Introduction to AI/ML

Hype

Bold predictions, two AI winters, and the hype cycle we are in now.

Hyping AI isn't new

1957 → 1967
1957
There are now in the world machines that think, that learn, and that create.— Herbert Simon
1958
Within ten years a digital computer will discover and prove an important new mathematical theorem.— Allen Newell
1961
Within our lifetime machines may surpass us in general intelligence.— Marvin Minsky
1965
Machines will be capable, within twenty years, of doing any work a man can do.— Herbert Simon
1967
Within a generation… few compartments of intellect will remain outside the machine's realm; the problem of creating "artificial intelligence" will be substantially solved.— Marvin Minsky

AI winter

1973: the Lighthill report, Artificial Intelligence: A General Survey; 2024: Goldman Sachs, Gen AI: too much spend, too little benefit?
The same doubts, 51 years apart: the Lighthill report (1973) and Goldman Sachs (2024)
MIT report: 95% of generative AI pilots at companies are failing
Paper: The Illusion of Thinking, on the limits of reasoning models

The third hype cycle

investment & research = f(expectations, results)

The third hype cycle: the original chart

Chart of AI investment and research over time: Dartmouth 1956, inflated hype in the 1970s, AI winter 1974 to 1980, new hopes in the 1980s, AI winter 1987 to 1993, data explosion 2007, concerns 2020, explosive growth