Hacker News new | ask | show | jobs
by jayant_kaushik 1267 days ago
What the near future would look like is- 1. Google: will implement some form of LLM on their engine and the contextual answers will significantly improve(~close to what ChatGPT produces) satisfying most people. But also displays the top links that contributed to the context. This is difficult to do but somehow they gotta figure out or risk their entire core business getting affected.

2. ChatGPT and other LLMs: will take some more time to meaningfully compete with google(as in becoming a noun) and in the mean time they would provide APIs to their models so that large orgs(businesses, scientists etc) can plug it in with their own data and use. They may or may not charge a fee for it.

In both cases there is one big pro and con: Pro: Death of SEO and ads as we know it Con: The model dictates what majority of internet users read, change their views or manipulate. And eventually becomes an AI weapon.