Right now, chatgpt is trained on a broad spectrum of slightly dated information, so hopefully the biases even out. And we know it hallucinates etc. In the fut

Right now, chatgpt is trained on a broad spectrum of slightly dated information, so hopefully the biases even out. And we know it hallucinates etc. In the future I see two big risks with the tech, approach and ownership: 1) short term, SEO is going to move into trying to influence the LLMs to push products unwittingly. Companies will be getting great offers from shady SEO outfits promising to make their products the default answers for broad questions? 2) mid term, now that all sense and pretence of safety and greater good are gone from openai, openai will be working out how to insert and see ads in the results? There will probably even be product placement in the results coming from the paid-for in-ms-office version.