ABCDEFGHIJKLMNOPQRSTU
1
Step1 🗂️2 📑3 🤔 4 🗣️5❓🤔6 🗣️
2
CommentaryWill the ycombinator startup be successful?The AI is a computer. Computer's speak ones and zeroes. The AI perceives in binary true/false statements. We represent that with a YES/NO perception. The AI can perceive that data and come to meaningful conclusions.Machine Readable Input Data - The AI can perceive in 1's and 0's. The human redable terms must be translated into a one or a zero. Then the AI can read the information and provide it's response.Matrix with rows matching input data features. Three features = 3 rows. Number of columns is memory capacity. Must have enough columns to maintain memory. It can be too dense and take too much compute and not be efficient. I can be not dense enough and not have the capacity required to be successful. Fewer than 2x input features is okay.The last layer needs to match the training data output dimentions with the density of the previos layer. 4 columns (from previous layer) x 1 output label. 4 x 1Test what the AI Matrix knows using the first input layer and a TANH function.Test what the AI Matrix knows using the second layer (output) without a non-linear function.
3
Simplified TerminologyYes/No Perceptions: AI requires true/false binary format to conclude an answer.Questions and Answers used for Teaching the AIAI Memory MatrixAI Response MatrixAI Rationalization / Thoughts about the QuestionAI Answers
4
5
ML TerminologyFeatures (unvectorized)Vectorized FeaturesHidden Layer WeightsOutput Layer WeightsHidden Layer ResultsOutput Layer Results
6
using AWS3+ team size
50%+ Rev Growth
onetwothree
7
Startup AYESYESYES1110.510.000.940.980.070.720.860.940.992.14
8
Startup BYESNONO1000.060.890.600.890.870.470.000.730.751.07
9
Startup CYESYESNO1100.340.380.160.600.470.510.710.910.951.96
10
Startup DYESNOYES1010.920.690.370.800.921.59
11
12
YouTube: https://youtu.be/xY4iKQ8TfiQ
13
14