[{"data":1,"prerenderedAt":320},["ShallowReactive",2],{"slug-how-to-play-tricks-on-artificial-intelligence":3},{"post":4,"relatedPosts":207,"relatedBooks":208},{"id":5,"date":6,"date_gmt":7,"guid":8,"modified":10,"modified_gmt":11,"slug":12,"status":13,"type":14,"link":15,"title":16,"content":18,"excerpt":21,"author":23,"featured_media":24,"comment_status":25,"ping_status":25,"sticky":26,"template":27,"format":28,"meta":29,"categories":30,"tags":34,"project_category":51,"contact_email_category":52,"yst_prominent_words":53,"class_list":65,"better_featured_image":91,"acf":141,"yoast_meta":151,"_links":154},25052,"2019-10-10T14:00:05","2019-10-10T12:00:05",{"rendered":9},"https:\u002F\u002Fmedicalfuturist.com\u002F?post_id=25052&_wpnonce=c07bc38614&status=auto-draft&type=post","2023-03-24T15:54:02","2023-03-24T14:54:02","how-to-play-tricks-on-artificial-intelligence","publish","post","https:\u002F\u002Fmedicalfuturist.com\u002Fhow-to-play-tricks-on-artificial-intelligence",{"rendered":17},"How To Play Tricks On Artificial Intelligence?",{"rendered":19,"protected":20},"\n\n\u003Ctable style=\"width: 100%; border-collapse: collapse; background-color: #eee; border-top: 4px solid #444;\" cellpadding=\"5px\">\n\u003Ctbody>\n\u003Ctr>\n\u003Ctd style=\"width: 100%;\">\n\u003Cp style=\"color: #555;font-size: 12px;line-height: 14px\">THIS ARTICLE HAS NOT BEEN UPDATED SINCE 2019. THE INFORMATION SHARED IN THE ARTICLE WAS ACCURATE AT THE TIME OF ITS PUBLICATION, BUT IT MAY BE OUT OF DATE NOW. \u003Ca href=\"https:\u002F\u002Fmedicalfuturist.com\u002Fmagazine\">BROWSE OUR LATEST ARTICLES HERE\u003C\u002Fa>\u003C\u002Fp>\n\u003C\u002Ftd> \u003C\u002Ftr>\u003C\u002Ftbody>\u003C\u002Ftable>\n\n\n\n\u003Cp>While in the last years, sci-fi stories, experts and even some opinion-leader public figures frequently spelled the end of mankind through artificial intelligence, lately, it has turned out just how easy it is to play tricks on smart algorithms and fool them into making errors. We looked around how researchers can hack A.I. and what that means for medicine and healthcare, mainly from a security perspective.\u003C\u002Fp>\n\n\n\n\u003Ch2 class=\"wp-block-heading\">\u003Cstrong>Does A.I.\nhave an Achilles heel?\u003C\u002Fstrong>\u003C\u002Fh2>\n\n\n\n\u003Cp>As the saying goes, a chain is as strong as its weakest link – and the ancient Greeks knew it. No matter that Achilles had extraordinary strength, courage, and loyalty, that he fought as the bravest in the Trojan war, as a tiny spot on his body was targeted, he was down. Recent examples show that something similar might happen in the backyard of artificial intelligence – if we are not vigilant enough and don’t do the necessary steps to counter it.\u003C\u002Fp>\n\n\n\n\u003Cp>\u003Cstrong>While in the latest years, we could hear a lot about the power and capabilities of A.I. – even to bring humanity to its knees -, recent news talk about the fact that researchers have found the Achilles heel of smart algorithms\u003C\u002Fstrong>. Remember when Stephen Hawking said that \u003Ca href=\"http:\u002F\u002Fwww.bbc.com\u002Fnews\u002Ftechnology-30290540\">the development of full artificial intelligence could spell the end of the human race\u003C\u002Fa>? Or when Elon Musk \u003Ca href=\"https:\u002F\u002Fwww.vanityfair.com\u002Fnews\u002F2017\u002F03\u002Felon-musk-billion-dollar-crusade-to-stop-ai-space-x\">told Bloomberg’s Ashlee Vance\u003C\u002Fa>, the author of the biography \u003Cem>Elon Musk\u003C\u002Fem>, that he was afraid that his friend Larry Page, a co-founder of Google and now the C.E.O. of its parent company, Alphabet, could have perfectly good intentions but still “produce something evil by accident”—including, possibly, “a fleet of artificial intelligence-enhanced robots capable of destroying mankind”? However, it’s not that easy to imagine a Matrix-like universe when carefully painted lines on the road might “capture” self-driving cars leaving them in an irresolvable loop. And that’s what is happening lately.\u003C\u002Fp>\n\n\n\n\u003Cp>\u003Cstrong>While the fears about A.I. developing capacities that could match human performance might be justified on some level – especially when looking at \u003Ca href=\"https:\u002F\u002Fnews.un.org\u002Fen\u002Fstory\u002F2019\u002F03\u002F1035381\">autonomous weapons\u003C\u002Fa> -, researchers and artists proved in several instances that artificial intelligence algorithms can be fooled, mainly through adversarial examples.\u003C\u002Fstrong>\u003C\u002Fp>\n\n\n\n\u003Cfigure class=\"wp-block-image\">\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"512\" height=\"288\" src=\"https:\u002F\u002F1712507217.rsc.cdn77.org\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FArtificial-Intelligence-512x288.jpg\" alt=\"hack A.I.\" class=\"wp-image-25054\" srcset=\"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FArtificial-Intelligence-512x288.jpg 512w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FArtificial-Intelligence-370x208.jpg 370w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FArtificial-Intelligence-768x432.jpg 768w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FArtificial-Intelligence.jpg 1068w\" sizes=\"auto, (max-width: 512px) 100vw, 512px\" \u002F>\u003Cfigcaption class=\"wp-element-caption\">Source: www.singularityhub.com\u003C\u002Ffigcaption>\u003C\u002Ffigure>\n\n\n\n\u003Ch2 class=\"wp-block-heading\">\u003Cstrong>Setting traps for self-driving cars\u003C\u002Fstrong>\u003C\u002Fh2>\n\n\n\n\u003Cp>What smart algorithms are excellent at is remembering huge chunks of data and categorizing, clustering, basically ordering information according to certain yardsticks. \u003Cstrong>That may seem like an endless pool of knowledge from certain viewpoints, but in other cases, it turns out that’s just information without context. This was the main takeaway of one of the first examples when \u003Ca href=\"https:\u002F\u002Ftechcrunch.com\u002F2017\u002F03\u002F17\u002Flaying-a-trap-for-self-driving-cars\u002F\">a performance artist laid a trap for self-driving cars\u003C\u002Fa>.\u003C\u002Fstrong> He imagined that among the first things that these cars learn are road signs – and the fact that it is never allowed to cross a solid line with a dashed one on the far side. When a self-driving car gets into a circle like this (as you are allowed to cross when the dashed line is closer to you, the vehicle can get in) – there’s no way out.\u003C\u002Fp>\n\n\n\n\u003Cp>\u003Cstrong>That’s not the only example of how embarrassingly easy it is to fool self-driving cars. In March 2019, \u003Ca href=\"https:\u002F\u002Fkeenlab.tencent.com\u002Fen\u002Fwhitepapers\u002FExperimental_Security_Research_of_Tesla_Autopilot.pdf?utm_campaign=the_algorithm.unpaid.engagement&utm_source=hs_email&utm_medium=email&utm_content=71373464&_hsenc=p2ANqtz--JBcpulYc-GW10QtUBBH_VTXHIiaAgdM-w3SdhQ1uop_m2MwFNQK8b-uDQ6hEgwH-08IpeSACOY432EYgtoku-uYAOZA&_hsmi=71373464\">Tencent’s Keen Security Lab released a report\u003C\u002Fa> where they detailed how they could trick a Tesla Model S into switching lanes and driving into oncoming traffic – by only placing three stickers on the road that formed a line\u003C\u002Fstrong>. The \u003Ca href=\"https:\u002F\u002Fwww.vox.com\u002Ffuture-perfect\u002F2019\u002F4\u002F8\u002F18297410\u002Fai-tesla-self-driving-cars-adversarial-machine-learning\">car’s autopilot system, which relies on computer vision, detected the stickers\u003C\u002Fa> and interpreted them to mean that the lane was veering left. Therefore, it steered the car that way. Now, that might have been lethal in real-world settings – but luckily that was just an experiment to assess how secure is the self-driving system against hackers. Well, it seems they still have some work to do at Tesla.\u003C\u002Fp>\n\n\n\n\u003Cfigure class=\"wp-block-image\">\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"512\" height=\"288\" src=\"https:\u002F\u002F1712507217.rsc.cdn77.org\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAutonomous-Trap-512x288.jpg\" alt=\"hack A.I.\" class=\"wp-image-25053\" srcset=\"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAutonomous-Trap-512x288.jpg 512w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAutonomous-Trap-370x208.jpg 370w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAutonomous-Trap-768x432.jpg 768w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAutonomous-Trap.jpg 1280w\" sizes=\"auto, (max-width: 512px) 100vw, 512px\" \u002F>\u003Cfigcaption class=\"wp-element-caption\">Source: www.fastcompany.com\u003C\u002Ffigcaption>\u003C\u002Ffigure>\n\n\n\n\u003Ch2 class=\"wp-block-heading\">\u003Cstrong>How to hide from facial recognition?\u003C\u002Fstrong>\u003C\u002Fh2>\n\n\n\n\u003Cp>Recently, other examples of algorithm-fooling tricks have also appeared. \u003Cstrong>A group of engineers from the University of KU Leuven in Belgium showed in a \u003Ca href=\"https:\u002F\u002Farxiv.org\u002Fpdf\u002F1904.08653.pdf\">study\u003C\u002Fa> in 2019 how simple printed patterns can fool an A.I. system that’s designed to recognize people in images\u003C\u002Fstrong>. \u003Ca href=\"https:\u002F\u002Fwww.theverge.com\u002F2019\u002F4\u002F23\u002F18512472\u002Ffool-ai-surveillance-adversarial-example-yolov2-person-detection\">As the researchers write\u003C\u002Fa>: “We believe that, if we combine this technique with a sophisticated clothing simulation, we can design a T-shirt print that can make a person virtually invisible for automatic surveillance cameras.” \u003C\u002Fp>\n\n\n\n\u003Cp>\u003Cstrong>Computer scientists regularly test deep learning\nsystems with so-called ‘adversarial examples’ crafted to make the A.I.s\nmisclassify them in order to find out the possible limitations of current deep\nlearning methods\u003C\u002Fstrong>. The T-shirt\nprint is one such example. \u003Ca href=\"https:\u002F\u002Fscience.sciencemag.org\u002Fcontent\u002F363\u002F6433\u002F1287\">Adversarial inputs were first formally described in\n2004\u003C\u002Fa> when spammers’\ntechniques were studied to circumvent spam filters. Typically, adversarial\nexamples are engineered by taking real data, such as a spam advertising\nmessage, and making intentional changes to that data designed to fool the\nalgorithm that will process it. Researchers have demonstrated the existence of\nadversarial examples for every type of machine-learning model ever studied and\nacross a wide range of data types, including images, audio, text, and other\ninputs.\u003C\u002Fp>\n\n\n\n\u003Cp>As an example, in a \u003Ca href=\"http:\u002F\u002Fwww.evolvingai.org\u002Ffooling\">research conducted by experts at the University of Wyoming\u003C\u002Fa>, scientists showed how easy it is to produce images that are completely unrecognizable to humans, but that state-of-the-art deep neural networks believe to be recognizable objects with 99.99 percent confidence. For example, they labeled white noise static to be a lion with certainty. \u003C\u002Fp>\n\n\n\n\u003Cp>In another project carried out in collaboration with Google, a research group of \u003Ca href=\"https:\u002F\u002Fpeople.eecs.berkeley.edu\u002F~dawnsong\u002F\">Dawn Song\u003C\u002Fa>, a professor at UC Berkeley, who specializes in studying the security risks involved with A.I. and machine learning, \u003Ca href=\"https:\u002F\u002Fwww.technologyreview.com\u002Fs\u002F613170\u002Femtech-digital-dawn-song-adversarial-machine-learning\u002F\">probed machine learning algorithms trained to generate automatic responses from e-mail messages\u003C\u002Fa> (in this case the \u003Ca href=\"https:\u002F\u002Fwww.cs.cmu.edu\u002F~.\u002Fenron\u002F\">Enron email data set\u003C\u002Fa>). The effort showed that by creating the right messages, it is possible to have the machine model spit out sensitive data such as credit card numbers. The findings were used by Google to prevent Smart Compose, the tool that auto-generates text in Gmail, from being exploited.\u003C\u002Fp>\n\n\n\n\u003Cfigure class=\"wp-block-image\">\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"512\" height=\"342\" src=\"https:\u002F\u002F1712507217.rsc.cdn77.org\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FFacial-recognition-512x342.jpg\" alt=\"hack A.I.\" class=\"wp-image-25055\" srcset=\"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FFacial-recognition-512x342.jpg 512w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FFacial-recognition-768x512.jpg 768w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FFacial-recognition.jpg 1000w\" sizes=\"auto, (max-width: 512px) 100vw, 512px\" \u002F>\u003Cfigcaption class=\"wp-element-caption\">Source: www.verdict.co.uk\u003C\u002Ffigcaption>\u003C\u002Ffigure>\n\n\n\n\u003Ch2 class=\"wp-block-heading\">\u003Cstrong>Adversarial examples are coming to medicine\u003C\u002Fstrong>\u003C\u002Fh2>\n\n\n\n\u003Cp>\u003Cstrong>These adversarial examples could also be used to trick\nmedical A.I. vision systems that are \u003Ca href=\"http:\u002F\u002Fscience.sciencemag.org\u002Fcontent\u002F363\u002F6433\u002F1287\">designed to identify diseases\u003C\u002Fa>\u003C\u002Fstrong>\u003Cstrong>. In that study, the researchers tested deep learning\nsystems with adversarial examples on three popular medical imaging tasks – classifying\n\u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FDiabetic_retinopathy\">diabetic retinopathy\u003C\u002Fa>\u003C\u002Fstrong>\u003Cstrong> from retinal images, \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FPneumothorax\">pneumothorax\u003C\u002Fa>\u003C\u002Fstrong>\u003Cstrong> from chest X-rays, and \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FMelanoma\">melanoma\u003C\u002Fa>\u003C\u002Fstrong>\u003Cstrong> from skin photos. In such attacks, pixels within\nimages are modified in a way that might seem like a minimal amount of noise to\nhumans but could trick these systems into classifying these images incorrectly\u003C\u002Fstrong>. The scientists noted that their attacks could make\ndeep learning systems misclassify images up to 100 percent of the time and that\nmodified images were imperceptible from real ones to the human eye. They add\nthat such attacks could work on any image and could even be incorporated\ndirectly into the image-capture process. \u003C\u002Fp>\n\n\n\n\u003Cp>Another \u003Ca href=\"https:\u002F\u002Fscience.sciencemag.org\u002Fcontent\u002F363\u002F6433\u002F1287\">study\u003C\u002Fa> published in Science about adversarial attacks on machine learning in medicine showed that if they take a mole classification algorithm for dermatology, and the software identifies a mole as benign with a >99 percent confidence, but they tweak the image a tiny bit, the algorithm can be fooled into thinking the mole is malignant with 100 percent confidence. \u003Cstrong>It’s not difficult to imagine how such tricks could be used to make insurance companies believe that the patient must be reimbursed for the treatment of a malignant mole. \u003C\u002Fstrong>\u003C\u002Fp>\n\n\n\n\u003Cfigure class=\"wp-block-image\">\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"512\" height=\"307\" src=\"https:\u002F\u002F1712507217.rsc.cdn77.org\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAdversarial-Attacks-Against-Medical-Deep-Learning-Systems-512x307.jpg\" alt=\"hack A.I.\" class=\"wp-image-25056\" srcset=\"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAdversarial-Attacks-Against-Medical-Deep-Learning-Systems-512x307.jpg 512w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAdversarial-Attacks-Against-Medical-Deep-Learning-Systems-768x461.jpg 768w, https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002FAdversarial-Attacks-Against-Medical-Deep-Learning-Systems.jpg 1024w\" sizes=\"auto, (max-width: 512px) 100vw, 512px\" \u002F>\u003Cfigcaption class=\"wp-element-caption\">Source: https:\u002F\u002Fcatalyst.nejm.org\u002Fvideos\u002Fhealth-superpowers-ai-good\u002F\u003C\u002Ffigcaption>\u003C\u002Ffigure>\n\n\n\n\u003Ch2 class=\"wp-block-heading\">\u003Cstrong>The time bomb is ticking\u003C\u002Fstrong>\u003C\u002Fh2>\n\n\n\n\u003Cp>The researchers who conducted the above study emphasized that the noise added to the image is definitely not random and has near-zero probability of occurring by chance. Thus, such \u003Cstrong>adversarial examples reflect not that machine-learning models are inaccurate or unreliable per se but rather that even otherwise effective models are susceptible to manipulation by inputs explicitly designed to fool them\u003C\u002Fstrong>.\u003C\u002Fp>\n\n\n\n\u003Cp>\u003Cstrong>And, that’s a worrying phenomenon – as hackers could easily find a way to attack artificial intelligence-based medical software through such ‘adversarial examples’ for medical fraud or for causing other types of harm on purpose\u003C\u002Fstrong>. That hasn’t happened yet – but it might only be a matter of time. Thus, we should be figuring out as soon as possible how to make smart algorithmic systems more robust in the face of attacks. And for that, it’s not enough to have A.I. experts and IT professionals, we should include medical doctors and even sociologists, artists, or philosophers who could look at the issues from a completely different perspective. These complex and diverse teams may be our best shot at staying always one step ahead of hackers with malevolent intentions.\u003C\u002Fp>\n\n\n\n\u003Ctable style=\"width: 100%; border-collapse: collapse; background-color: #eee; border-top: 4px solid #444;\" cellpadding=\"10px\">\n\u003Ctbody>\n\u003Ctr>\n\u003Ctd style=\"width: 100%;\">\n\u003Cp>\u003Cstrong>At The Medical Futurist, we are building a community for making a bold vision about the future of healthcare reality today.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Ciframe loading=\"lazy\" width=\"560\" height=\"315\" src=\"https:\u002F\u002Fwww.youtube.com\u002Fembed\u002FFIbMejImnxs\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture\" allowfullscreen=\"\">\u003C\u002Fiframe>\n\u003Cp>If you’d like to support this mission, we invite you to \u003Ca href=\"https:\u002F\u002Fwww.patreon.com\u002Fthemedicalfuturist\">join The Medical Futurist Patreon Community\u003C\u002Fa>. A community of empowered patients, future-oriented healthcare professionals, concerned health policymakers, sensible health tech developers, and enthusiastic medical students. If there were ever a time to join us, it is now. Every contribution, however big or small, powers our research and sustains our future.\u003C\u002Fp>\n\u003Cp>\u003Ca href=\"https:\u002F\u002Fwww.patreon.com\u002Fthemedicalfuturist\">\u003Cstrong style=\"background-color: #e71d3299; color: #000;\">Click here to support The Medical Futurist from as little as $3\u003C\u002Fstrong>\u003C\u002Fa> – it only takes a minute. Thank you.\u003C\u002Fp>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\n\u003C\u002Ftable>\n\u003Cp> \u003C\u002Fp>\n",false,{"rendered":22,"protected":20},"\u003Cp>While in the last years, sci-fi stories, experts and even some opinion-leader public figures frequently spelled the end of mankind through artificial intelligence, lately, it turned out just how easy it is to play tricks on smart algorithms and fool them into making errors. We looked around how researchers can hack A.I. and what that means for medicine and healthcare, mainly from a security perspective.\u003C\u002Fp>\n",6,25057,"closed",true,"","standard",{"_acf_changed":20,"footnotes":27},[31,32,33],504,521,799,[35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50],144,817,206,1228,207,1481,246,1503,271,1504,275,1505,425,1506,688,134,[],[],[54,55,56,57,58,59,60,61,62,63,64],1789,1801,1819,1883,2695,2739,3481,3539,1661,1715,1723,[66,14,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90],"post-25052","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-artificial-intelligence","category-future-medicine","category-security-privacy","tag-artificial-intelligence","tag-security","tag-digital","tag-artificial","tag-digital-health","tag-self-driving","tag-future","tag-hack","tag-health","tag-adversarial","tag-healthcare","tag-adversarial-example","tag-technology-2","tag-autonomous","tag-data-security","tag-ai",{"id":24,"alt_text":92,"caption":27,"description":93,"media_type":94,"media_details":95,"post":5,"source_url":140},"tmf_hack A.I.","TMF","image",{"width":96,"height":97,"file":98,"sizes":99,"image_meta":137},1920,1080,"2019\u002F10\u002F114_How-to-play-tricks-on-AI.png",{"medium":100,"large":106,"thumbnail":111,"medium_large":115,"1536x1536":116,"large_old_512x288":121,"thumbnail_old_150x150":126,"medium_old_370x208":128,"medium_large_old_768x432":131,"large_old_512x288_old_512x288":134},{"file":101,"width":102,"height":103,"mime-type":104,"source_url":105},"114_How-to-play-tricks-on-AI-370x208.png",370,208,"image\u002Fpng","https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002F114_How-to-play-tricks-on-AI-370x208.png",{"file":107,"width":108,"height":109,"mime-type":104,"source_url":110},"114_How-to-play-tricks-on-AI-768x432.png",768,432,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002F114_How-to-play-tricks-on-AI-768x432.png",{"file":112,"width":113,"height":113,"mime-type":104,"source_url":114},"114_How-to-play-tricks-on-AI-150x150.png",150,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002F114_How-to-play-tricks-on-AI-150x150.png",{"file":107,"width":108,"height":109,"mime-type":104,"source_url":110},{"file":117,"width":118,"height":119,"mime-type":104,"source_url":120},"114_How-to-play-tricks-on-AI-1536x864.png",1536,864,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002F114_How-to-play-tricks-on-AI-1536x864.png",{"file":122,"width":123,"height":124,"mime-type":104,"source_url":125},"114_How-to-play-tricks-on-AI-512x288.png",512,288,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002F114_How-to-play-tricks-on-AI-512x288.png",{"file":112,"width":127,"height":127,"mime-type":104,"source_url":114},"150",{"file":101,"width":129,"height":130,"mime-type":104,"source_url":105},"370","208",{"file":107,"width":132,"height":133,"mime-type":104,"source_url":110},"768","432",{"file":122,"width":135,"height":136,"mime-type":104,"source_url":125},"512","288",{"aperture":138,"credit":27,"camera":27,"caption":27,"created_timestamp":138,"copyright":27,"focal_length":138,"iso":138,"shutter_speed":138,"title":27,"orientation":138,"keywords":139},"0",[],"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F10\u002F114_How-to-play-tricks-on-AI.png",{"related_posts_footer":142,"related_posts":20,"cta_type":147,"cta_color":148,"subtitle":27,"related_books":149},[143,144,145,146],13650,14848,15909,16745,"collaboration","blue",[150],24762,{"yoast_wpseo_title":152,"yoast_wpseo_metadesc":153,"yoast_wpseo_canonical":15},"How To Play Tricks On Artificial Intelligence? - The Medical Futurist","The Medical Futurist looked around how researchers can hack A.I. and what that means for medicine and healthcare, mainly from a security perspective.",{"self":155,"collection":161,"about":164,"author":167,"replies":170,"version-history":173,"predecessor-version":177,"wp:featuredmedia":181,"wp:attachment":184,"wp:term":187,"curies":203},[156],{"href":157,"targetHints":158},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fposts\u002F25052",{"allow":159},[160],"GET",[162],{"href":163},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fposts",[165],{"href":166},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Ftypes\u002Fpost",[168],{"embeddable":26,"href":169},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fusers\u002F6",[171],{"embeddable":26,"href":172},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fcomments?post=25052",[174],{"count":175,"href":176},8,"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fposts\u002F25052\u002Frevisions",[178],{"id":179,"href":180},50255,"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fposts\u002F25052\u002Frevisions\u002F50255",[182],{"embeddable":26,"href":183},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fmedia\u002F25057",[185],{"href":186},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fmedia?parent=25052",[188,191,194,197,200],{"taxonomy":189,"embeddable":26,"href":190},"category","https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fcategories?post=25052",{"taxonomy":192,"embeddable":26,"href":193},"post_tag","https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Ftags?post=25052",{"taxonomy":195,"embeddable":26,"href":196},"project_category","https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fproject_category?post=25052",{"taxonomy":198,"embeddable":26,"href":199},"contact_email_category","https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fcontact_email_category?post=25052",{"taxonomy":201,"embeddable":26,"href":202},"yst_prominent_words","https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fyst_prominent_words?post=25052",[204],{"name":205,"href":206,"templated":26},"wp","https:\u002F\u002Fapi.w.org\u002F{rel}",[],[209],{"id":150,"date":210,"date_gmt":211,"guid":212,"modified":214,"modified_gmt":215,"slug":216,"status":13,"type":217,"link":218,"title":219,"content":221,"excerpt":223,"author":225,"featured_media":226,"comment_status":25,"ping_status":25,"template":27,"yst_prominent_words":227,"class_list":229,"better_featured_image":232,"acf":267,"yoast_meta":288,"_links":291},"2021-03-03T21:58:00","2021-03-03T20:58:00",{"rendered":213},"https:\u002F\u002Fapi.medicalfuturist.com\u002F?post_type=book&p=24762","2024-05-29T23:37:45","2024-05-29T21:37:45","a-guide-to-artificial-intelligence-in-healthcare","book","https:\u002F\u002Fapi.medicalfuturist.com\u002Fbooks\u002Fa-guide-to-artificial-intelligence-in-healthcare\u002F",{"rendered":220},"A Guide to Artificial Intelligence in Healthcare",{"rendered":222,"protected":20},"\n\u003Cp>Can we stay human in the\nage of A.I.? To go even further, can we grow in humanity, can we shape a\nmore humane, more equitable and sustainable healthcare?\u003C\u002Fp>\n\n\n\n\u003Cp>Our e-book aims to prepare\nhealthcare and medical professionals for the era of human-machine collaboration.\nRead The Medical Futurist’s guide to understanding, anticipating and\ncontrolling artificial intelligence.\u003C\u002Fp>\n",{"rendered":224,"protected":20},"\u003Cp>Can we stay human in the age of A.I.? To go even further, can we grow in humanity, can we shape a more humane, more equitable […]\u003C\u002Fp>\n",10,56335,[228,64],1683,[230,217,231,68,70,71],"post-24762","type-book",{"id":226,"alt_text":27,"caption":27,"description":27,"media_type":94,"media_details":233,"post":150,"source_url":266},{"width":234,"height":235,"file":236,"filesize":237,"sizes":238,"image_meta":264},1700,2200,"2021\u002F03\u002Fai-in-healthcare_2024-v6-1-1.png",785627,{"medium":239,"large":243,"thumbnail":248,"medium_large":252,"1536x1536":253,"2048x2048":258},{"file":240,"width":102,"height":103,"mime-type":104,"filesize":241,"source_url":242},"ai-in-healthcare_2024-v6-1-1-370x208.png",49495,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2021\u002F03\u002Fai-in-healthcare_2024-v6-1-1-370x208.png",{"file":244,"width":108,"height":245,"mime-type":104,"filesize":246,"source_url":247},"ai-in-healthcare_2024-v6-1-1-768x994.png",994,349457,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2021\u002F03\u002Fai-in-healthcare_2024-v6-1-1-768x994.png",{"file":249,"width":113,"height":113,"mime-type":104,"filesize":250,"source_url":251},"ai-in-healthcare_2024-v6-1-1-150x150.png",23049,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2021\u002F03\u002Fai-in-healthcare_2024-v6-1-1-150x150.png",{"file":244,"width":108,"height":245,"mime-type":104,"filesize":246,"source_url":247},{"file":254,"width":255,"height":118,"mime-type":104,"filesize":256,"source_url":257},"ai-in-healthcare_2024-v6-1-1-1187x1536.png",1187,849543,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2021\u002F03\u002Fai-in-healthcare_2024-v6-1-1-1187x1536.png",{"file":259,"width":260,"height":261,"mime-type":104,"filesize":262,"source_url":263},"ai-in-healthcare_2024-v6-1-1-1583x2048.png",1583,2048,1578683,"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2021\u002F03\u002Fai-in-healthcare_2024-v6-1-1-1583x2048.png",{"aperture":138,"credit":27,"camera":27,"caption":27,"created_timestamp":138,"copyright":27,"focal_length":138,"iso":138,"shutter_speed":138,"title":27,"orientation":138,"keywords":265},[],"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2021\u002F03\u002Fai-in-healthcare_2024-v6-1-1.png",{"leanpub_url":268,"preview":269,"buy_button_text":287},"https:\u002F\u002Fleanpub.com\u002FArtificialIntelligenceinHealthcare",[270,271,273,275,277,279,281,283,285],{"image":266},{"image":272},"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F09\u002FArtificialIntelligenceinHealthcare-3-1.png",{"image":274},"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F09\u002FArtificialIntelligenceinHealthcare-4-1.png",{"image":276},"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F09\u002FArtificialIntelligenceinHealthcare-5-1.png",{"image":278},"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F09\u002FArtificialIntelligenceinHealthcare-6-1.png",{"image":280},"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F09\u002FArtificialIntelligenceinHealthcare-7-1.png",{"image":282},"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F09\u002FArtificialIntelligenceinHealthcare-8-1.png",{"image":284},"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F09\u002FArtificialIntelligenceinHealthcare-9-1.png",{"image":286},"https:\u002F\u002Fcdn.medicalfuturist.com\u002Fwp-content\u002Fuploads\u002F2019\u002F09\u002FArtificialIntelligenceinHealthcare-10-1.png","Get it on Leanpub",{"yoast_wpseo_title":289,"yoast_wpseo_metadesc":290,"yoast_wpseo_canonical":218},"A Guide to Artificial Intelligence in Healthcare - The Medical Futurist","The Guide To Artificial Intelligence In Healthcare aims to prepare healthcare and medical professionals for the era of human-machine collaboration.",{"self":292,"collection":297,"about":300,"author":303,"replies":306,"wp:featuredmedia":309,"wp:attachment":312,"wp:term":315,"curies":318},[293],{"href":294,"targetHints":295},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fbook\u002F24762",{"allow":296},[160],[298],{"href":299},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fbook",[301],{"href":302},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Ftypes\u002Fbook",[304],{"embeddable":26,"href":305},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fusers\u002F10",[307],{"embeddable":26,"href":308},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fcomments?post=24762",[310],{"embeddable":26,"href":311},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fmedia\u002F56335",[313],{"href":314},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fmedia?parent=24762",[316],{"taxonomy":201,"embeddable":26,"href":317},"https:\u002F\u002Fapi.medicalfuturist.com\u002Fwp-json\u002Fwp\u002Fv2\u002Fyst_prominent_words?post=24762",[319],{"name":205,"href":206,"templated":26},1790184587418]