An Anarchy of Methods: Current Trends in How Intelligence Is Abstracted in AI

Joel Lehman, Jeff Clune, Sebastian Risi
2014 IEEE Intelligent Systems  
of connectionist neural networks, while others use mathematical models of decision processes or view intelligence as symbol manipulation. Similarly, researchers focus on different processes for generating intelligence, such as learning through reinforcement, natural evolution, logical inference, and statistics. The result is a panoply of approaches and subfi elds. Because of independent vocabularies, internalized assumptions, and separate meetings, AI subcommunities can become increasingly
more » ... ated from one another even as they pursue the same ultimate goal. Further deepening the separation, researchers may view other approaches only in caricature, unintentionally simplifying the motivations and research of other researchers. Such isolation can frustrate timely dissemination of useful insights, leading to wasted effort and unnecessary rediscovery. To address such dangers, we organized an AAAI Fall Symposium called "How Should Intelligence Be Abstracted in AI Research" that gathered experts with diverse perspectives on biological and synthetic intelligence. The hope was that such a meeting might lead to a productive examination of the value and promise of different approaches, and perhaps even inspire syntheses that cross traditional boundaries. However, organizing a crossdisciplinary symposium has risks as well. Discussion could have focused narrowly on intractable disagreements, or on which singular abstraction is "the best." An unhelpful slugfest of ideas could have emerged instead of collaborative cross-pollination, leading to a veritable AI Tower of Babel. In the end, there were world-class keynote speakers spanning AI and biology (see Table 1 ), and participants were indeed collaborative. Some traveled to the United States from as far as Brazil, Australia, and Singapore; but beyond geographic diversity, there were representatives from many disciplines and approaches to AI (see Figure 1) . Drawing from the symposium's talks and events, we now summarize recent progress across AI fi elds, as well as the key ideas, debates, and challenges identifi ed by the attendees. (See also the sidebar, "Straight from the Experts," which showcases and summarizes the direct viewpoints of some of the keynote speakers.) Key Ideas Discussed One controversial topic was deep learning, which has recently shattered many performance records over an impressive spectrum of machine learning tasks. 1,2 The central idea behind deep learning is that large hierarchical artifi cial neural networks (ANNs), inspired by those found in the neocortex, can be trained on big data (for example, millions of images) to learn a hierarchy of increasingly abstract features 3 (see Figure 2) . Overall, participants agreed that recent progress in deep networks was a signifi cant step forward for processing streams of high-dimensional raw data into meaningful abstract representations, which is required for tasks like recognizing faces from unprocessed pixel data. But there was also agreement that much work remains to create algorithms that leverage such representations to produce intelligent behavior and learn in real-time from feedback; in other words, scaling deep learning to more cognitive behavior may prove problematic. W hile researchers in AI all strive to create intelligent machines, separate AI communities view intelligence in strikingly different ways. Some abstract intelligence through the lens
doi:10.1109/mis.2014.92 fatcat:khlf4nvqgvfzvcy37rfksuij5q