4 ms·
The problem with AI such as GPT4 is training data. As it usage increases, most of its training data will be data generated by GPT4, hence creating a positive fe
by streetcat1 4y ago
The problem with AI such as GPT4 is training data. As it usage increases, most of its training data will be data generated by GPT4, hence creating a positive feedback loop. This would actually increase the value of human knowledge workers.
- anonzzzies 4y agoNah; the model will improve to learn faster and more from less data. Or AI winter.
- daveguy 4y agoI'm betting on an AI winter before a fundamental change in the difficulty of computation.
- anonzzzies 4y agoYep, that is a likely outcome.
- waboremo 4y agoEntire farms of people dedicated to providing accurate training data. Like battery farms from the matrix but more boring since you'll need a masters degree to become a happy little data provider. Joking, but training data isn't as big of a problem as you suggest. Since so much about AI is still very centralized it's relatively low effort for them to tag generations in a way that makes sense, avoiding a lot of effects of a loop.
- meh8881 4y agoIt’s ok to train a model on model outputs so long as you have someone to curate them.