The sensible path toward human level AI

Inspired by the brain: we present Braifun, an innovative machine learning tool.

PHP: A New Brain Puzzle Game

How many trials you need to predict a hidden pattern? In this puzzle you need to predict a grid of 16 hidden true(T) and false(F) squares. Initially all set to F. To check your score, press the Check Score button. Change the squares and Check your score again. Do that repeatedly until you reach the score of 16 (which means you have predicted all the answers). You shouldn’t use the Reveal Answer button because this is not a memory test. The optimal solution should reach the perfect score in minimal number of trials and this is very relevant to how our brain works. Learn more:

Checkout Our New Video

Artificial Particle Life simulator is a simple fun program to build complex self organizing patterns from very simple rules similar to Conway’s game of life. We built this project to study complexity, chaos, and also to learn more about the human Brain. Our brain is trying to model these complex phenomena by making relations across its neurons. Many of these emerging complex patterns are computationally irreducible making them very good candidates to benchmark Artificial Intelligence algorithms. Unlike text and image data, these patterns can provide unlimited variations of meaningful shapes & semi-predictable behaviors

What Can We Do To Make Life better?

DNA-based adaptation is slow, costly, and inefficient. On the other hand, the brain’s intellectual approach to learn and generalize is fast and efficient but it is still prone to be stuck on local maxima due to our limited life span. The intellectual process operated by our brain is still largely a puzzle. Solving this puzzle can help us to find the path toward the Global Maximum and solve many problems we currently suffer from. To learn more about our view, watch this video:

Why Choosing Braifun?

If you are looking for an easy and powerful tool to develop intelligent systems, Braifun is for you. Deep Learning tools like Tensorflow and Pytorch are data and power inefficient because they rely on backpropagation which is not optimal for real time learning. We believe human level intelligence cannot be attained unless we first achieve data efficiency and real time learning. Think about the difference between Web 1.0 and Web 2.0. If other Machine Learning tools are ML 1.0, Braifun is ML 2.0. Braifun uses an innovative ML algorithm we call it PHUN

Language Modeling

Unsupervised training on textual data.


Learns and detects shapes few shots.


Learns and detects audio and other complex time-series signals.

Reinforcement Learning

Efficient learning from feedback and self-supervision

What is PHUN?

PHUN (pronounced as Fun) stands for Predictive Hebbian Unified Neurons. PHUN builds up on well-known terms and concepts rooted in brain research and neuroscience studies. It employs a technique similar to Hebbian and spike timing dependent plasticity for its learning. PHUN’s advantages are:



ML Algorithms

One-shot LearningIterative Learning
Online LearningOffline Learning
Computationally EfficientComputationally Inefficient
Biologically PlausibleBiologically Less Plausible

ML Tools

BraifunPytorch / TensorFlow
Easy to LearnSteep Learning Curve
Uses PHUNUses Backpropagation
runs on CPUruns on CPU/GPU
Coming Soon!Available

Sign up to be notified once Braifun released