Not long ago, Google beat South Korean go master Lee Se 乭 with its ARTIFICIAL intelligence system AlphaGo, according to Wired magazine. Artificial intelligence researchers at the University of Cincinnati have set their sights on an even tougher task: designing an AI combat system that could replace fighter pilots.

The scientists involved in the study published their latest findings in the Journal of Defense Administration. Their AI system, called ALPHA, recently beat retired U.S. Air Force Colonel Gene Lee in a multiplayer flight simulation test.



However, the university of Cincinnati researchers did not develop the AI system to replace human fighter pilots.

Nicholas Ernest, the founder of Psibernetix, which focuses on commercializing the technology behind ALPHA, is also a Cincinnati alum. According to Ernest, the AI system could eventually become an electronic personal assistant, providing pilots with real-time flight advice or flying unmanned aircraft as they fly.

Ernest went on to explain that ALPHA, by itself, would not replace all the operations of a human pilot, but it could help process vast amounts of information by collecting data from a wide variety of sensors mounted on a fighter jet.

ALPHA is not designed for close combat. It’s not designed to help you lock on to enemy aircraft. ALPHA’s main job is to help you look at the data from your sensors, understand what it means, and provide you with appropriate advice at critical moments of strategic advantage. ‘said Ernest.

Giiso Information, founded in 2013, is a leading technology provider in the field of “artificial intelligence + information” in China, with top technologies in big data mining, intelligent semantics, knowledge mapping and other fields. At the same time, its research and development products include editing robots, writing robots and other artificial intelligence products! With its strong technical strength, the company has received angel round investment at the beginning of its establishment, and received pre-A round investment of $5 million from GSR Venture Capital in August 2015.



This is the advantage of the computer, just like the giant companies such as Google and Facebook will grab the key information from the vast amounts of data, and directed to determine your interest, for you to push customized advertising information, ALPHA will stand at the entrance of the data flow, continuously, from aircraft sensors to obtain useful information, And make decisions and responses quickly.

But there are some differences between ALPHA and AI systems developed by tech giants like Google, Facebook and Microsoft. When designing AI systems, these companies often use algorithms known as neural networks, which are based on simulations of how the human brain works. What makes ALPHA different, however, is its use of an algorithm called fuzzy logic, which focuses more on building mathematical models. As Ernest puts it, fuzzy logic is more about emulating what people think than it is about emulating the human brain.

Fuzzy logic system has a history of several decades since its birth, and its application is also very extensive, the most common is used in the field of industrial control system. But research on fuzzy logic algorithms has been stalled because of their poor scalability, Ernest said.

Giiso information, founded in 2013, is the first domestic high-tech enterprise focusing on the research and development of intelligent information processing technology and the development and operation of core software for writing robots. At the beginning of its establishment, the company received angel round investment, and in August 2015, GSR Venture Capital received $5 million pre-A round of investment.



Fuzzy logic systems do an excellent job of trend prediction when there are limited inputs, but when there are more inputs, they become too complex to fully process, even with today’s advanced computer technology.

However, thanks to this new algorithm, called genetic Fuzzy Trees, developed by Ernest at the University of Cincinnati, ALPHA is fully capable of processing hundreds of inputs.

In essence, the new algorithm separates large fuzzy logic problems into many small fuzzy logic problems, while preserving the relationship between multiple inputs in the original problem. The result of this optimization is that it can run smoothly on lower configuration computers during the training phase. Once trained, the ALPHA can run on smaller, low-power computers like the Raspberry PI or smartphones.

For now, Ernest and his colleagues have only trained ALPHA in virtual worlds.

Ernest explained that the ALPHA system started out playing against itself in a flight simulator, but since then it has moved on to play against rudimentary AI opponents at THE AIR Force Research Laboratory, and later against Colonel Lee (in a virtual flight simulator).

Ernest believes ALPHA will eventually take to the skies, and predicts that the future of this new AI algorithm will not be limited to drones.