justaguy1212

Yeah this is a bizarre statement considering that GPT-3 is definitely not any of those things really. GPT-3 is far, far too computationally expensive to have value in industry. A linear CRF is more useful than most NN approaches in industry right now, just simply because in many circumstances you want to have something that you can apply to a few billion documents and get the result within a few hours, then tweak a few things and repeat if you like. These simple models also have the ability to be predictable as well. Some transformer or lstm methods can be useful in industry, but it really depends on the application. I certainly would not be using GPT-like systems for much in industry, other than gimmicks for marketing. GPT-3 is useful for academia - not industry.

Stackexchange sites are far better than subreddit if you're looking for more reasonable and constructive discussion.

I actually don't see this from his data. It seems more like he started with a hypothesis, wrote the article as if he was correct, and then gathered data to try to support it. Then after seeing the data wasn't as commensurate with his idea, he wrote around the inconveniences.