WEBVTT

00:00.000 --> 00:12.560
The Neuron analogy

00:12.560 --> 00:15.560
You just recognized a friend in a crowd, instantly.

00:15.560 --> 00:21.520
Your brain did something extraordinary, and you didn't even think about it.

00:21.520 --> 00:25.960
Your brain has about 86 billion of these tiny cells called neurons.

00:25.960 --> 00:32.120
Each one gather signals, adds them up, and passes a signal on, but only if the total is strong enough.

00:32.120 --> 00:36.040
Engineers were inspired by this, they built an artificial neuron.

00:36.040 --> 00:39.720
It looks similar, inputs come in, and output goes out.

00:39.720 --> 00:47.280
But inside, there's no chemistry, just multiplication and addition.

00:47.280 --> 00:48.520
Give it numbers.

00:48.520 --> 00:53.920
It multiplies each by a weight, adds them, and sends one number forward to the next neuron.

00:53.920 --> 00:57.440
It's the whole signal and response.

00:57.440 --> 01:04.480
One neuron is simple, but connect many in layers, and they can do what your brain did at the start, recognize a face.

01:04.480 --> 01:10.200
The UAE's AI labs run on exactly this idea.

01:10.200 --> 01:18.400
This whole unit, you'll open the black box, weights, bias, layers, and run real code the kind UAE engineers use.

01:18.400 --> 01:19.400
Let's begin.