Thanks for point this out. The previous Twitter was misleading,...

Thanks for point this out.
The previous Twitter was misleading, author was referring to the recent article on Supercomputer but saying stuff around SSL.

Regards to Mark's post on SSL, he is actually referring to data2vec from Meta Ai which is a new learning mechanism to unify CV speech and text into one more generalised AI modal.

This is what the team from Meta AI said:
"We also hope data2vec will bring us closer to a world where computers need very little labeled data in order to accomplish tasks. Since it is difficult and sometimes impossible to collect annotated examples — to train speech recognition models for thousands of languages, for example — data2vec is an important step toward more general AI"

SSL will eventually replace human labeled data, but the question is when will that happen, in the next 5 years or 10 years?
If SSL fully matures and AI can learn everything by itself, surely companies like Appen will be no more so will millions of other jobs/companies that can be easily replaced by AI/machines/bots.

data2vec is a step close but definitely not hyped like what the media is saying, if you read the research paper on data2vec a lot of the methods used were already built and pre existing.