Are we on the brink of achieving human-level AI? What’s holding us back from creating truly intelligent machines?<\/p>\n\n\n\n
In their book Rebooting AI<\/em>, Gary Marcus and Ernest Davis argue that current AI approaches<\/a> fall short of human-level intelligence. They propose that AI research should draw inspiration from the human brain’s efficiency and structure.<\/p>\n\n\n\n Continue reading for Marcus and Davis’s insights on building machines that think more like us.<\/p>\n\n\n\n\n\n\n\n Narrow AIs are computational black boxes where information goes in, passes through a single (if convoluted) algorithm, and then comes out reprocessed as a new result. That’s not how the human brain works<\/a>, and Marcus and Davis argue that strong, human-level AI shouldn’t work like that either. Instead, AI research should draw on the efficiencies of the human brain, such as how it uses different processes for different types of information, how it creates abstract models of the world with which it interacts, and how it goes beyond correlating data to think in terms of causality<\/em>\u2014using mental models<\/a> to understand how the external world changes over time.<\/p>\n\n\n\n Unlike narrow, single-system AI, the brain is a collection of systems that specialize in different types of information\u2014sight, sound, and touch, for example\u2014while regulating different forms of output, such as conscious and unconscious bodily functions. Likewise, Marcus and Davis write that strong AI should incorporate multiple processing systems and algorithms<\/strong> to handle the panoply of inputs and problems it will encounter in the real world. Also like the human brain, strong AI must be flexible enough to combine its different systems in whatever arrangement is needed at the moment, as humans do when we associate a memory with a smell, or when an artist engages both her visual and manual skills to produce a painting or a sculpture.<\/p>\n\n\n\n (Shortform note: The idea of modeling machine intelligence on the brain can be traced back to computer scientist Norbert Wiener<\/a>\u2019s book Cybernetics<\/em><\/a>, published in 1948. However, most brain-based AI research has focused on simulating the behavior of neurons<\/a> with little attention given to replicating the brain\u2019s larger structures. That trend may be changing, with the development of a new type of neural network called a transformer. It mimics the brain\u2019s hippocampus<\/a>, which is closely tied to learning and memory<\/a>. Transformers have proven useful in solving translation issues that plagued older AI frameworks<\/a>, hinting that Marcus and Davis are correct that closer brain emulation is key to building more powerful AIs in the future.)<\/p>\n\n\n\n A more challenging but necessary step to strong AI will be designing knowledge frameworks that will let computers understand the relationships among entities, objects, and abstractions.<\/strong> These relations form the bedrock of human thought in the form of the mental models we use to interpret the world around us. Davis and Marcus state that modern AI developers largely disregard knowledge frameworks as an AI component. But without knowing how information interrelates, AI is unable to synthesize data from different sources and fields of knowledge. This ability is as crucial to advancing scientific progress<\/a> as it is to driving a car in the snow while adjusting your schedule because of school closings\u2014all of which AI could help with.<\/p>\n\n\n\n (Shortform note: Before AI can understand relationships between the real world and abstract concepts, it will first have to be able to perform abstract reasoning<\/a>\u2014solving problems conceptually <\/em>through logic and imagination without complete data or exact prior experience. Swiss researchers have made progress in this field using a combination of machine learning and a \u201cvector symbolic\u201d architecture<\/a> that\u2019s conceptually similar to the knowledge frameworks Davis and Marcus recommend. Meanwhile, software engineer Fran\u00e7ois Chollet has created the Abstraction and Reasoning Corpus<\/a>, a tool for measuring how effectively AI systems perform basic abstract reasoning tasks as compared to humans presented with the same challenge.)<\/p>\n\n\n\n In addition to having a knowledge framework, strong AI must understand causality\u2014how and why objects, people, and concepts change over time.<\/strong> Marcus and Davis say this will be hard to achieve because causality leads to correlation, but not all correlations are causal. For example, though most children like cookies (a correlation), enjoying cookies doesn\u2019t cause <\/em>childhood. Furthermore, AI will have to juggle causality in multiple fields of knowledge at once. For instance, an AI working on a legal case will have to understand how human motivations interact with physical evidence. At present, the only way for a computer to model causality is to run multiple simulations, but that\u2019s far less efficient than how the human brain works. Therefore, we must design our strong AIs to learn these concepts the way that humans learn. <\/p>\n\n\n\n (Shortform note: It\u2019s not accurate to say that our human understanding of causality isn\u2019t based on simulations\u2014we just call them stories <\/em>instead. In Wired for Story<\/em><\/a>, Lisa Cron explains that the human brain uses story structure to store information and make causal predictions<\/a>, such as what will happen if you poke a sleeping bear. The difference between stories and computer simulations is that stories are largely symbolic<\/a>, revolving around a few key elements rather than a plethora of details, and are therefore more versatile and efficient than full simulations. AI developers are now starting to use narrative structures<\/a> to organize AI-generated content and to shape the data used for AI training<\/a> in much the same way that the human brain parses data.)<\/p>\n\n\n\n When it comes to teaching computers how to think, we can draw from millions of years of human evolution<\/a> and incorporate facets of the human brain\u2019s tried-and-true approach to learning into machine cognition. While current machine learning operates from a starting point of pure data, Davis and Marcus argue that preprogrammed knowledge about the world similar to what humans are born with can facilitate stronger AI development. The authors describe how the brain learns<\/a> by combining first-hand experience with preexisting knowledge, how AIs could use this tactic to construct their own knowledge frameworks, and why a hybrid approach to machine learning would be more powerful than the current data-only techniques.<\/p>\n\n\n\n When learning, people draw from two sources\u2014high-level conceptualizations that are either instinctive or taught to us by others, and low-level details that we absorb through day-to-day experiences. Our preexisting, high-level knowledge provides a vital framework through which we interpret whatever we discover on our own.<\/strong> For example, we\u2019re born knowing that food goes into our mouths and tastes good\u2014as children, we then use this framework to determine what does and doesn\u2019t qualify as food. However, Marcus and Davis report that AI developers shun the idea of preprogramming knowledge into neural networks, preferring their systems to learn from data alone, free of any context that would help it make sense.<\/p>\n\n\n\n (Shortform note: Psychologists classify humans\u2019 instinctive knowledge<\/a> into two broad categories\u2014individual instincts, such as eating and self-protection, and social instincts like reproduction and play. Instincts are different from autonomic functions such as digestion and blood circulation in that they trigger complex behaviors and are tied to emotional responses. In Instinctive Computing<\/em><\/a>, computer science professor Yang Cai discusses how and what types of instincts should be programmed into AI to bring it closer in line with biological intelligence while laying the groundwork for AI self-awareness, a topic that Marcus and Davis barely broach.)<\/p>\n\n\n\n Preprogrammed knowledge frameworks\u2014like a set of \u201cinstincts\u201d an AI would be born with\u2014 could greatly advance AI language<\/a> comprehension. When humans read or listen to language, we construct a mental model of what\u2019s being described based on our prior understanding of the world. Davis and Marcus argue that giving an AI a preprogrammed basis for connecting language to meaning would let it construct its own knowledge frameworks<\/strong>, just as humans learn over time. By insisting that AIs learn from raw data alone, developers tie their hands behind their backs and create an impossible task for AI, like giving a calculator the complete works of Shakespeare and expecting it to deduce the English language from scratch.<\/p>\n\n\n\nHuman-Level AI<\/h2>\n\n\n\n
Mental Models and Causality<\/h3>\n\n\n\n
Brain-Like Learning<\/h3>\n\n\n\n