Yuxi on the Wired - The Perceptron Controversy
Connectionism died in the 60s from technical limits to scaling, then resurrected in the 80s after backprop allowed scaling. The Minsky–Papert anti-scaling hypothesis explained, psychoanalyzed, and buried.
The Perceptron Controversy – Yuxi on the Wired Abstract During the 1950s and 1960s, the connectionist and symbolic schools of artificial intelligence competed for researcher allegiance and funding, with the symbolic school winning by 1970. This “first neural network winter” lasted until the rise of the connectionist school in the 1980s. It is often stated that neural networks were killed off by the 1969 publication of Perceptrons by Marvin Minsky and Seymour Papert. This story is wrong on multiple accounts: Minsky and Papert had been working towards killing off neural networks since around 196
Explore this link on the map →related reading
- Perceptrons (book) - Wikipediaen.wikipedia.org
- Notes for a talk on the history of neural networks – Yuxi on the Wiredyuxi.ml
- Marvin Minsky - Wikipediaen.wikipedia.org
- Connectionism - Wikipediaen.wikipedia.org
- Neural networks and deep learningneuralnetworksanddeeplearning.com
- On Really Trying · Gwern.netgwern.net
- Why the Godfather of A.I. Fears What He’s Built | The New Yorkernewyorker.com
- Zoom In: An Introduction to Circuitsdistill.pub
- On neural scaling and the quanta hypothesisericjmichaud.com
- Richard Feynman and The Connection Machine — LONG NOW IDEASlongnow.org
- An Abbreviated Personal Historyweb.stanford.edu
- 2023 letter | Zhengdongzhengdongwang.com