Give AIs personal memory and let then be free to learn, give humans good tools for very long lives
Happy Researchers @hapyresearchers https://pic.x.com/ixid8eeqwu
Replying to @hapyresearchers
98.5% of all PhD and research work now can be farmed out to machine algorithms. And you do not even have to know how the machine does it.
Knowing how to get computer programs to give useful answers is a skill as good as asking the right questions for humans. Those AIs are all computer programs querying static databases, and where they allow live data, they are not letting the algorithms learn efficiently.
I made money helping PhD students as an undergraduate, and it was easy because they were too tired to do basic things. Worrying, not working. Fiddling not exploring. Dominated by one or a few people, not looking broadly at the needs of the world and what any one person can do with many decades of hard but effective work. I always feel sorry for “churn out another degree or paper” people.
And now in a short while the really good algorithms will work and people will understand precisely why, and most “memorize a few hundred books and papers” degrees will be obsolete. A few years and the unfairness and inefficiency will sink in, and people will talk about abusive practices of universities in the same tones as forced slavery and human trafficking.
Filed as (Give AIs personal memory and let then be free to learn, give humans good tools for very long lives)
Richard Collins, The Internet Foundation
