We probably will end up with AI using search engines to tell us what we are looking for. The process of using your eyes to look at search results is something a

We probably will end up with AI using search engines to tell us what we are looking for. The process of using your eyes to look at search results is something an AI can do for us. That's different from a pre-trained model that already knows what you are looking for and doesn't need to go out and look for it. That's what chat-gpt is today. All Google needs to do is upgrade the frontend to their search engine to be a conversational AI that knows how to query their search engine effectively. Of course easier said then done and Google's inability to implement big technical changes to their core products is becoming a problem. This should be in their comfort zone; but it isn't. It's not about what they can do but about what they are willing to do. And they seem hesitant to commit to doing anything, often withdrawing products before they've even had a chance to fail properly. The big hurdles here are going to be cost and controversy. Doing this stuff at scale is going to be expensive. That comes straight out of their profit margin. And inevitably there are going to be controversies with AI getting it wrong, being biased, offensive, etc. Also not great for profits. Google is the wrong company to be doing this. That's why OpenAI is going where Google would be well capable of going without actually going there. One of those two companies is going to get billions of revenue in paid accounts next year for AI services. The other one is going to be obsessing about their add revenue. I expect a few more startups might jump in this space once the money start flowing.