Chapter 1 · The dawn of AI (1943–1990)1943
The first artificial neuron
McCulloch and Pitts showed a simple on/off unit that sums inputs can compute logic.
Chapter 1 · The dawn of AI (1943–1990)1943
McCulloch and Pitts showed a simple on/off unit that sums inputs can compute logic.
Inputs 1 + 0 = 1. Threshold is 2. Not reached, so it stays quiet.
Each input is either on (1) or off (0). The neuron adds up what arrives. If the total reaches the threshold, it fires; otherwise it stays quiet.
In 1943 the scientists Warren McCulloch and Walter Pitts described a very simplified model of a nerve cell. It is either firing or not, like a switch.
Set two weights and a threshold. The truth table shows when your neuron fires. Can you make each gate? A simplified teaching version.
| A | B | Sum | Fires? |
|---|---|---|---|
| 0 | 0 | 0 | 0 |
| 0 | 1 | 1 | 0 |
| 1 | 0 | 1 | 0 |
| 1 | 1 | 2 | 1 |
Your neuron behaves like: AND
Now flip the inputs yourself:
Tip: can you make “one or the other, but not both”? Try as hard as you like. A later lesson explains why a single neuron cannot.
When does a threshold neuron fire?
Two inputs have weight 1 and the threshold is 2. Which gate is this?
What does a negative weight do?