User:NickGilman/sandbox

Mounir Shita

Mounir Shita (b. 1976) is a Norwegian serial entrepreneur and artificial intelligence (AI) theoretician and researcher focusing on Artificial General Intelligence (AGI). He was born and grew up in Oslo, Norway, the son of Moroccan immigrants.

Shita currently resides in the USA in the Portland Metro Area of Oregon and is the Founder and CEO of Kimera Systems, Inc.

Theories of Artificial General Intelligence (AGI)

Shita's AGI theories center on what has traditionally been called Strong AI, a sharp contrast to the more typical and common Weak or Narrow AI, a type of AI pre-programmed to accomplish a specific task or set of tasks. Shita's AGI focus is born of his belief that disparate objects narrowly programmed with AI lack scalability and cannot achieve a future in which human beings are served by an intelligent global network that efficiently orchestrates the interoperation and coordination of many devices to help humans accomplish their individual and broader societal goals.

Defining Intelligence

Shita began his research into AI by asking questions about the nature of intelligence: What is it? This line of inquiry was fundamentally different from the neuroscience-centric approach, which presupposes the human brain is only valid model of intelligence. Additionally, using the brain as the model for AI is problematic, since, Shita argues, so many brain functions are still not well understood and objective testing is difficult, if not impossible. Shita believed the best path to AGI is in understanding the nature of intelligence: What is intelligence?

The research efforts first suggested an assumption that more intelligent lifeforms are effectively better at achieving a variety of goals than lesser intelligent lifeforms. This, in turn, led Shita to define goal achievement by examining intelligence at the most fundamental level of spacetime. For instance the goal of moving a cup from one side of a table to the other. It 's a simple goal easily achieved, but how is it best described? Shita concluded, "From the most basic perspective of physics, this new state can be described as a collection of particles, the cup, being moved to a new location, changing the composition of spacetime. The new version of spacetime is interpreted as achievement of this goal. While the cup example is simplistic, it turns out that any goal, easy or complicated, requires the movement of particles in order to achieve the desired result. There is no conceivable goal can be accomplished without moving particles. This led Shita to postulate: Intelligence is the process of changing the composition of spacetime." This quantum physics approach has arguably allowed Kimera Systems to claim that it has achieved a human-like intelligence algorithm decades sooner than a great many scientists thought possible.

AGI Single-Algorithm Solution

Producing a theory of intelligence at the fundamental particle level allowed Shita's company, Kimera Systems, to develop a machine algorithm for implementation that differentiated it from other AI research and solutions. While other AI companies are largely focused on recognizing narrowly-defined patterns and making predictions, Kimera’s solution is based upon a universal theory of intelligence that uses a single algorithm to solve virtually any problem, thus truly putting the general into Artificial General Intelligence (AGI). Shita believes the advantage of this approach is extremely powerful. AGI is able to learn in any domain and apply that knowledge in any other domain where it has value. As a comparison, an existing AI that is designed to play chess will learn to play chess better and better till it becomes unbeatable. However, existing AI programs are domain-specific or domain-limited and cannot apply that knowledge in any other area, such as in a game of poker or industrial controls or any other area. An AGI single-algorithm solution will be able to apply that knowledge in any other domain where it has value.

From a technical perspective, Shita sees AGI as an essential technology to transform the global network from passive to proactive. For example, AGI should be intelligent enough to summon a ride share or cab automatically when dinner ends and already knows your desired destination. Kimera’s AGI uses a three-layered decentralized architecture:

Layer 1: On the devices themselves

Layer 2: An open-source agent that people can host anywhere they want on the global network

Layer 3: An AGI Node, which will initially be hosted by Kimera and will expand to be hosted on networks anywhere in the world. Kimera’s AGI is an infrastructure technology where AGI Nodes running on networks around the globe deliver information about how to establish peer-to-peer connections to any other network-connected device.

AGI-Powered Connections Across the Global Network

Consider the following analogy: A human who wants to exit through a doorway will first notice whether the door is open or closed. It is the brain that comprehends and evaluates the position of the door and, if it is closed, the brain will communicate to the appropriate muscles the necessary actions the body needs to take to open the door. The brain determines which actions are required and communicates with the body to complete the task. Similarly, an AGI Node acts as the brain and the devices—software applications, content and services found across the global network—act as the body parts. As the AGI Node understands what needs to change in a user's environment to move the user closer to realizing goals, it tells these “body parts” how to increase the probability of influencing changes in the environment and achieving the intended goal. AGI proactively establishes peer-to-peer connections across the global network to bring together the right agents, software and content to influence the right changes in the environment. These peer-to-peer connections provide great benefits to participants and software and content developers:

1) Some decision making becomes autonomous. How much authority an Agent has for autonomous decision making is controlled by each person.

2) When Agents cannot, or do not have authority to make autonomous decisions, the right software is automatically delivered to people’s devices, providing them with the tools they need to move closer to their goals.

3) Software developers who develop these applications (be it a shopping list, profile viewer, public transportation application, etc.) don’t have to worry about app distribution and marketing, because the AGI Node will match a person’s need to an app, the AGI Node will recommend to a person’s Agent to surface the app on their device.

Kimera Systems

Kimera Systems was founded by Shita in June, 2012. The Company's purpose is to further refine and commercialize Shita's single-algorithm AGI technology, which has been named Nigel.