Brain Simulator II v1.0 Release Adds Features to Speed Artificial General Intelligence Development

Share Article

Open-source software platform designed for proving how AGI, the next phase of AI, will emerge.

Charles J. Simon, creator of the Brain Simulator II.

These new features, coupled with continued experimentation into varied AI algorithms, are eliminating the limitations of computer power and accelerating the evolution of AI into AGI, with true AGI expected by the end of this decade.

On the heels of publishing Brain Simulator II: The Guide for Creating Artificial General Intelligence, Artificial Intelligence (AI) expert Charles Simon has launched a new version of Brain Simulator II, the free, open-source software platform for proving how Artificial General Intelligence (AGI) will emerge.

The point of the Brain Simulator II is to understand that we don’t know precisely how to create AGI, that human intelligence is the best AGI model we have, and that today’s narrow AI approaches cannot gradually expand to AGI. To that end, the Brain Simulator creates a biologically plausible spiking neural platform and experimental research toolkit for trying out alternative approaches to today’s AI.

Brain Simulator II creates an end-to-end AGI system with modules for vision, hearing, robotic control, learning, internal modeling, planning, imagination, and forethought – in short, the ability to understand, learn, and respond to any intellectual task a human can perform. Today, the system functions with a small number of objects and words, but these basics will be expanded in the future to provide a more general solution.

Features in the new version of Brain Simulator II include:

  • A huge performance increase – Brain Simulator’s new Neuron Engine can process up to 2.5 billion synapses per second on a multicore desktop computer with neuron and synapse counts limited only by available RAM;
  • The ability to split a network across multiple servers – For even larger networks, the new Neuron Server can run in parallel on any number of networked computers, with performance estimates indicating that full neocortex emulation could consist of as few as 160 servers;
  • A quantum User Interface improvement – With only a small fraction of human DNA contributing to the brain’s thinking neocortex, it must consist of repeating simple neuron clusters and generic rules for connecting them so the upgraded User Interface makes experimenting with this kind of neural design quick and easy;
  • Numerous new modules and applications – With more than 50 modules and applications demonstrating knowledge representation, vision, learning, and many others, the Brain Simulator has an expanding library to fit many facets of AGI.

“These new features, coupled with continued experimentation into varied AI algorithms, are eliminating the limitations of computer power and accelerating the evolution of Artificial Intelligence into AGI, with true AGI expected by the end of this decade,” Simon explains.

By advancing new algorithms to simulate biological neuron circuits coupled with high-level AI techniques, Brain Simulator II already is able to implement an artificial entity known as “Sallie” who can move about, see, touch, and smell objects, and learn to understand speech within a simulated environment

According to Simon, current examples of AI, such as ALEXA and SIRI, fall short in the area of reasoning about time and space – things any three-year-old can do. Sallie, on the other hand, furthers existing technologies by recognizing objects with binocular vision, associating them with words she hears, planning a sequence of actions, and manipulating objects to achieve a goal.

As Sallie advances to understanding her world using information captured with biologically plausible techniques in a Universal Knowledge Store (UKS), interfaces already exist for cameras, microphones, and robotic control to bring AGI to life. “Relating information from multiple senses then is the key to comprehending that objects exist in a physical environment – an insight which underlies all human thought,” Simon concludes.

Earlier this month, Simon published Brain Simulator II: The Guide for Creating Artificial General Intelligence, a companion book to provide neuroscientists, cognitive scientists, and programmers interested in AGI, neuron limitations, or spiking AI software with everything they might need to build a system that can learn about object permanence, spatial relationships, and time/causality, and ultimately help AI to become AGI.

For more information about Brain Simulator II and its new features, visit http://brainsim.org. The companion book, Brain Simulator II: The Guide for Creating Artificial General Intelligence (ISBN 978-1-7326872-4-0) is available at: https://futureai.guru/BrainSimGuide.aspx.

Share article on social media or email:

View article via:

Pdf Print

Contact Author

Matthew Pugh
@PughTillerPR
Follow >
Visit website