CogniGron Seminar: Carlo Reita (CEA - LETI, France) - "Human Brain or Bee Brain? Hardware targets for AI"

In recent years, the field of Neural Networks has found a new golden age after nearly twenty years of lessened interest. Under the heading of Artificial Intelligence (AI) a large number of Deep Neural Netwooks (DNNs) have recently find application in image processing, management of information in large databases, decision aids, natural language recognition, etc. Most of these applications rely on algorithms that runs on standard computing systems and sometimes make use of specific accelerators like Graphic Processor Units (GPUs) or dedicated highly parallel processors. However, when comparing with biological system, the most striking differences are the energy necessary for the computation and the capability of real time interaction with the environment. While there are discussion about AI systems to be superior to the human brain, in reality present systems are in a sense even less capable than an insect. A revised hardware sense-process-act loop as well as a more closed algorithm-hardware link are necessary before we start to approach biological entities capability and efficiency, and the goal of a bee’s brain is probably a good starting point.