What are Neural Networks || How AIs think

  • Published: 21 May 2018
  • Big thanks to Brilliant.org for supporting this channel check them out at https://www.brilliant.org/CodeBullet

    check out Brandon Rohrers video here: https://www.youtube.com/watch?v=ILsA4nyG7I0&t=638s

    Become a patreon to support my future content as well as sneak peaks of whats to come.

    Check out my Discord server

Comments • 368

  • Hi there!
    Hi there!  3 weeks back

    6:25 smooth transition there

    • anton broos
      anton broos  3 weeks back

      So Ben is a XOR-gate??

      • IceCubeGold ll Diamond PB2

        Multiple Neurons Of Code Bullet Learn How To Explain How They Work.

        • Illusion
          Illusion  1 months back

          I'm having an AI test in 6 hours
          thank you Code Bullet

          • Spencer Jessee
            Spencer Jessee  2 months back

            I would greatly appreciate the followup video you mentioned about the connection of genetic weight evolution with neural networks

            • TJ.man. Alex
              TJ.man. Alex  2 months back

              Is it just me or is this just crazy complicated Boolean logic

              • Blue 64
                Blue 64  2 months back

                @Code Bullet
                You could use an Info Card to take people to the Video/Channel you mentioned at 5:00

                • elcidbob
                  elcidbob  2 months back

                  Seems like it would be more effective, albeit more complicated to implement how with neurons there's not two states, there's three. Neurons always fire. They have what's known as a resting rate which is just the rate it fires with no external influence. When inputs stimulate the neuron, it fires at a faster rate. When an input inhibits a neuron, it fires at a slower rate. When there's no input or the inhibitory input equals the stimulating input, there's no change.

                  • Jay T
                    Jay T  2 months back

                    at 6:48 why did you multiply bias node by -1?

                    • Nicholas Nelson
                      Nicholas Nelson  2 months back

                      Jay T
                      The bias node is connected with a blue line. If you pay close attention, he drags his mouse over the blue line as he says it’s converted into a minus 1

                  • Christopher Dibbs
                    Christopher Dibbs  2 months back

                    Neuroscientists are just brains trying to figure themselves out...

                    • Adam Opalecky
                      Adam Opalecky  1 months back

                      Holy shet. This is too real for me :D

                  • Amethystic
                    Amethystic  2 months back

                    This was surprising easy to follow

                    • Aron Highgrove
                      Aron Highgrove  2 months back

                      That was a little chaotic, and therefore hard to follow.
                      The video you linked used a more consistent and logical color scheme: more intense / higher value = white, less intense / lower value = black. That's also how light works and more intuitive.
                      Not sure why you got so much views, the presentation wasn't really helpful.

                      • QuasiGame0
                        QuasiGame0  2 months back

                        So, in a way, aren't ai neural networks, more or less a large string of AND functions with counters that increase/decrease if the result was favourable/correct or not and then comparing the counters to determine in the AND function which should be chosen?

                        • QuasiGame0
                          QuasiGame0  2 months back

                          @Selsion Thanks for the response! I'm trying to recreate by using littlebigplanet's logic
                          Wasn't expecting on a year old video haha

                        • Selsion
                          Selsion  2 months back

                          For each neuron, rather than a question of "does the input have this feature", it's more of a question of "how well is this feature found in the input". So you might get values of 1.2 if it's detected, 0 if not, -2.6 if the opposite was found, etc. So it's not really a bunch of AND functions, but rather a linear combination of the inputs. eg if you have inputs x1, x2, x3, weights w1, w2, w3, and bias b, then you compute z = x1w1 + x2w2 + x3w3 + b. We then compute y = f(z) for some activation function f, like sigmoid. This activation function lets our network approximate non-linear functions.

                          Also, the usual way that we train networks isn't through increasing/decreasing counters. Instead, we define a loss function which describes how badly the network performed on the given data. Then, we use calculus to figure out how our loss function changes with respect to the weights. This will tell us how to change our weights to decrease the loss, so we update the weights a tiny bit given that info. We keep repeating that until we're satisfied with the network.

                      • Engr. Abidur Rahman Alvi
                        Engr. Abidur Rahman Alvi  2 months back

                        Nailed it bro! Beautiful

                        • TC TrainConstruct
                          TC TrainConstruct  3 months back

                          A neuron doesn't sum up the inputs then uses the activation function, but to the inputs it adds a "bias".
                          Output = H(b+Si)
                          where H is the activation function, b the bias and Si the sum of hte inputs.

                          • Syed Ali
                            Syed Ali  3 months back

                            3:35 Best part of the video

                            • Mateo Gonzalez Cano
                              Mateo Gonzalez Cano  4 months back

                              Which web u use to programm

                              • Venus Dandan
                                Venus Dandan  4 months back

                                I looked away for like 2 seconds and suddenly I didn't understand and had to rewind. I love the drawings <3

                                • Mynameisevanbro _
                                  Mynameisevanbro _  4 months back

                                  Is it just me or when I put 1,-1,-1,-1 into the input layer I get 1. If everyone has this issue. Then I found a flaw

                                  • Vastag Botond
                                    Vastag Botond  4 months back

                                    Very helpful:)

                                    • Jose BERGER
                                      Jose BERGER  5 months back

                                      WHERES THE PROGRAMME THAT ISNT MUCH OF A HELP!!!!

                                      • Planet Dilien
                                        Planet Dilien  5 months back

                                        He's more likely to be triggered
                                        Now I know what you mean!

                                        • ston219
                                          ston219  5 months back

                                          When u learn maths better and more in one Video then the whole school year

                                          • Amir
                                            Amir  5 months back

                                            Fantastic man. Your videos are great...

                                            • James Collins
                                              James Collins  5 months back

                                              Can you do an example for your actual projects, AI learning to play games rather than image recognition.

                                              • hehe boi
                                                hehe boi  5 months back

                                                I didnt understand an word.

                                                • 4372696D736F6E
                                                  4372696D736F6E  6 months back

                                                  I always thought that neural network genetic algorithm fusion is made by tensorflow lambda layer. I was wrong.

                                                  • David Eikeland
                                                    David Eikeland  6 months back

                                                    2:48 to 3:29 is incorrect, a neuron outputs the same value to all outputs...

                                                    • VanLifeCrisis
                                                      VanLifeCrisis  6 months back

                                                      *eyes glaze over*

                                                      • NathanGG
                                                        NathanGG  6 months back

                                                        So if the brain is made up of 100 billion neurons, does that mean that if we had computers powerful enough to simulate evolution with creatures of 100 billion neurons, they could eventually become as intelligent as us?

                                                        • Otaku-chan
                                                          Otaku-chan  2 months back

                                                          ​@Арднер Уса You're right, but on the other hand technology is always advancing, just as if you told people a hundred years ago that it's possible to send a human to that one ball in the sky that goes from crescent to gibbous and back again, or if you told them that a computer could be smaller than your hand and be used to hear someone's voice instead of being a gigantic machine that's as big as two rooms (and can only do basic arithmetic lol) they wouldn't believe you.

                                                          Quantum computers can already create states that are _neither 0 nor 1_ which opens the possibility for neuron contacts and probabilities. It's already possible to mimic the action of inhibitors and circles, and perhaps in less than a century making a "are you human?" captcha will not even exist because there'll be no problem AI cannot solve.

                                                          However if by intelligence you mean emotional intelligence...no one can say. Honestly it boggles my mind to think that someone someday can type code which enables a computer to feel emotions like genuine frustration or excitement which isn't pre-programmed. But it may just happen, who knows.

                                                        • Otaku-chan
                                                          Otaku-chan  2 months back

                                                          And that is why people think that Artificial Intelligence might someday develop far enough to rival human beings' smarts.

                                                        • Арднер Уса
                                                          Арднер Уса  4 months back

                                                          Real brain has much more variety in neiron contacts than just 0 and 1. There are ingibitors, circles, and more and more... So we not even close to create computers which could support it. Its much easy to brootforce rar with password than doing so

                                                      • dnabre
                                                        dnabre  6 months back

                                                        You make it sound like using genetic algorithms are the only/normal way to train a neural net. While they are used, they aren't the standard/normal approach to training. More structured learning models are generally used.

                                                        • Imicoz
                                                          Imicoz  6 months back

                                                          6:30 and onwards reminded me so much of tecmath's channel... Like holy shit xD

                                                          • Billiam _
                                                            Billiam _  7 months back

                                                            Hey CodeBullet. I’m an upcoming senior in HS and over the past year I’ve found a passion for coding. I’ve been trying to get into ai and ML for the past few months but with no luck. Could you go more into depth with this specific neural network?

                                                            • Cireng Enak
                                                              Cireng Enak  7 months back

                                                              Just use the dino "hack"

                                                              • Rocketjake1001
                                                                Rocketjake1001  7 months back

                                                                This video not only explains how Neural Networks are made but how Code Bullet was created into our mortal world.

                                                                • PsychoGoth
                                                                  PsychoGoth  7 months back

                                                                  No one cares about the SOMA? Really??? That’s one of the most important parts of a biological neural network... it’s basically a calculator which performs a number crunch on all of a neurones dendrite inputs coming in... kind of converting binary into a decimal value... It also determines if the message is strong enough to be used or should be ignored causing the dendrites to reset to accept new messages.

                                                                  • Ibrahim Abdalla
                                                                    Ibrahim Abdalla  8 months back

                                                                    I dare you to learn binary, I heard it's a pain in the ass

                                                                    • dudebobmac
                                                                      dudebobmac  8 months back

                                                                      I love this, and I'd love to see a video on how to actually combine NNs and the genetic algorithm! Keep up the awesome work :D

                                                                      • Indrajeet Javeri
                                                                        Indrajeet Javeri  8 months back

                                                                        5:00 bottom right corner.3B1B spotted

                                                                        • ElijahTommy 777
                                                                          ElijahTommy 777  8 months back

                                                                          So, if you have a neuron that you want to require 3 positive connections for it to activate, then you would make the bias -2 right?

                                                                        • Carter
                                                                          Carter  8 months back

                                                                          Neural networks are paper airplanes if you see the youtube thumbnail.

                                                                          • Paweł Gontowski
                                                                            Paweł Gontowski  8 months back

                                                                            Very helpfull ʕ•ᴥ•ʔ

                                                                            • John Dough
                                                                              John Dough  8 months back

                                                                              me, a biologist, hearing him explain biology...yeah thats about right

                                                                            • unicon
                                                                              unicon  8 months back

                                                                              Why do o fell like a perceptron will be a robot in the next fallout game

                                                                              • Cain XVII
                                                                                Cain XVII  8 months back

                                                                                You look like a pokemon

                                                                                • Jamman
                                                                                  Jamman  9 months back

                                                                                  6:32 what happened to the quillity

                                                                                  • powerline
                                                                                    powerline  9 months back

                                                                                    Let the AI learn to play superhexagon

                                                                                    • Kaatan Nogard
                                                                                      Kaatan Nogard  9 months back

                                                                                      So a neural network is basically a serie of gates that allow the machine to draw conclusions out of simpler facts