It's probably no secret to those who do a lot of space-focused work, but when it comes to generative AI it's quickly becoming apparent that how a user interacts with generative models and systems is at least as important as training. and the underlying inference. technology. The last example comes across OpenAI Chat GPT, which was released as a free research preview for anyone to try this week.
Chat GPT is basically an implementation of their new GPT-3.5 natural language generation technology, but implemented in such a way that you just chat with it in a web browser as if you were slacking off with a colleague or interacting with a customer support agent on a website.
OpenAI it's already made waves with its DALL-E imaging technology, and its GPT series has turned heads with each successive release (and occasional existential dread from writers). But the latest chat-style iteration has apparently broadened its appeal and audience.
The examples so far in the wild seem to show that it's actually getting better at the term paper thing, but that it still has some work to do when it comes to avoiding some typical AI chatbot pitfalls, including presenting information. wrong as a fact. But obviously his commitment is through the roof, and people seem to be much more impressed with Chat GPT than with GPT-3.
The potential of something like Chat GPT of OpenAI to eventually supplant a search engine like Google is not a new idea, but this delivery of the underlying technology of OpenAI it's the closest approximation yet to how it would actually work on a fully fledged system, and you should fear Google.
Chat GPT it's also the best expression yet of something startups and space-watching entrepreneurs should already know: The gold rush in generative AI will be driven by the development of novel and defensible businesses built around how it shows up.
