Ok, I’ll call it: We are closing this chapter of the AI wars.
Since January or so, tech leaders, their customers and reporters have been haggling over advancements in large language models. We’ve been consumed by models, their names and which 0.5 version of a model is better than another.
I don’t think any of that is going to matter next year. My takeaway from the last few weeks—anchored by major releases from Google—is that all these models are converging very, very quickly. Yes, we can debate this or that benchmark, and researchers will continue to boast about outperforming their rivals in one way or another.
0 Commentaires