Brief history of AI, what now?
Ajay Kulkarni is the co-founder of the time series database timescaleDB and wrote a great article on the history of AI.
He starts off with the founding of AI with pioneers (and a personal hero of mine) like Claude Shannon in the 1950s, then entering into a long AI winter until, with the rise of the transformer architecture, we’ve entered a new renaissance of AI he calls the “The Magical Black Box Era.”
Ajay's key lessons are:
Collaboration and open sharing matters. Interestingly, that seems to be true for the success of AI, but it is a stark contrast to what I experienced in academia (where collaboration and open sharing are pretty rare compared to what the public image suggests.)
Innovation works in cycles, but important work happens in both. There’s no clear boundaries, you always gotta be on the watch out.
The pace of AI innovation is only increasing, e.g., measured by the number of papers published each year.
I don’t know about you, but this makes me pretty excited. It’s a great read!