Here’s a question for ChatGPT: why should Big Tech control conversational search assistants?
Microsoft and Google’s race for control over the market for large language model (LLM) conversational search assistants reveals more than just evidence of one company that has stolen a march on another, which is now running around like a headless chicken.
First of all, as impressive as the “new Bing” launched by Microsoft may seem, it was not designed to do this. Bizarre errors, ethical dilemmas, unfathomable secrets and all sorts of incidents suggest that ChatGPT is simply a technology designed to “look human”, to be able to hold a conversation based on a pre-trained algorithm sourced from a huge database, but not a resource designed to generate quality answers, tell the truth or provide reasonable answers. ChatGPT can provide some adequate answers and pass some exams and tests, but that doesn’t count for much when, in practice, we are entrusting ourselves to a mathematical model — not to “intelligence” as such — that assembles and writes documents from certain sources and that, unfortunately, many people accept as gospel.
The problem with an arms race to incorporate LLMs into search engines, office automation or a thousand other things is that we are starting from the wrong premise: because of OpenAI’s unbridled ambition, it’s now “official” that we have a…
