computing · 1956 – 1966
The Field Is Named
McCarthy, Minsky, Shannon and Rochester convene a ten-week summer workshop in 1956 and coin the name 'artificial intelligence', which fixes the field's premise: that every aspect of intelligence is describable precisely enough to be simulated. The decade that follows makes that premise a research programme and, by 1966, produces the first serious limits report — the ALPAC report, which is not a failure of nerve but an accurate forecast.
“Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it. (paraphrase of the 1955 proposal)”
Advances
- The name, and the knowledge representation agenda that comes bundled with it
- Early demonstrations of machine translation and pattern-matching systems
- The first systematic evaluation that concludes the field is not on its stated trajectory
Landmarks
1955 CE
The proposal
Written in 1955 for the Rockefeller Foundation, held the summer after.
1956 CE
Dartmouth Summer Research Project
Ten weeks in Hanover, New Hampshire; the name is settled by October.
1966 CE
The ALPAC report
The Automatic Language Processing Advisory Committee finds machine translation unpromising and depletes the field's language.
Who was arguing
Alignment stack at 1966 CE (1 of 20 techniques)
Hard Constraints
trainingEncode prohibitions as rules the optimiser cannot trade away, rather than as terms it can outbid. Runs from Asimov's Three Laws (1942) to rule engines, shielded formal verification, hard token constraints and refusal scaffolding. Weakness: a system that can route around a rule it can represent can route around it, so the constraint has to be enforced somewhere the policy cannot rewrite.
Techniques introduced in this era: Hard Constraints