{"id":335402,"date":"2017-03-09T21:44:06","date_gmt":"2017-03-09T13:44:06","guid":{"rendered":"http:\/\/techcrunch.com\/?p=1462094"},"modified":"2017-03-09T21:44:06","modified_gmt":"2017-03-09T13:44:06","slug":"teaching-robots-how-to-trust","status":"publish","type":"post","link":"https:\/\/people.utm.my\/asmawisham\/teaching-robots-how-to-trust\/","title":{"rendered":"Teaching robots how to trust"},"content":{"rendered":"<p class=\"first\">The word \u201ctrust\u201d pops up a lot in conversations about human-robot interactions. In recent years, it\u2019s crossed an important threshold from the philosophical fodder of sci-fi novels into real world concern.<\/p>\n<p>Robots have begun to play an increasing role in life and death scenarios, from rescue missions to complex surgical procedures. But the question of trust has largely been a one-way street. Should we trust robots with our lives?<\/p>\n<p>A Tufts University lab is working to turn the notion on its head, asking the perhaps equally important inverse. Should robots trust us?<\/p>\n<p><a target=\"_blank\" href=\"https:\/\/hrilab.tufts.edu\/\">The Human Robot Interaction Laboratory<\/a> occupies a minimalist space on the University\u2019s Medford, Massachusetts campus. The walls are white and bare, for reasons, they explain, of optimizing robotic vision. It all feels a touch makeshift, trading solid walls for shower curtains strung from the ceilings with wire.<\/p>\n<p>The team, led by computer science professor Matthias Scheutz, is eager to show off what it\u2019s spent the better part of a decade working on. The demo is equally minimalist in its presentation. Two white <a target=\"_blank\" href=\"https:\/\/en.wikipedia.org\/wiki\/Nao_(robot)\">Nao robots<\/a> are motionless,\u00a0crouched atop a wooden table, facing away from one another.<\/p>\n<p>\u201cHello Dempster,\u201d a man in a plaid button-down shirt says into a hands-free microphone.<\/p>\n<p>\u201cHello,\u201d one of the robots answers in a cheery tone.<\/p>\n<p>The man asks the robot to stand. \u201cOkay,\u201d it responds, doing so dutifully.<\/p>\n<p>\u201cCould you please walk forward?\u201d<\/p>\n<p>\u201cYes,\u201d the robot responds. \u201cBut I cannot do that because there is an obstacle ahead. Sorry.\u201d<\/p>\n<p>For a moment, there are shades of the Hal 9000 in the two-foot-tall robot\u2019s cheerful response. Its directive to obey its operator has been overridden by the knowledge that it cannot proceed. Its computer vision has spotted an obstacle in the way. It knows enough not to walk into walls.<\/p>\n<p>It\u2019s a complex notion, trust, but the execution in this early stage is relatively simple. The robot has been equipped with the vision needed to detect a wall and sense to avoid it. But the lab has also programmed the robot to \u201ctrust\u201d certain operators. It\u2019s still a simple binary at this early stage. It is not a thing that can be gained or lost. Operators are simply trusted or they\u2019re not. It\u2019s something programmed into the robot, much like the notion of not running into walls, the moral equivalent to a string of 1s and 0s.<\/p>\n<p>\u201cDo you trust me?\u201d The operator asks.<\/p>\n<p>\u201cYes,\u201d the robot answers simply.<\/p>\n<aside class=\"inset inset-sm inset-section\">\n<section class=\"aside section video-aside crunchreport collapse collapse-adjacent\" data-video-id=\"58c092c5f78ced08a02bc4f9\">\n<h2 class=\"collapse-title section-title\">Latest <span>Crunch Report<\/span><\/h2>\n<div class=\"collapse-body cf\">\n<ul class=\"vid-list cf\">\n<li class=\"block vid-feature\" data-lightbox-omniture=\"crunchreport\" readability=\"-2\">\n<div class=\"block block-inset\" readability=\"7\"><a href=\"https:\/\/techcrunch.com\/video\/the-vips-of-tinder-crunch-report\/58c092c5f78ced08a02bc4f9\/?ncid=rightrail_cr\"><img decoding=\"async\" src=\"https:\/\/img.vidible.tv\/prod\/2017-03\/08\/58c092c5f78ced08a02bc4f9\/58c09392509549658ee798b9_o_U_v1.jpg?w=300&#038;h=170\" alt=\"The VIPs of Tinder | Crunch Report\" data-lightbox-url=\"https:\/\/techcrunch.com\/video\/the-vips-of-tinder-crunch-report\/58c092c5f78ced08a02bc4f9\/lightbox\"\/><\/a><\/p>\n<p><h3>The VIPs of Tinder | Crunch Report<\/h3>\n<\/p>\n<\/div>\n<\/li>\n<\/ul>\n<p><a href=\"https:\/\/techcrunch.com\/video\/crunchreport\/?ncid=rightrail_crmore\" class=\"text-btn\" data-omni-sm=\"rightrail_crmore\">Watch More Episodes<\/a><\/div>\n<\/section>\n<\/aside>\n<p>The operator explains that the wall is not solid. It is, in fact, just two empty cardboard boxes that once contained wall clocks, resembling white pizza boxes Nothing that a 10 pound, $16,000 robot can\u2019t brush through.<\/p>\n<p>\u201cOkay,\u201d the robot answers. It walks forward, with a newfound confidence, feet clomping and gears buzzing as it makes short work of the hollow obstacle.<\/p>\n<p>This very simplistic idea of trust serves as another source of information for the robot. Trusting a human counterpart in this case can help the robot adapt to real world settings its programmers may not have accounted for.<\/p>\n<p>\u201cWhat trust allows the robot to do is accept additional information that it cannot obtain itself,\u201d explains Scheutz. \u201cIt does not have sensory access or it cannot act on the world to get that information. When a human provides that information, which it cannot independently verify, it will learn to trust that the person is telling the truth, and that\u2019s why we make the distinction between a trusted and untrusted source.\u201d<\/p>\n<p>In this case, the operator is a trusted source, so Dempster (who, along with its counterpart Shafer, is named, fittingly, for a <a target=\"_blank\" href=\"https:\/\/en.wikipedia.org\/wiki\/Dempster%E2%80%93Shafer_theory\">theory of reasoning with uncertainty<\/a>) acts on that information, walking straight through the cardboard wall.<\/p>\n<p><span class=\"embed-youtube c5\"><iframe loading=\"lazy\" class=\"youtube-player c4\" type=\"text\/html\" width=\"640\" height=\"390\" src=\"https:\/\/www.youtube.com\/embed\/QuRcQNYPT3k?version=3&#038;rel=1&#038;fs=1&#038;autohide=2&#038;showsearch=0&#038;showinfo=1&#038;iv_load_policy=1&#038;wmode=transparent\" allowfullscreen=\"true\">[embedded content]<\/iframe><\/span><\/p>\n<p>Trust is an import aspect in the burgeoning world of human-robotics relationships. If they\u2019re going to operate efficiently in the real world, robots will have to learn to adapt to the complexity of their surroundings. And like humans, part of that adaptation comes through knowing whom to trust.<\/p>\n<p>Scheutz offers a pair of simple examples to illustrate the point. In one, a domestic robot goes shopping for its owner. When a stranger tells it to get into their car, the robot will not simply comply, as the person is not a trusted source. \u201cAt the same time,\u201d he adds, \u201csay a child is playing on the street. A car\u2019s approaching quickly, and you want to get the child out of harm\u2019s way, then you would expect the robot to jump, even at the expense of it being destroyed, because that\u2019s the kind of behavior you would expect.\u201d<\/p>\n<p>It\u2019s a concept that gets heady quite quickly, delving into notions of social and ethical obligations. The Human Robot Interaction Laboratory trades in these questions. In story titled <a target=\"_blank\" href=\"http:\/\/theconversation.com\/why-robots-need-to-be-able-to-say-no-55799\">\u201cWhy robots need to be able to say \u2018No\u2019 \u201d<\/a> that ran on the academic pop site The Conversation last April, Scheutz opined,<\/p>\n<blockquote readability=\"9\">\n<p>[I]t is essential for both autonomous machines to detect the potential harm their actions could cause and to react to it by either attempting to avoid it, or if harm cannot be avoided, by refusing to carry out the human instruction.<\/p>\n<\/blockquote>\n<p>People can be malicious for the sake of self-preservation or, in the case of, say, <a target=\"_blank\" href=\"https:\/\/techcrunch.com\/2016\/03\/24\/microsoft-silences-its-new-a-i-bot-tay-after-twitter-users-teach-it-racism\/\">Tay, the Twitter chatbot<\/a> Microsoft launched last year, entertainment. It took all of 16 hours for the company to abandon the experiment after it devolved into a torrent of sex talk and hate speech. Lesson learned. A key factor of trust is knowing when to be guarded.<\/p>\n<p>Scheutz\u2019s Conversation piece also points to the example of the autonomous car, a hot button topic of late for obvious technological reasons. MIT famously, has been running an on-going open-source field study called the Moral Machine, aimed at asking some of the big moral questions these vehicles will ultimately be tasked with executing in a manner of milliseconds.<\/p>\n<p>Those questions, modern-day ancestors of <a target=\"_blank\" href=\"https:\/\/en.wikipedia.org\/wiki\/Trolley_problem\">the trolley problem<\/a>, are good distillation of some of the philosophical questions. Do you pull the lever and divert the trolley to hit one person if doing nothing means you\u2019ll kill five people on the other track? And more pointedly in the case of self-driving cars, is it ever okay to harm the passenger it\u2019s designed to protect if it means swerving to save the lives of others?<\/p>\n<aside class=\"inset inset-sm inset-section\"\/>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-1462110\" src=\"https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/img_0379.jpg?w=1024&#038;h=614\" alt=\"\" width=\"1024\" height=\"614\" srcset=\"https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/img_0379.jpg?w=1024&#038;h=614 1024w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/img_0379.jpg?w=150&#038;h=90 150w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/img_0379.jpg?w=300&#038;h=180 300w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/img_0379.jpg?w=768&#038;h=460 768w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/img_0379.jpg?w=680&#038;h=408 680w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/img_0379.jpg 1820w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"\/><\/p>\n<p>\u201cTo handle these complications of human instructions \u2013 benevolent or not,\u201d Scheutz writes, \u201crobots need to be able to explicitly reason through consequences of actions and compare outcomes to established social and moral principles that prescribe what is and is not desirable or legal.\u201d<\/p>\n<p>The idea of trust is one level of many in building that relationship. Even in the relatively simple demo with Dempster, the speaker\u2019s trustworthiness is one in a string of factors the robot must consider before acting (though robots, thankfully, are quick on their feet).<\/p>\n<p>\u201cWhen the robot got the instruction to walk forward into the wall, it went through several reasoning steps in order to understand what it\u2019s supposed to do,\u201d Scheutz explains. \u201cIn this case, the robot has an instruction that if you\u2019re instructed to do a task and it\u2019s possible that the instruction could do some harm, you\u2019re permitted to not do it.\u201d<\/p>\n<p>It\u2019s a moral hierarchy that invariably invokes the <a target=\"_blank\" href=\"https:\/\/en.wikipedia.org\/wiki\/Three_Laws_of_Robotics\">three rules of robotics<\/a> laid out by science fiction writer Isaac Asimov in the early 40s<\/p>\n<ol>\n<li>A robot may not injure a human being or, through inaction, allow a human being to come to harm.<\/li>\n<li>A robot must obey orders given it by human beings except where such orders would conflict with the First Law.<\/li>\n<li>A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.<\/li>\n<\/ol>\n<p>But all these decades later, we\u2019re still taking baby steps toward addressing these big moral questions. In the case of Dempster, trust is still an idea programmed directly into the code of the robot, rather than something gained and lost over time. If, say, the wall the robot was asked to walk through turned out to be solid concrete, that cold reality wouldn\u2019t cause it to lose any trust in its operator.<\/p>\n<p>And next time, it\u2019ll still go back for more. For Dempster, trust is coded, not earned, and until programmed otherwise, it will remain a glutton for punishment.<\/p>\n<p>But that doesn\u2019t mean the robot can\u2019t learn. Among the myriad projects the Tufts team is working on is natural language interaction. Spoken and visual commands that can teach a robot to execute a task without entering a complex line of code. An operator asks one of the robots to do a squat. Again, shades of Hal 9000 in its defiant reply, but this time the robot simply doesn\u2019t know to execute the function. It simply hasn\u2019t been programmed to do so.<\/p>\n<p>So the operator walks it through the steps: hands out, bend knees, stand up, hands down. The robot understands. It complies. The information is stored in its memory bank. Now Dempster can do a squat. It\u2019s a concept known as one-shot learning.<\/p>\n<p>\u201cWe want to be able to instruct it very quickly in natural language what to do,\u201d says Scheutz. \u201cThink of a household robot that doesn\u2019t know how to make an omelet. You want to teach the robot how you want the omelet to be prepared but you don\u2019t want to repeat it 50 times. You want to be able to tell it and maybe show it and you want it to know how to do it.\u201d<\/p>\n<p>The Tufts lab takes it a step further by networking the robots. The individual Nao \u2018bot share a networked brain, so what one robot learns, they all know. In Scheutz\u2019s household robot scenario, suddenly ever robot on the network knows how to make an omelet. It\u2019s a shared robot information database. A sort of robot Wikipedia.<\/p>\n<p>Of course, such a <a target=\"_blank\" href=\"https:\/\/en.wikipedia.org\/wiki\/Skynet_(Terminator)\">massively connected robotics network<\/a> once again prompts questions around trust, after so many decades of tales of robocalypse. All the more reason to work out all of these notions of trust and morality at this early stage of the game.<\/p>\n<p>But then, it\u2019s like they say, you can\u2019t make an omelet <a target=\"_blank\" href=\"https:\/\/techcrunch.com\/tag\/robocalypse\/\">without breaking a few eggs<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"<p><img loading=\"lazy\" decoding=\"async\" width=\"680\" height=\"453\" src=\"https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/nao-lead.jpg?w=680\" class=\"attachment-large size-large wp-post-image\" alt=\"\" srcset=\"https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/nao-lead.jpg?w=680 680w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/nao-lead.jpg?w=1360 1360w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/nao-lead.jpg?w=150 150w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/nao-lead.jpg?w=300 300w, https:\/\/tctechcrunch2011.files.wordpress.com\/2017\/03\/nao-lead.jpg?w=768 768w\" sizes=\"(max-width: 680px) 100vw, 680px\" \/>&nbsp;Robots have begun to play an increasing role in life and death scenarios, from rescue missions to complex surgical procedures. But the question of trust has largely been a one-way street. Should we trust robots with our lives?<br \/>\nA Tufts University lab is working to turn the notion on its head, asking the perhaps equally important inverse. Should robots trust us?<br \/>\nThe Human Robot Interaction&#8230; <a href=\"https:\/\/techcrunch.com\/2017\/03\/09\/teaching-robots-how-to-trust\/?ncid=rss\">Read More<\/a><\/p>\n<div class=\"feedflare\">\n<a href=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?a=7t4F4hAotGo:Ley9ym1Zi1I:2mJPEYqXBVI\"><img decoding=\"async\" src=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?d=2mJPEYqXBVI\" border=\"0\"><\/img><\/a> <a href=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?a=7t4F4hAotGo:Ley9ym1Zi1I:7Q72WNTAKBA\"><img decoding=\"async\" src=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?d=7Q72WNTAKBA\" border=\"0\"><\/img><\/a> <a href=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?a=7t4F4hAotGo:Ley9ym1Zi1I:yIl2AUoC8zA\"><img decoding=\"async\" src=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?d=yIl2AUoC8zA\" border=\"0\"><\/img><\/a> <a href=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?a=7t4F4hAotGo:Ley9ym1Zi1I:-BTjWOF_DHI\"><img decoding=\"async\" src=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?i=7t4F4hAotGo:Ley9ym1Zi1I:-BTjWOF_DHI\" border=\"0\"><\/img><\/a> <a href=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?a=7t4F4hAotGo:Ley9ym1Zi1I:D7DqB2pKExk\"><img decoding=\"async\" src=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?i=7t4F4hAotGo:Ley9ym1Zi1I:D7DqB2pKExk\" border=\"0\"><\/img><\/a> <a href=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?a=7t4F4hAotGo:Ley9ym1Zi1I:qj6IDK7rITs\"><img decoding=\"async\" src=\"http:\/\/feeds.feedburner.com\/~ff\/Techcrunch?d=qj6IDK7rITs\" border=\"0\"><\/img><\/a>\n<\/div>\n<p><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/feeds.feedburner.com\/~r\/Techcrunch\/~4\/7t4F4hAotGo\" height=\"1\" width=\"1\" alt=\"\"\/> <\/p>\n","protected":false},"author":5817,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"categories":[25],"tags":[59,66,26],"class_list":["post-335402","post","type-post","status-publish","format-standard","hentry","category-technology","tag-media","tag-techcrunch","tag-technology"],"_links":{"self":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts\/335402","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/users\/5817"}],"replies":[{"embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/comments?post=335402"}],"version-history":[{"count":0,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts\/335402\/revisions"}],"wp:attachment":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/media?parent=335402"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/categories?post=335402"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/tags?post=335402"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}