{"id":6010,"date":"2015-05-27T10:35:15","date_gmt":"2015-05-27T16:35:15","guid":{"rendered":"https:\/\/www.roughtype.com\/?p=6010"},"modified":"2017-09-09T11:01:21","modified_gmt":"2017-09-09T15:01:21","slug":"can-computers-improvise","status":"publish","type":"post","link":"https:\/\/www.roughtype.com\/?p=6010","title":{"rendered":"Can computers improvise?"},"content":{"rendered":"

\"eminem\"<\/a><\/p>\n

Bust this:<\/p>\n

Girl I\u2019m down for whatever cause my love is true
\nThis one goes to my man old dirty one love we be swigging brew
\nMy brother I love you Be encouraged man And just know
\nWhen you done let me know cause my love make you be like WHOA<\/i><\/p><\/blockquote>\n

These rap lyrics, cobbled together by a computer from a database of lines from actual rap songs, “rival those of Eminem,” wrote<\/a>\u00a0Esquire<\/em>‘s Jill Krasny last week.\u00a0I have to think that’s the biggest dis\u00a0ever thrown Eminem’s way. But Krasny\u00a0was\u00a0not the only one gushing over the witless mashup. A Mashable<\/em> headline<\/a>\u00a0said the program, dubbed\u00a0DeepBeat<\/a> by its Finnish creators, “produced rap lyrics that rival human-generated rhymes.” Quartz<\/em>‘s Adam Epstein\u00a0suggested<\/a> that robots can\u00a0now be considered “lyrical wordsmiths.”\u00a0Reporte<\/a>d<\/a> UPI: “Even rappers might soon lose their jobs to robots.”<\/p>\n

I guess it must have been\u00a0a slow news day.<\/p>\n

Silly as it is, the story is not\u00a0atypical, and it illuminates something important about our sense of the possibilities and threats presented by\u00a0computers.\u00a0Our expectations about artificial intelligence have raced ahead of the reality, and that’s\u00a0skewing our view not only of the future but of the very real accomplishments being made in the AI and robotics fields. We take\u00a0a\u00a0modest but meaningful advance<\/a> in natural-language processing \u2014 DeepBeat fits lines together through a statistical analysis of rhyme, line length, and wording, its choices constrained\u00a0by a requirement that a specified keyword (“love” in the example above) appear in every line* \u2014 and we\u00a0leap\u00a0to the conclusion\u00a0that computers are\u00a0mastering wordplay and, by implication, encroaching on the human facility for creativity and improvisation. In the process, we denigrate the accomplishments of talented people \u2014 just to make the case for the computer seem a little more compelling.<\/p>\n

We humans have a well-documented tendency to perceive human characteristics in, and attribute human agency to, inanimate objects. That’s\u00a0a side effect, scientists believe, of the\u00a0human\u00a0mind’s exquisite sensitivity to social signals.\u00a0A hint of human-like cognition or behavior triggers a sense that we’re in the presence of a human-like being. The\u00a0bias becomes particularly strong when we observe\u00a0computers and automatons performing manual or analytical tasks similar to those we do\u00a0ourselves.\u00a0Joseph Weizenbaum, the MIT computer scientist who wrote the program for the early chatbot ELIZA<\/a>, limned\u00a0the phenomenon in a 1966 paper<\/a>:<\/p>\n

It is said that to explain is to explain away. This maxim is nowhere so well fulfilled as in the area of computer programming, especially in what is called heuristic programming and artificial intelligence. For in those realms machines are made to behave in wondrous ways, often sufficient to dazzle even the most experienced observer. But once a particular program is unmasked, once its inner workings are explained in language sufficiently plain to induce understanding, its magic crumbles away; it stands revealed as a mere collection of procedures, each quite comprehensible.<\/p><\/blockquote>\n

The procedures have grown more complex and impressive since then, and their practical applications have widened enormously, but Weizenbaum’s point still holds. We’re quick to mistake clever programming for actual talent.<\/p>\n

Last week, in a New York Times<\/em> op-ed<\/a>\u00a0examining human and machine error, I wrote a couple of sentences that I suspected\u00a0would raise the odd hackle: “Computers are wonderful at following instructions, but they\u2019re terrible at improvisation. Their talents end at the limits of their programming.” And hackles were raised. Krasny hooked her Esquire<\/em> piece on DeepBeat to those lines, arguing that the program’s\u00a0ability to spot\u00a0correlations\u00a0in spoken language\u00a0is an example of machine\u00a0improvisation that proves me wrong. On Twitter, the sociologist and technology writer Zeynep Tufekci suggested I was\u00a0“denying the true state of advances in artificial intelligence.” She wrote<\/a>: “I don’t agree that [computers]\u00a0are unable to ‘improvise’ in the most practical way.”<\/p>\n

The quotation marks in Tufekci’s\u00a0statement are revealing. If we redefine what we mean by improvisation<\/em>\u00a0to encompass\u00a0a computer’s ability to respond programmatically to events within a constrained field of activity, then, sure, we can say that computers “improvise.” But that’s not what we really mean by improvisation. To improvise \u2014 the word derives from a Latin term meaning “without preparation” \u2014 is to act without instruction, without programming, in novel and unforeseen situations. To improvise is to go off script<\/em>.\u00a0Our talent for improvisation, a talent we share with other animals,\u00a0stems from the mind’s ability to translate particular experiences into a store of general know-how, or common sense, which then can be deployed, fluidly and often without conscious deliberation, to meet new challenges in new circumstances.<\/p>\n

No computer has demonstrated an act of true improvisation, an act that can’t be explained by the instructions written by its programmers.\u00a0Great strides are being made in machine learning and other AI techniques,** but the programming of common sense remains out of reach.\u00a0The cognitive scientist\u00a0Gary Marcus, in a recent New Yorker<\/em> essay, “Hyping Artificial Intelligence, Yet Again,” explains<\/a>:<\/p>\n

Trendy new techniques like deep learning<\/a> and neuromorphic engineering give A.I. programmers purchase on a particular kind of problem that involves categorizing familiar stimuli, but say little about how to cope with things we haven\u2019t seen before. As machines get better at categorizing things they can recognize, some tasks, like speech recognition, improve markedly, but others, like comprehending what a speaker actually means, advance more slowly.<\/p><\/blockquote>\n

Marcus is\u00a0hardly blas\u00e9 about advances in artificial intelligence. He thinks<\/a> it likely that “machines will be smarter than us before the end of the century \u2014 not just at chess or trivia questions but at just about everything, from mathematics and engineering to science and medicine.” But he stresses\u00a0that we’re still a long way from building machines with common sense, much less an ability to program\u00a0themselves, and he’s skeptical that existing AI techniques will get us there.<\/p>\n

Since we don’t know how the minds of human beings and other animals develop common sense, or gain\u00a0self-awareness, or learn to improvise in novel situations,\u00a0we have\u00a0no template to follow in designing machines with such abilities. We’re working in the dark. As University of California, Berkeley, professor Michael Jordan, a leading expert in machine intelligence, said<\/a> in an IEEE Spectrum<\/em> interview with Lee Gomes last year, when it comes to “issues of higher cognition \u2014 how we perceive, how we remember, how we act \u2014 we have no idea how neurons are storing information, how they are computing, what the rules are, what the algorithms are, what the representations are, and the like. So we are not yet in an era in which we can be using an understanding of the brain to guide us in the construction of intelligent systems.”<\/p>\n

And yet the assumption that computers are replicating our own thought processes\u00a0persists. In reporting on DeepBeat’s use of a neural network, two\u00a0Wall Street Journal<\/em> bloggers wrote<\/a> that the software “is based on a field of artificial intelligence that mimics the way the human brain works.” Writing about neural nets in general, Wired<\/em>‘s Cade Metz says<\/a>\u00a0that\u00a0“these systems approximate the networks of neurons inside the human brain.” As\u00a0Jordan cautions,\u00a0“people continue to infer that something involving neuroscience is behind [neural nets], and that deep learning is taking advantage of an understanding of how the brain processes information, learns, makes decisions, or copes with large amounts of data. And that is just patently false.”<\/p>\n

Even Marcus’s expectation of the arrival of human-level artificial\u00a0intelligence in the next eighty or so years is based on a faith that we will find a way, without a map in hand or in the offing, to cross the undiscovered country that lies between where we are today\u00a0and the promised land of human-level AI. That’s not to say it can’t happen. Our own minds would seem to be\u00a0proof that common sense and improvisational skill can come from an assemblage of inanimate components. But it is to say that predictions that it will happen \u2014 in twenty years or fifty years or a hundred years \u2014 are speculations, not guarantees. They assume a lot of things that haven’t happened yet.<\/p>\n

If in interpreting the abilities of machines we fall victim to our anthropomorphizing instinct, in forecasting the progress of machine abilities\u00a0we’re often misled by our tendency to place unwarranted faith in our prediction techniques. Many of the predictions for the rapid arrival of human-like artificial intelligence, or “superintelligence” that exceeds human intelligence, begin with reference to “the exponential advance of computer power.” But even if we assume a doubling of available\u00a0computer-processing power every year or two indefinitely into the future, that doesn’t tell us much about how techniques for programming AI\u00a0will unfold. In warning against AI hype in another IEEE Spectrum<\/em> interview, published earlier this year, Facebook’s director of AI research, Yann LeCun,\u00a0described<\/a>\u00a0how easy it is to be led astray in anticipating ongoing exponential advances in AI programming:<\/p>\n

As\u00a0Neil Gershenfeld\u00a0has noted, the first part of a\u00a0sigmoid\u00a0looks a lot like an exponential. It\u2019s another way of saying that what currently looks like exponential progress is very likely to hit some limit \u2014 physical, economical, societal \u2014 then go through an inflection point, and then saturate. I\u2019m an optimist, but I\u2019m also a realist.<\/p><\/blockquote>\n

Here’s a\u00a0sketch\u00a0of the sigmoidal pattern of progress that LeCun is talking about:<\/p>\n

\"sigmoid\"<\/a><\/p>\n

It’s very easy to get carried away during that exponential phase (“computers will soon do everything that people can do!”; “superintelligence will arrive in x<\/em> years!”), and the more carried away you get, the more disheartened you’ll be when the plateau phase arrives. We’ve seen this cycle of hype and disappointment before in the progress of artificial intelligence, and\u00a0it hasn’t just\u00a0distorted our sense of the future of AI; it has also had a debilitating\u00a0effect on the research itself. Coming after a period of hype, the plateau\u00a0stage tends to provoke a sharp drop in both interest and investment, which ends up\u00a0extending the plateau. Notes LeCun, “AI has gone through a number of AI winters because people claimed things they couldn\u2019t deliver.”<\/p>\n

To be skeptical about the promises being made for\u00a0AI is not to denigrate the ingenuity of the people who are actually doing the work.\u00a0Appreciating the difficulties, uncertainties, and limits in pushing AI forward makes the achievements\u00a0of computer scientists and programmers in the field seem more impressive, even heroic. If computers were actually capable of improvisation, a lot of the hardest programming challenges\u00a0would evaporate. It’s useful, in this regard, to look at how Google has gone about programming its “autonomous” car to deal with unusual traffic situations. If you watch the car perform tricky maneuvers, you might be tempted to think that it has common sense\u00a0and a talent for improvisation. But what it really has is very good and diligent programmers. Here’s how Astro Teller, the head of Google X, the R&D unit developing the vehicle, explains<\/a> how the\u00a0process has proceeded:<\/p>\n

When we started, we couldn\u2019t make a list of the 10,000 things we\u2019d have to do to make a car drive itself. We knew the top 100 things, of course. But pretty good, pretty safe, most of the time isn\u2019t good enough. We had to go out and just find a way to learn what should be on that list of 10,000 things. We had to see what all of the unusual real world situations our cars would face were. There is a real sense in which the making of that list, the gathering of that data, is fully half of what is hard about solving the self driving car problem.<\/p><\/blockquote>\n

The Google team, Teller says,\u00a0“drives a thousand miles of city streets every single day, in pursuit of moments that stump the car.” As the team methodically\u00a0uncovers novel\u00a0driving challenges \u2014 challenges that human drivers routinely handle with ease, without requiring new instructions \u2014 it updates the car’s software to give the vehicle the ability to handle new categories\u00a0of situations:<\/p>\n

When we produce a new version of our software, before that software ends up on our actual cars, it has to prove itself in tens of thousands of [possible situations]\u00a0in our simulator, but using real world data. We show the new software moments like this and say \u201cand what would you do now?\u201d Then, if the software fails to make a good choice, we can fail in simulation rather than in the physical world. In this way, what one car learns or is challenged by in the real world can be transferred to all the other cars and to all future versions of the software we\u2019ll make so we only have to learn each lesson once and every rider we have forever after can get the benefit from that one learning moment.<\/p><\/blockquote>\n

Behind the illusion of machine improvisation lies a whole lot of painstaking effort. In an article this week about the Darpa Robotics Challenge, an annual event that tests the limits of robots, New York Times<\/em> science reporter John Markoff emphasized<\/a> this point:<\/p>\n

\n

Pattern recognition hardware and software has made it possible for computers to make dramatic progress in computer vision and speech understanding. In contrast, [Darpa program manager Gill]\u00a0Pratt said, little headway has been made in \u201ccognition,\u201d the higher-level humanlike processes required for robot planning and true autonomy. As a result, both in the Darpa contest and in the field of robotics more broadly, there has been a re-emphasis on the idea of human-machine partnerships.\u00a0\u201cIt is extremely important to remember that the Darpa Robotics Challenge is about a team of humans and machines working together,\u201d he said. \u201cWithout the person, these machines could hardly do anything at all.\u201d<\/p>\n<\/blockquote>\n

If you’re worried about a robot or an AI algorithm taking your job, you can\u00a0take a little\u00a0comfort in what I’ve written here. But you shouldn’t take a lot of comfort. As the Google car shows, computers can take over a whole lot of sophisticated manual and intellectual work without demonstrating any common sense or improvisational skill. Many years ago, Alan Turing observed<\/a> that, as computers sped up and databases swelled, the “ingenuity” of programmers\u00a0would be able to be substituted for the “intuition” of skilled professionals in many fields. We’re seeing that today on a broad scale, and we’re going to be seeing even more of it tomorrow. But Turing also concluded\u00a0that there would always be limits to the use\u00a0of ingenuity. There would always be an important place for intuition \u2014 for “spontaneous judgments which are not the result of conscious trains of reasoning.” Computers would not be able to substitute for talented, experienced people in all situations. And we’re seeing plenty of evidence of that today, too. If you’re a rapper, you may need to worry about shifts in fashion \u2014 capriciousness is another human quality\u00a0that computers can’t match \u2014 but you can rest assured that robot rappers pose no threat whatsoever to your livelihood. Computers in the future will be able to\u00a0do more than we assume but less than we fear.<\/p>\n

Ever since our ancestors\u00a0made the first tools, we have been dividing labor between ourselves and our technologies. And the line between human effort and machine effort is always changing, for better and for worse. It would probably be a good idea to\u00a0spend a little less time worrying about, or yearning for, a future in which robots take all our jobs and a little more time thinking about how to divide labor between people and computers in the wisest possible way.<\/p>\n

______<\/p>\n

*The Finnish researchers admit that when they apply their statistical model of rap lyrics to Eminem’s work, it scores poorly. The reason? Eminem is a master at\u00a0“bending” \u2014 altering the\u00a0pronunciation of words to create assonance and other rhymes where the rules say they shouldn’t exist. Eminem, in other words, improvises.<\/p>\n

**Tomorrow, for example, Berkeley researchers will present<\/a>\u00a0a paper on a machine-learning technique that appears to enable a robot to master\u00a0simple new tasks through a process of trial and error.<\/p>\n

Image: Marvel<\/a>.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"

Bust this: Girl I\u2019m down for whatever cause my love is true This one goes to my man old dirty one love we be swigging brew My brother I love you Be encouraged man And just know When you done let me know cause my love make you be like WHOA These rap lyrics, cobbled […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"jetpack_post_was_ever_published":false},"categories":[1],"tags":[],"class_list":["post-6010","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/www.roughtype.com\/index.php?rest_route=\/wp\/v2\/posts\/6010","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.roughtype.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.roughtype.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.roughtype.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.roughtype.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=6010"}],"version-history":[{"count":10,"href":"https:\/\/www.roughtype.com\/index.php?rest_route=\/wp\/v2\/posts\/6010\/revisions"}],"predecessor-version":[{"id":8101,"href":"https:\/\/www.roughtype.com\/index.php?rest_route=\/wp\/v2\/posts\/6010\/revisions\/8101"}],"wp:attachment":[{"href":"https:\/\/www.roughtype.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6010"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.roughtype.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6010"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.roughtype.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6010"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}