Hacker News Reader

Comment by jcfrei | Hacker News Reader

Google's core business model and that of the new AI companies (ie. ChatGPT) are diametrically opposed to each other: Google's most valuable business is having corporations bid on keywords such that their products/services will appear very high up when someone uses the search bar. The business model of the new AI companies is to have users pay a monthly fee for getting very good answers to questions about products/services or just general queries. Right now Google somehow tries to shoehorn their existing business model into these new chat based interfaces. This is obviously not going to work out because either you respond with shitty answers (with high bids) or disregard the bids but then you lack monetization. We're probably going to see a two tier market here (just as we do with Android/iOS): Free AI bots that give shitty replies influenced by bids/ads and premium ones that require a subscription.

View on HN

Or, the core business model of LLM products today is diametrically opposed to the inevitable core business model of LLM products 5 years from now.

The business model today is very good answers to questions, but eventually corporations will bid on keywords such that their products/services will appear in LLM output.

Just like streaming services used to be cheap and didn't have ads when they were new, but today they're not cheap and do have ads. There's no excuse for us to fall for this again.

On a lot of things "good answers" include personal value judgments - having a product that directly delivers those (unlike traditional search) is not trivial and needs noticeable personalization which is comparatively expensive. Not clear if that model is cost efficient.

I am completely bewildered why google hasn't started looking into a "Google Prime" where you get access to tons of google services, while also putting your privacy first.

Why? Their business is ads, not services for users

Adapt or die

Speaking of incompatible business models, I'm still not clear how AI/LLM driven search is supposed to be sustainable when it means that most content is consumed through a middleman without the original publisher seeing a cent of revenue from it. If the AI companies manage to get the precedent that training is transformative enough to be considered fair use to stick then not even paywalled outlets are safe, someone can buy a single subscription, throw all their content into the LLM blender, and sell it on as their own product. Hell if they're feeling bold then they can just pirate the training material, as we saw with that enormous archive of pirated ebooks which is being passed around as an "AI research dataset".

Like everyone out there farming the Amazon, they don't care whether it's sustainable or not. I suppose this is how we end up with Butlerian Jihad moments.

there are two types of people. 1) who trusts everything on the internet 2) who doesn't trust everything on the internet.

whether its free or premium, the second group still wouldn't trust the search or ai results. they apply their own due diligence. sadly it makes a tiny percentage of the world people now. we should strive to increase this number and in next decade or so, mostly it happens.

Reply on news.ycombinator.com