Binary Neurons Network (BNN)

This is an attempt to create AI but not AI in the modern sense of the word.
It is AI in the original meaning coinciding with the meanings of the following synonyms:

• Artificial general intelligence (AGI);
• Strong artificial intelligence (Strong AI);
• Human-level artificial intelligence (HLAI);
• True artificial intelligence (True AI).

Looking for a sponsor

Usage

#include <iostream>
#include <thread>

#include "common/architecture.h"

int main()
{
    constexpr bnn_settings bs
    {
        .quantity_of_neurons_in_power_of_two = 12, // 2^12=4096
        .input_length = 31,
        .output_length = 8,
        .motor_binaries_per_motor = 8,
        .random_size_in_power_of_two = 22,
        .quantity_of_threads_in_power_of_two = 1, // 2^1=2
    };

    bnn::architecture bnn(bs);
    bnn.initialize();
    bnn.start();
    do {} while(bnn.get_state() != bnn_state::started);
    bool stop{false};

    auto manager_thread = std::thread([&stop]()
    {
        std::this_thread::sleep_for(std::chrono::seconds(2));
        stop = true;
        std::this_thread::sleep_for(std::chrono::seconds(1));
    });

    while(!stop)
    {
        static char input[bs.input_length + 1]{};
        static char output[bs.output_length + 1]{};
        static bool value;

        for (u_word i = 0; i < bs.input_length; i++)
        {
            value = rand() % 2;

            // Put data in BNN
            bnn.set_input(i, value);

            input[i] = value + 48;
        }

        for (u_word i = 0; i < bs.output_length; i++)
        {
            // Get data from BNN
            value = bnn.get_output(i);

            output[i] = value + 48;
        }

        std::cout << "input=" << input << " output=" << output << std::endl;
        std::this_thread::sleep_for(std::chrono::milliseconds(100));
    }

    bnn.stop();
    manager_thread.join();

    return 0;
}

Build

CMake

Example projects for BNN

• ./src/examples/minimal/ - Contains minimal project
• ./src/examples/web/ - (support is suspended) The input to the brain is a picture from the browser, and the output is the movement of the gaze of the brain. QT Creator is required
• https://github.com/cortl0/device - The practical implementation of a physical device with the binary neurons
• https://github.com/cortl0/device_3d - Testing the BNN in 3d world

Author

Ilya Shishkin
e-mail: cortl@yandex.ru

Achievements

• The first mention of BNN in the literature:
  http://gcrinstitute.org
  http://gcrinstitute.org/papers/055_agi-2020.pdf
  (see page 52)

• BNN was placed in the Arctic Code Vault
  https://archiveprogram.github.com/arctic-vault

Links

• http://8iter.ru/ai.html
• https://t.me/artificial_intelligence_bnn
• https://t.me/artificial_intelligence_bnn_grp
• https://www.youtube.com/watch?v=z-TKgo2b8mg&t
• http://www.gotai.net/forum/default.aspx?threadid=192413
• http://www.gotai.net/forum/default.aspx?threadid=276702
• https://www.cyberforum.ru/ai/thread1834551.html
• https://agi.place/?ai=9&id=24
• https://agi.place/?ai=9&id=26
• https://agi.place/?ai=9&id=34
• https://agi.place/?ai=9&id=48
• https://github.com/cortl0/binary_neurons_network

License

This project is licensed under the GPL v3.0 - see the LICENSE file for details