Microsoft co-founder Paul Allen announces that he will invest $500 million to fund two teams of researchers pursuing breakthroughs in artificial intelligence. Allen, who is 62 and has an estimated worth of $17 billion, has developed a plan and a research team to systematically attack gaps in artificial intelligence know-how. “We are starting with biology” he explains, “But first you have to figure out how you represent that knowledge in a software database. I wish I could say our understanding of the brain could inform that, but we’re probably a decade away from that. Our understanding of the brain is so elemental at this point that we don’t know how language works in the brain.”
To start, Allen has formed two teams that will pursue artificial intelligence from fundamentally opposite starting points. The first will be led by artificial intelligence scientist Oren Etzioni who has joined Allen from the University of Washington as the new chief executive of the Allen Institute for Artificial Intelligence. Etzioni and his team will leverage existing principles of AI design to build a man-made brain capable of independently passing a high school science test. The underlying goal of this effort is to create a brain not only capable of reading and responding to complex questions, but to code a system capable of abstract reasoning.
While complex deep learning algorithms have been created in the past with genuine success, these algorithms were highly specific in their purpose and design, and would fall flat if asked to pivot and apply their reasoning to a new problem. Developing a system with the ability to apply reason to a broad range of problems would be an entirely new accomplishment for the artificial intelligence and computer science fields.
While Etzioni works on building a brain capable of reasoning, a separate team working under Caltech neuroscientist Christof Koch, resident and chief scientific officer at the Allen Institute for Brain Science, will tackle the same problem with the exact opposite approach. Rather than attempting to build a brain from scratch, this team will dissect the human brain with imaging technology and then use information extracted during that process to create analytics models and simulations.
In the end, Allen hopes to merge the discoveries of both teams to push AI from science fiction to reality, resulting in a computer system that can perform outside of the narrowly-defined purpose and develop a general knowledge and ability to reason that could make it useful in a wide variety of everyday tasks.