I can think of plenty of tasks that even the closed frontier models fail miserably at. From my perspective we have a long way to go for these models to get better.
replyI'm sure they exist, but it's not the bulk of the work.
replyMy prediction is that in in a few years, maybe more, the current models will be seen as so primitive that they're not worth mentioning outside of a historical context.
replyWhat a bold prediction /s
reply