Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

The concept of a moat for facebook (via network effects) and google (via scale, habits and learning effects) has worked well when it comes to printing cash.

Moats don't last forever, doesn't mean they aren't real.

The guy writing the post was writing about AI research at google. Not generally at Google, or for search.



> via scale, habits and learning effects

Then OpenAI/Google still has the moat for LLMs, for being most reliable, updated and trustworthy.

Facebook example made sense in pre short video era where connections meant something personal.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: