[rorcodes]

I'm wondering if, with the recent influx of GPT-3 account approvals and now the rise in competition for API based ML, there will be a slow trend away from learning ML as people start productionising with GPT-3 calls. There are negatives and positives towards this (ie - black box nature of the API / inconsistency in results as it trains making it non testable etc).

As someone with a masters in machine learning I look forward to watching this all unfold, very interested in where this part of the engineering industry is headed.