1
00:00:04,450 --> 00:00:02,210
Heat current machine learning systems

2
00:00:06,710 --> 00:00:04,460
are mind-blowing and what they can do

3
00:00:08,509 --> 00:00:06,720
but they don't do it like biological

4
00:00:11,810 --> 00:00:08,519
organisms at all it's it's very

5
00:00:15,709 --> 00:00:11,820
different cells do not form change in

6
00:00:18,349 --> 00:00:15,719
which they wait uh sums of real numbers

7
00:00:20,450 --> 00:00:18,359
that there is something going on that is

8
00:00:22,010 --> 00:00:20,460
roughly similar to it but there's a

9
00:00:23,929 --> 00:00:22,020
self-organizing system that designs

10
00:00:25,609 --> 00:00:23,939
itself from the inside out not by a

11
00:00:27,650 --> 00:00:25,619
measure learning principle that applies

12
00:00:29,630 --> 00:00:27,660
to the outside and updates weights after

13
00:00:31,790 --> 00:00:29,640
reading and comparing them and Computing

14
00:00:34,190 --> 00:00:31,800
gradients to the system so this

15
00:00:36,950 --> 00:00:34,200
perspective of local self-organization

16
00:00:40,069 --> 00:00:36,960
by reinforcement agents that try to

17
00:00:41,389 --> 00:00:40,079
trade rewards with each other that is a