4 ms·

I've worked on such projects, too. They were carefully selected to have minimum impact when the models are wrong. In some of these cases, as you say, the KPI w

I've worked on such projects, too. They were carefully selected to have minimum impact when the models are wrong. In some of these cases, as you say, the KPI was higher, e.g. in classification tasks (where previously a clerk was supposed to label some cases). An LLM produced decent results most of the time - and didn't get tired. But when I looked deeper, it turned out the folks who decided to use LLMs could have used classical ML learning methods as well with similar efficiency. I believe we arrived to the point where people just got lazy and just use LLM for anything.

Arrowmaster 1y ago
This absolutely seems to be the way things are going. We keep hearing about how 'AI' models have done these amazing things and learn they were specially trained models. Most news reports don't even get into if it was an LLM or ML. Then business leaders think they can just pay for a subscription to an all encompassing LLM model and get the same results. It's just an evolution of the long standing tradition of tech being pushed from the top down because the sales guys impressed the C-suites.