesto
2 days ago
> Meta has 3.5 billion users and projected ~$200b revenue in 2025
Meta makes about $200B on ads, Google makes about $235B on ads. Advertising is roughly 1.5% of the total GDP of the US and hasn't changed in 20+ years. So what you have is a big ass pie with a few players fighting for it that barely grows every year.
OpenAI has to somehow:
1. Compete directly with Google Gemini and Meta's Llama for a piece of users pie with a product that has very little differentiator (functionality and technically speaking).
2. Have to prove to advertisers that their single dollar ad purchase on OpenAI is categorically worth more than any other channel.
3. Have enough forward capital to continue purchasing capital-intense hardware purchases.
4. Having enough capital to weather any potential economic headwinds.
I know where my bet is.
OpenAI has branding power, a clear product focused mindset, focused attention, and moves faster than Google and Meta. Nearly 1 billion users in 3 years is no joke.
We're on Hacker News. Y Combinator literally teaches their companies that they can beat incumbents using focus and speed.
My bet is on OpenAI. When they IPO, I can easily see them with $1 trillion in valuation and raise the a record amount of money in an IPO.
If Meta and Google don't see OpenAI and LLMs as an existential threat, they wouldn't invest so much. I think AI has that potential to completely disrupt Google and Meta because it fundamentally changes the way people behave. It's a paradigm shift. It isn't just playing the same game.
You’re betting on OpenAI a company that has literally never sold an online ad against the two kings of online advertising…cool.
It took tiktok just 5 years to go from ~no revenue to ~$12b annual in the US.
ChatGPT has roughly the same MAU as tiktok. I don't see why their ad business wouldn't meet or exceed what tiktok was able to do in less than 5 years.
4 replies →
I agree with the gist of your statement but the same could've been said for a number of new companies against entrenched players. Heck, I'm pretty sure google was that company against the existing search.
You'll probably argue that this time it's different but no one knows what's different until it's already changed.
1 reply →
> OpenAI has branding power
With consumers right now? Sure, but so does WhatsApp and IG, both Meta properties. Meta and Google also WAY better brand power with advertisers. So there's that.
> I can easily see them with $1 trillion in valuation and raise the a record amount of money in an IPO.
They have agreements of roughly in $1.5T infra spend (and that doesn't include their own S&M and R&D spend) for the next 5 years. They have to have a combined amount of cashflow to cover that $1.5T (mix of income, debt financing, and stock financing) + all their other spending. The CFO admitted that they may need to bail out data centers to cover this to stay solvent in the long run.
> Y Combinator literally teaches their companies that they can beat incumbents using focus and speed.
YC is literally not God when it comes to advice, so this point is moot. Meta and Google didn't come out of YC and yet still beat incumbents.
With AI. OpenAI/ChatGPT is synonymous with AI. People say "ask ChatGPT" the same way people say "Google it".
I'm sure their $1.5t infrastructure commitments are based on hitting certain goals. Their comment about government support for data center is isn't a call for a bailout and taken out of context/exaggerated by mass media.
Yes but Google also beat the incumbents in Yahoo, AOL. People thought no way back in 2000 as well. Heck, Google wanted to sell itself to Yahoo.
If OpenAI (or any LLM chat product) starts shoving ads into people's faces, they aren't going to have a billion users anymore.
I already get mad at gemini when it shoves a barely-related youtube link into the chat with someone's huge stupid youtube face on it. Its a major reason why I rarely use it.
People didn't leave when Instagram, FB, Google showed them ads on a daily basis.
Hacker News commentator !== normal people.
As devil's advocate, the innovator's dilemma is said to be the cause of incumbent disruption. But in the case of AI, we're seeing incumbents over-correct into rapid AI adoption. It is a mess, but I wouldn't then use history to predict who the winners will be.
I don't know if they over correct. I think they rightly see the existential impact of AI on their existent business.
I think people who say they're only doing this for investors is wrong. I think management at Google and Meta truly think they're f'ed if they don't get AI right.
They have to compete against entities who really know what they are doing.
What about money spent on software developers, other knowledge work… AI could take 10-20% of spending on humans… that would be a few trillion per year.