{"id":14137,"date":"2018-01-23T16:24:11","date_gmt":"2018-01-23T15:24:11","guid":{"rendered":"https:\/\/timoelliott.com\/blog\/?p=14137"},"modified":"2018-01-23T16:24:11","modified_gmt":"2018-01-23T15:24:11","slug":"whats-real-artificial-intelligence","status":"publish","type":"post","link":"https:\/\/timoelliott.com\/blog\/2018\/01\/whats-real-artificial-intelligence.html","title":{"rendered":"What&#8217;s &#8220;Real&#8221; Artificial Intelligence?"},"content":{"rendered":"<p><a href=\"https:\/\/i0.wp.com\/timoelliott.com\/blog\/wp-content\/uploads\/2018\/01\/clown-unsplash.jpg?ssl=1\"><img data-recalc-dims=\"1\" loading=\"lazy\" decoding=\"async\" class=\"alignnone size-medium wp-image-14140\" src=\"https:\/\/i0.wp.com\/timoelliott.com\/blog\/wp-content\/uploads\/2018\/01\/clown-unsplash.jpg?resize=608%2C399&#038;ssl=1\" alt=\"\" width=\"608\" height=\"399\" srcset=\"https:\/\/i0.wp.com\/timoelliott.com\/blog\/wp-content\/uploads\/2018\/01\/clown-unsplash.jpg?resize=608%2C399&amp;ssl=1 608w, https:\/\/i0.wp.com\/timoelliott.com\/blog\/wp-content\/uploads\/2018\/01\/clown-unsplash.jpg?resize=768%2C504&amp;ssl=1 768w, https:\/\/i0.wp.com\/timoelliott.com\/blog\/wp-content\/uploads\/2018\/01\/clown-unsplash.jpg?w=800&amp;ssl=1 800w\" sizes=\"auto, (max-width: 608px) 100vw, 608px\" \/><\/a><\/p>\n<p>With all the current hype around artificial intelligence (AI), I was asked the other day how somebody could differentiate between &#8220;real&#8221; AI and fake AI. Ultimately, it&#8217;s probably not a very useful question, but here goes:<\/p>\n<p>First, artificial intelligence isn&#8217;t a technology, it&#8217;s a \u201csociotechnical construct\u201d \u2014 it\u2019s used to refer to any use of computers to do things that previously only computers to do<\/p>\n<p>There&#8217;s a rough consensus for <a href=\"https:\/\/bdtechtalks.com\/2017\/05\/12\/what-is-narrow-general-and-super-artificial-intelligence\/\" target=\"_blank\" rel=\"noopener noreferrer\">different types of AI<\/a>:<\/p>\n<ul>\n<li>\u201cWeak\u201d or \u201cnarrow\u201d AI refers to niche uses of technology for a single defined purpose. It is the only form of AI actually available today.<\/li>\n<li>\u201cStrong\u201d AI would be able to address a wide range of different tasks and approach human reasoning.<\/li>\n<li>\u201cSuper\u201d AI would be (will be?!) when computers have higher general intelligence than humans<\/li>\n<\/ul>\n<p>So\u00a0you may hear\u00a0things like \u201cthat\u2019s not real AI\u201d when the example concerns only narrow AI. For example, there&#8217;s been some <a href=\"http:\/\/www.rogerschank.com\/fraudulent-claims-made-by-IBM-about-Watson-and-AI\" target=\"_blank\" rel=\"noopener noreferrer\">pushback<\/a>\u00a0on\u00a0describing solutions as &#8220;cognitive&#8221;\u00a0that don&#8217;t really live up to the term:<\/p>\n<blockquote><p>They are not doing &#8220;<a href=\"http:\/\/www.rogerschank.com\/fraudulent-claims-made-by-IBM-about-Watson-and-AI\" target=\"_blank\" rel=\"noopener noreferrer\">cognitive computing<\/a>&#8221; no matter how many times they say they are &#8212; <a href=\"http:\/\/www.rogerschank.com\/things-I-have-done\" target=\"_blank\" rel=\"noopener noreferrer\">Roger Shank, AI researcher<\/a><\/p><\/blockquote>\n<p>The main technology associated with the recent surge in AI buzz is <a href=\"https:\/\/en.wikipedia.org\/wiki\/Machine_learning\" target=\"_blank\" rel=\"noopener noreferrer\">machine learning<\/a>\u00a0(ML).\u00a0Inevitably, there are different definitions of the term available, but the consensus\u00a0is that it refers to any algorithm that automatically updates itself based on the data. This means it includes\u00a0<a href=\"http:\/\/a248.g.akamai.net\/n\/248\/420835\/91770f24e9570d450e5107ed43e880a3d476b2cdbc65b5cc33afbc6c8debce39\/sapasset.download.akamai.com\/420835\/sapcom\/docs\/2016\/02\/c4d70a6c-617c-0010-82c7-eda71af511fa.pdf\" target=\"_blank\" rel=\"noopener noreferrer\">automated forms of traditional statistical methods<\/a> as well as neural network and \u201c<a href=\"https:\/\/dam.sap.com\/mac\/preview\/a\/67\/XwXJUvyxXXOlxAJgmgwxmOA7lPHJywAUSPSIUEOSPIXEXOyp\/38002_WP_Deep_Learning_USen_final.htm\" target=\"_blank\" rel=\"noopener noreferrer\">deep learning<\/a>\u201d methods.<\/p>\n<p><a href=\"https:\/\/i0.wp.com\/timoelliott.com\/blog\/wp-content\/uploads\/2018\/01\/Mark_I_perceptron.jpg?ssl=1\"><img data-recalc-dims=\"1\" loading=\"lazy\" decoding=\"async\" class=\"wp-image-14138 alignright\" src=\"https:\/\/i0.wp.com\/timoelliott.com\/blog\/wp-content\/uploads\/2018\/01\/Mark_I_perceptron.jpg?resize=147%2C181&#038;ssl=1\" alt=\"\" width=\"147\" height=\"181\" \/><\/a>Neural networks <a href=\"https:\/\/en.wikipedia.org\/wiki\/Perceptron\" target=\"_blank\" rel=\"noopener noreferrer\">date back to the 1950s<\/a>, but the rise in computing power and the quantity of data available means that they can be made more powerful and sophisticated (e.g. <a href=\"https:\/\/en.wikipedia.org\/wiki\/Generative_adversarial_network\" target=\"_blank\" rel=\"noopener noreferrer\">generative adversarial networks<\/a>, where different neural networks compete against each other, strengthening the quality of the overall algorithm)<\/p>\n<p>Ultimately, the biggest factor in whether ML works well or not is typically not the choice of algorithm, but the quantity and quality of data available.<\/p>\n<p>The &#8220;real&#8221; opportunity today is weak AI, which despite the name is poised to make a very big difference to <a href=\"https:\/\/timoelliott.com\/blog\/2017\/06\/the-future-of-digital-business-self-improving-products.html\" target=\"_blank\" rel=\"noopener noreferrer\">almost all aspects of modern business<\/a>, by automating complex-but-repetitive processes and allowing more intuitive interactions with data.<\/p>\n<p>For example, today, there are armies of people in shared finance centers whose job is to match invoices sent to customers with the corresponding bank payments. Today, it&#8217;s typically a messy process, with\u00a0as few as 40% matched automatically based on amounts, reference numbers etc. The remaining 60% have to be treated manually, because there are two invoices for one payment, or the amounts are slightly different, etc. But using ML, <a href=\"https:\/\/events.sap.com\/sapandasug\/en\/session\/32268\" target=\"_blank\" rel=\"noopener noreferrer\">organizations have seen this rise quickly to 96% matching and beyond <\/a>\u2014 a huge savings in time and effort.\u00a0And that\u2019s just one example of the hundreds of different business processes that will be automated using ML in the years to come.<\/p>\n<p>And new <a href=\"https:\/\/blogs.sap.com\/2017\/05\/10\/sap-copilot-1705-now-available-to-customers\/\" target=\"_blank\" rel=\"noopener noreferrer\">enterprise digital assistants<\/a> are allowing business people to work with their corporate systems through chat and voice, as easily as they ask a question to Siri, Cortana, or Alexa.<\/p>\n<p>Compared to\u00a0strong AI\u00a0that would be able to figure out what\u2019s wrong with you better than a doctor, these types of application may seem tame,\u00a0but <strong>the\u00a0big exciting opportunity for AI is the boring stuff<\/strong> \u2014 these\u00a0applications are very real today, while the more glamorous use cases are not!<\/p>\n<p><em>Photo by\u00a0<a href=\"https:\/\/unsplash.com\/photos\/f46l9gZJLqs?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText\">Diana Feil<\/a>\u00a0on\u00a0<a href=\"https:\/\/unsplash.com\/?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText\">Unsplash<\/a><\/em><\/p>\n","protected":false},"excerpt":{"rendered":"<p>How can you tell the difference between real and fake artificial intelligence? <\/p>\n","protected":false},"author":2,"featured_media":14140,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"jetpack_post_was_ever_published":false,"_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}},"categories":[14],"tags":[1177,1176,1341,1175,1340],"class_list":["post-14137","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-thoughts","tag-ai","tag-artificial-intelligence","tag-fake-ai","tag-machine-learning","tag-real-ai"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"https:\/\/i0.wp.com\/timoelliott.com\/blog\/wp-content\/uploads\/2018\/01\/clown-unsplash.jpg?fit=800%2C525&ssl=1","jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/p3X9RF-3G1","_links":{"self":[{"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/posts\/14137","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/comments?post=14137"}],"version-history":[{"count":0,"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/posts\/14137\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/media\/14140"}],"wp:attachment":[{"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/media?parent=14137"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/categories?post=14137"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/timoelliott.com\/blog\/wp-json\/wp\/v2\/tags?post=14137"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}