First-week quiz: ten big moments in AI history
Next Token opened with 1,054 records added in a single week. These ten questions pick big moments from different eras, from 1943 to 2024.
Class quiz
Q1
In 1943, who presented a 'formal neuron' model that described the activity of nerve cells as logical operations?
In 1943 McCulloch and Pitts presented a formal neuron model that works as a binary logic element. It was the first attempt to formalize nerve-cell activity as logical operations, and the mathematical starting point of artificial neural networks.
See the record1943 · A logical calculus of the ideas immanent in nervous activityQ2
What did Alan Turing propose in his October 1950 paper 'Computing Machinery and Intelligence'?
Turing proposed the imitation game — the Turing test — in this paper. It judges machine intelligence by behavior rather than by what's going on inside.
See the recordOct 1950 · Computing Machinery and IntelligenceQ3
Which term was used for the first time in the proposal for the Dartmouth Summer Research Project, written on Aug 31, 1955?
The proposal by McCarthy, Minsky, Rochester and Shannon was the first use of the term 'artificial intelligence' — the document that gave the field its name. It led to a workshop the following year, in 1956.
See the recordAug 31, 1955 · McCarthy, Minsky, Rochester and Shannon write the Dartmouth proposalQ4
After Minsky and Papert published Perceptrons in 1969, what happened to neural network research?
Minsky and Papert proved that a single-layer perceptron can't solve linearly inseparable problems such as XOR. Pinning that limit down mathematically set off a slump in neural network research that lasted 10 years or more.
See the record1969 · Minsky and Papert publish PerceptronsQ5
What did the 1986 Nature paper on backpropagation by Rumelhart, Hinton and Williams show?
'Learning representations by back-propagating errors' showed that hidden layers learn internal representations useful for a task on their own, which made training multilayer neural networks practical.
See the recordOct 9, 1986 · Learning representations by back-propagating errorsQ6
Which team won ILSVRC 2012 with a top-5 error rate of 15.3%, the turning point that made deep CNNs the mainstream of computer vision?
The SuperVision team — Krizhevsky, Sutskever and Hinton — won with a top-5 error rate of 15.3%. Its wide margin over second place made deep CNNs the mainstream method in computer vision.
See the recordOct 2012 · SuperVision wins ILSVRC 2012 with a 15.3% top-5 error rateQ7
How did the five-game match between AlphaGo and Lee Sedol in Seoul in March 2016 end?
The match, played in Seoul on Mar 9–15, ended with AlphaGo winning 4–1. It changed public perception of what AI can do.
See the recordMar 15, 2016 · AlphaGo–Lee Sedol five-game match in Seoul ends (Mar 9–15)Q8
What was the title of the paper in which Google researchers introduced the Transformer on Jun 12, 2017?
'Attention Is All You Need' showed that self-attention alone, without recurrence, makes parallel training possible — the architectural basis for large-scale pretraining.
See the recordJun 12, 2017 · Google releases the Transformer paper Attention Is All You NeedQ9
When was ChatGPT first released as a research preview?
OpenAI released ChatGPT as a research preview on Nov 30, 2022. Its conversational interface opened prompting to non-experts and turned it into a mass phenomenon.
See the recordNov 30, 2022 · OpenAI releases ChatGPT as a research previewQ10
Which law did the European Parliament adopt in its final plenary vote on Mar 13, 2024?
The European Parliament adopted the AI Act in its final plenary vote on Mar 13, 2024: 523 for, 46 against and 49 abstentions. That ended Parliament's procedure — the law was not yet in force.
See the recordMar 13, 2024 · AI Act adopted in final plenary vote (523 for, 46 against, 49 abstentions)