{"id":855,"date":"2016-08-24T01:53:08","date_gmt":"2016-08-24T00:53:08","guid":{"rendered":"http:\/\/eden-study.org\/?page_id=855"},"modified":"2022-08-11T14:51:10","modified_gmt":"2022-08-11T13:51:10","slug":"adviser","status":"publish","type":"page","link":"https:\/\/eden-study.org\/?page_id=855","title":{"rendered":"Advice"},"content":{"rendered":"<p>[vc_row][vc_column][vc_column_text]<\/p>\n<h1>Advice<\/h1>\n<p>Dr Eden provides expert advice on strategic decisions which require both of scientifically-informed academic research and hands-on experience. His clients included policy makers, hedge funds, venture capital investors, corporations, start-ups, and charities (<a href=\"https:\/\/eden-study.org\/?page_id=38\">details<\/a>).<\/p>\n<p>Advice is offered in three areas:<\/p>\n<ul>\n<li><a href=\"#advise-ml\">Machine learning<\/a><\/li>\n<li><a href=\"https:\/\/eden-study.org\/?page_id=855#advise-sen\">Software development<\/a><\/li>\n<li><a href=\"#advise-forecast\">Technological forecasting<\/a><\/li>\n<\/ul>\n<p><a href=\"https:\/\/eden-study.org\/?page_id=38\">See also: Professional\u00a0 experience<\/a>[\/vc_column_text][vc_column_text]<\/p>\n<blockquote><p>The difference between theory and practice is greater in practice than in theory<\/p>\n<footer><cite>\u2014 Origin unknown<\/cite><\/footer>\n<\/blockquote>\n<p>[\/vc_column_text][\/vc_column][\/vc_row][vc_row][vc_column][vc_tta_tabs style=\u201doutline\u201d color=\u201dmulled-wine\u201d spacing=\u201d5\u2033 active_section=\u201d1\u2033 pagination_style=\u201doutline-square\u201d css=\u201d.vc_custom_1475083268948{border-radius: 4px !important;}\u201d][vc_tta_section title=\u201dTechnological Forecasting\u201d tab_id=\u201dadvise-forecast\u201d][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dtop\u201d][vc_column_inner width=\u201d1\/2\u2033][vc_column_text]Experience gained with \u2014<\/p>\n<ul class=\"ul-boxed list-unstyled\">\n<li>Tracking, measuring & forecasting machine intelligence<\/li>\n<li>Training in Effective Forecasting (<a href=\"http:\/\/www.goodjudgment.com\/\">Good Judgement Project <\/a>)<\/li>\n<\/ul>\n<p>[\/vc_column_text][\/vc_column_inner][vc_column_inner width=\u201d1\/2\u2033][vc_column_text]Expertise offered\u00a0in \u2014<\/p>\n<ul class=\"ul-boxed list-unstyled\">\n<li>Future of artificial intelligence<\/li>\n<li>Future of algorithmic trading<\/li>\n<li>Cryptocurrency & Smart Contract evolution<\/li>\n<\/ul>\n<p>[\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_row_inner][vc_column_inner][vc_cta h2=\u201dQuestions\u201d]<img loading=\"lazy\" decoding=\"async\" class=\"wp-image-2012 alignright\" style=\"padding-left: 12px;\" src=\"https:\/\/eden-study.org\/wp-content\/uploads\/2016\/08\/Future-Vision-e1475776547578-384x193.jpg\" width=\"215\" height=\"103\" \/><\/p>\n<p>Which investment algorithms will <a href=\"https:\/\/www.wired.com\/2016\/01\/the-rise-of-the-artificially-intelligent-hedge-fund\/\">hedge funds<\/a>\u00a0use next decade? How will AI affect computer trading, now that <a href=\"http:\/\/www.bloomberg.com\/news\/articles\/2016-01-26\/high-speed-firms-now-oversee-almost-all-stocks-at-nyse-floor\">almost all stocks are traded by algorithms<\/a>? How will driverless cars affect transportation? Will AI\u00a0<a href=\"http:\/\/www.nytimes.com\/2016\/02\/28\/magazine\/the-robots-are-coming-for-wall-street.html\">displace 40% of jobs<\/a> within 20 years? When will <a href=\"https:\/\/www.researchgate.net\/publication\/272521896_How_predictable_is_technological_progress\">solar energy<\/a> be cheaper than fossil fuel, if at all?<\/p>\n<p>Making reliable technological forecasts seems to require a crystal\u00a0ball.[\/vc_cta][\/vc_column_inner][\/vc_row_inner][vc_column_text]<\/p>\n<blockquote><p>Organizations\u00a0spend\u00a0staggering amounts of time and money trying to predict the future but no time or money measuring their accuracy or improving on their ability to do it<\/p>\n<footer><cite>\u2014\u00a0<a href=\"https:\/\/www.wired.com\/2011\/07\/overcoming-our-aversion-to-acknowledging-our-ignorance\/\">Philip Tetlock<\/a><\/cite><\/footer>\n<\/blockquote>\n<p>[\/vc_column_text][vc_row_inner][vc_column_inner width=\u201d1\/2\u2033][vc_cta h2=\u201dProphecy was given to the fools\u2026\u201d]Who expected\u00a0the\u00a0World Wide Web\u00a0and mobile phones to take the 1990s by storm? Or BigDog, cryptocurrencies, and\u00a0<a href=\"http:\/\/www.bloomberg.com\/news\/articles\/2016-01-26\/high-speed-firms-now-oversee-almost-all-stocks-at-nyse-floor\">computer trading overtaking stock exchanges<\/a> less than a decade later? Disruptive technology emerges\u00a0at a breakneck speed as paradigm shifts accelerate. As for experts, when they\u2019re pushed to make accurate forecasts they perform worse\u00a0on average than simply expecting the future to repeat the past, according to the\u00a0Good Judgement Project. [vc_single_image image=\u201d1882\u2033 img_size=\u201dfull\u201d add_caption=\u201dyes\u201d style=\u201dvc_box_shadow_3d\u201d onclick=\u201dlink_image\u201d][\/vc_cta][\/vc_column_inner][vc_column_inner width=\u201d1\/2\u2033][vc_cta h2=\u201d\u2026 but technological forecasting is a science\u201d]Forecasting has been disrupted during the last decade or so by four discoveries:<\/p>\n<ul class=\"ul-boxed list-unstyled\">\n<li>Big\u00a0Data<br \/>\n<h6>Recording\u00a0very large, statistically-significant #data points<\/h6>\n<\/li>\n<li><a href=\"\/adviser\/#advise-ml\">Machine learning<\/a><br \/>\n<h6><a href=\"https:\/\/eden-study.org\/research\/#research-ai\">Artificial intelligence<\/a>\u00a0systems with\u00a0human capabilities at superspeed<\/h6>\n<\/li>\n<li>Universal Laws<br \/>\n<h6>Stable trends across continents & centuries<\/h6>\n<\/li>\n<li><a href=\"http:\/\/www.penguinrandomhouse.com\/books\/227815\/superforecasting-by-philip-e-tetlock-and-dan-gardner\/9780804136716\/\">Superforecasting<\/a><br \/>\n<h6>Experiments with accurate forecasting (the Good Judgement Project)<\/h6>\n<\/li>\n<\/ul>\n<p>[\/vc_cta][\/vc_column_inner][\/vc_row_inner][vc_row_inner][vc_column_inner][vc_cta h2=\u201dUniversal trends\u201d]<a href=\"http:\/\/pcdb.santafe.edu\/\">Comprehensive analysis<\/a> of millions of data points collected from <span class=\"fontstyle0\">research articles, government\u00a0reports, and market research<\/span>\u00a0spanning decades\u00a0and continents\u00a0shows that\u00a0technology progresses relatively uniformly. Wars, recessions, and natural calamities have short-term effects\u00a0and the trends quickly stabilize,\u00a0as the graphs below (and many others) reveal. That, among others, allows reliable\u00a0inference.[\/vc_cta][vc_media_grid item=\u201d1931\u2033 grid_id=\u201dvc_gid:1660225268824-1bf33c64-3161-10\u2033 include=\u201d1720,962,1926\u2033][vc_column_text]<\/p>\n<blockquote><p>The further backward you look, the further forward you can see<\/p>\n<footer><cite>\u2014\u00a0Winston Churchill<\/cite><\/footer>\n<\/blockquote>\n<p>[\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_row_inner][vc_column_inner width=\u201d1\/2\u2033][vc_cta h2=\u201dForecasting with AI\u201d]After beating the experts in predicting the oscars, college football, and the Stanley Cup, <em>Unanimous AI<\/em>\u00a0predicted the results of the Kentucky Derby at odds of 540:1 using\u00a0<em>swarm<\/em> <em>intelligence<\/em>\u00a0which combines predictions from individual experts\u00a0whose predictions were only\u00a023% accurate using AI, thereby reaching impressive accuracy \u2014 100% in this example. In many domains (such as energy and stock prices) predictions with high reliability (72% and above) can be made using \u2018big data\u2019, data mining, and predictive modelling techniques, ranging from simple statistical and regression methods to more sophisticated <a href=\"https:\/\/eden-study.org\/adviser\/#advise-ml\">machine learning<\/a> inference algorithms.[vc_single_image image=\u201d2033\u2033 img_size=\u201dfull\u201d][\/vc_cta][\/vc_column_inner][vc_column_inner width=\u201d1\/2\u2033][vc_cta h2=\u201dHyped \u2260 Worthless\u201d h4=\u201dEven productive technologies get overhyped\u201d][vc_single_image image=\u201d960\u2033 img_size=\u201dfull\u201d add_cption=\u201dyes\u201d onclick=\u201dlink_image\u201d]Space travel,\u00a0submarines, smartphones, and artificial intelligence\u00a0were foreseen\u00a0centuries\u00a0ahead. Once a\u00a0solid proof of concept is demonstrated early expectations are inflated by fantasy literature\u00a0and pop\u00a0images years and decades\u00a0before the technology matures.<\/p>\n<p>Compare for example the state of Enterprise 3D Printing in 2016 according to Gartner\u2019s Hype Cycle (above), which is approaching the Plateau of Productivity, with Machine Learning which is at the Peak of Inflated Expectations. Both technologies are revolutionary. Unrealistic\u00a0expectations do\u00a0not prove that either technology is\u00a0worthless.[\/vc_cta][\/vc_column_inner][\/vc_row_inner][vc_column_text]<\/p>\n<blockquote><p>The future is already here \u2014 it\u2019s just not evenly distributed<\/p>\n<footer><cite> \u2014 Attributed to William Gibson<\/cite><\/footer>\n<\/blockquote>\n<p>[\/vc_column_text][\/vc_tta_section][vc_tta_section title=\u201dMachine Learning\u201d tab_id=\u201dadvise-ml\u201d][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dtop\u201d][vc_column_inner width=\u201d1\/2\u2033][vc_column_text]Experience gained with \u2014<\/p>\n<ul class=\"ul-boxed list-unstyled\">\n<li>Learning \u2018hard\u2019 (imprecisely defined) classes<br \/>\n<h6>E.g. image\/speech recognition<\/h6>\n<\/li>\n<li>Implementing machine learning\u00a0systems<\/li>\n<li>Recommender systems<\/li>\n<\/ul>\n<p>[\/vc_column_text][\/vc_column_inner][vc_column_inner width=\u201d1\/2\u2033][vc_column_text]Expertise offered\u00a0in \u2014<\/p>\n<ul class=\"ul-boxed list-unstyled\">\n<li class=\"\">Computer trading (algotrading)<\/li>\n<li>Choosing machine learning algorithms<br \/>\n<h6>Selecting most suitable [un]supervised learning<\/h6>\n<\/li>\n<li>Symbolic\u00a0and nonsymbolic\u00a0learning<br \/>\n<h6>Symbolic: reasoning with 1st-order & temporal logic<\/h6>\n<h6>Non-symbolic: nearest-neighbour, neural nets, genetic algorithms\u2026<\/h6>\n<\/li>\n<\/ul>\n<p>[\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_row_inner][vc_column_inner][vc_column_text]Machine learning \u2014 a field of expertise <a href=\"https:\/\/eden-study.org\/pubs\/msc\/\">since 1991 <\/a>\u2014 is replacing humans in\u00a0investment planning, fraud detection, personal assistants, recommender systems, speech and image recognition, and many other\u00a0tasks. Early\u00a0classifiers such as decision trees and bayesian nets were superseded by deep learning neural nets \u2014 multi-layered networks that gradually build an \u2018ontological hierarchy\u2019 them much more complex reasoning.<\/p>\n<p>Genetic algorithms, instance-based learning, or neural nets? Programmers today can use machine learning libraries,\u00a0each suitable for different tasks:[\/vc_column_text][vc_single_image image=\u201d840\u2033 img_size=\u201dfull\u201d add_caption=\u201dyes\u201d style=\u201dvc_box_shadow_3d\u201d onclick=\u201dlink_image\u201d][vc_column_text]Our machine learning research started in 1990 in asking\u00a0<a href=\"\/pubs\/msc\/\">how humans classify everyday objects quickly and without any formal criteria<\/a>\u00a0and whether a learning algorithm can perform\u00a0as well as a human. Since then the variety of available neural network types has exploded.[\/vc_column_text][vc_single_image image=\u201d1703\u2033 img_size=\u201dfull\u201d add_caption=\u201dyes\u201d style=\u201dvc_box_shadow_3d\u201d onclick=\u201dlink_image\u201d][\/vc_column_inner][\/vc_row_inner][\/vc_tta_section][vc_tta_section title=\u201dSoftware Design\u201d tab_id=\u201dadvise-sen\u201d][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dtop\u201d][vc_column_inner width=\u201d1\/2\u2033][vc_column_text]Experience gained with \u2014<\/p>\n<ul class=\"ul-boxed list-unstyled\">\n<li>Object-oriented programming: planning, staff training, and migrating legacy\u00a0code (re-engineering)<\/li>\n<li>Design & architecture of software platforms, infrastructure & applications: planning, evolving, and implementing<\/li>\n<\/ul>\n<p>See also:<\/p>\n<ul>\n<li><a href=\"\/bio\/prof\/\">Professional record<\/a><\/li>\n<li><a href=\"\/research\/#research-sen\">Software engineering research record<\/a><\/li>\n<\/ul>\n<p>[\/vc_column_text][\/vc_column_inner][vc_column_inner width=\u201d1\/2\u2033][vc_column_text]Expertise offered\u00a0in \u2014<\/p>\n<ul class=\"ul-boxed list-unstyled\">\n<li>Software design & architecture<\/li>\n<li>Re-engineering & evolution of legacy software<\/li>\n<li>Software visualization & reverse engineering<\/li>\n<li>Software quality<\/li>\n<li>Staff training for technology migration<\/li>\n<\/ul>\n<p>[\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_column_text]<\/p>\n<blockquote><p>There are two ways of constructing a software design. One way is to make it so simple that there are obviously no deficiencies. The other way is to make it so complicated that there are no obvious deficiencies.<\/p>\n<footer><cite>\u2014 C.A.R. Hoare (Turing Award Lecture)<\/cite><\/footer>\n<\/blockquote>\n<p>[\/vc_column_text][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dmiddle\u201d][vc_column_inner width=\u201d5\/12\u2033][vc_cta h2=\u201dFailure is common\u201d]It is estimated that up to 65% of large software projects fail disastrously. In 2013 the BBC wrote off \u00a3100 on an overambitious\u00a0IT project.\u00a0In 2016 an incorrectly configured software update\u00a0destroyed the $286m Japanese satellite Hitomi, and the Scottish Police Authority scrapped \u00a360m on an IT\u00a0project because of \u201cinsurmountable flaws\u201d.\u00a0Also in 2016 the NHS revealed that its IT call handling and IT project is four years late and \u00a340m over budget.[\/vc_cta][\/vc_column_inner][vc_column_inner width=\u201d7\/12\u2033][vc_line_chart x_values=\u201dBBC; NHS; Hitomi; SPA\u201d values=\u201d%5B%7B%22title%22%3A%22Cost%20of%20write-off%20(%C2%A3m)%22%2C%22y_values%22%3A%22100%3B%2040%3B%20200%3B%2060%22%2C%22color%22%3A%22mulled-wine%22%7D%5D\u201d][\/vc_column_inner][\/vc_row_inner][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dmiddle\u201d][vc_column_inner width=\u201d3\/12\u2033][vc_single_image image=\u201d1665\u2033 img_size=\u201dfull\u201d add_caption=\u201dyes\u201d][\/vc_column_inner][vc_column_inner width=\u201d9\/12\u2033][vc_cta h2=\u201dPatching only goes so far\u2026\u201d]Failures occur not common for lack of skill, but\u00a0because\u00a0<a href=\"#advise-forecast\">technology changes at a breakneck speed;<\/a>\u00a0hardware improves exponentially;\u00a0software libraries, environments, and even programming languages are replaced before reaching maturity; and because markets dictate unrealistic deadlines. In this climate design and reengineering\u00a0seem\u00a0a luxury and there is only time for \u2018firefighting\u2019. Before you know it, your software is increasingly a\u00a0hopeless mess.[\/vc_cta][\/vc_column_inner][\/vc_row_inner][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dmiddle\u201d][vc_column_inner width=\u201d1\/2\u2033][vc_cta h2=\u201dResearch-informed solutions\u201d]Decades of studying government, health, open-source, and financial projects have taught us that successful projects have dedicated effort to reduce complexity. How? Unfortunately, there is <a href=\"https:\/\/en.wikipedia.org\/wiki\/No_Silver_Bullet\">no silver bullet<\/a> (yet). In reality, each project has its own needs. A\u00a0stable\u00a0<a href=\"\/pubtype\/software-architecture\/\">architecture<\/a>\u00a0can emerge from research-informed\u00a0application of\u00a0cutting-edge\u00a0software <a href=\"\/pubtype\/tools\/\">tools <\/a>and\u00a0<a href=\"\/pubtype\/modelling\/\">modelling<\/a>\u00a0techniques (not only UML). Bottlenecks can be diagnosed using <a href=\"\/software-visualization\/\">software visualization<\/a> and analysis techniques for reengineering and <a href=\"\/pubtype\/software-evolution\/\">evolution<\/a>. <a href=\"\/research\/#research-flexibility\">Flexibility<\/a> can be built in using techniques such as <a href=\"\/pubtype\/design-patterns\/\">design patterns<\/a>. And the performance of critical functions such as security can be enforced using industrial\u00a0<a href=\"\/pubtype\/formal-methods\/\">formal methods<\/a>. The\u00a0knowledge required comes from\u00a0experimenting with research-informed <a href=\"\/research\/#research-sen\">software engineering<\/a>\u00a0tools and practices and keeping up with the scientific literature. This is where we could help.[\/vc_cta][\/vc_column_inner][vc_column_inner width=\u201d1\/2\u2033][vc_single_image image=\u201d1396\u2033 img_size=\u201dfull\u201d add_caption=\u201dyes\u201d onclick=\u201dlink_image\u201d][vc_single_image image=\u201d1641\u2033 img_size=\u201dfull\u201d add_caption=\u201dyes\u201d onclick=\u201dlink_image\u201d][\/vc_column_inner][\/vc_row_inner][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dmiddle\u201d][vc_column_inner][vc_column_text]<\/p>\n<blockquote><p>He who loves practice without theory is like the sailor who boards ship without a rudder and compass and never knows where he may cast.<\/p>\n<footer><cite>\u2014 Leonardo Da Vinci<\/cite><\/footer>\n<\/blockquote>\n<p>[\/vc_column_text][\/vc_column_inner][\/vc_row_inner][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dmiddle\u201d][vc_column_inner width=\u201d1\/3\u2033][vc_single_image image=\u201d922\u2033 img_size=\u201dfull\u201d add_caption=\u201dyes\u201d style=\u201dvc_box_shadow_3d\u201d onclick=\u201dlink_image\u201d][\/vc_column_inner][vc_column_inner width=\u201d2\/3\u2033][vc_cta h2=\u201dReports are useless\u201d]There is little point in\u00a0writing\u00a0lengthy documents. Nobody reads them \u2014 usually they\u2019re obsolete anyway.<\/p>\n<p>But some documentation must record clearly and ambiguously the following:\u00a0What exactly can the user expect from the application\u00a0you\u2019re developing? Are these expectations realistic? Which changes in the requirements can be expected within 6 months from release? Within two or five years? Which <a href=\"\/pubs\/2005hicss\/\">strategic\u00a0design<\/a> decisions\u00a0will ensure smooth transition to\u00a0the next version? How should\u00a0programmers <a href=\"\/pubtype\/modelling\/\">see<\/a> and understand these decisions? How can we ensure that future changes will not violate these decisions accidentally?[\/vc_cta][\/vc_column_inner][\/vc_row_inner][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dmiddle\u201d][vc_column_inner width=\u201d5\/12\u2033][vc_cta h2=\u201dLet\u2019s Communicate. <em>Precisely<\/em>.\u201d]Instead of lengthy descriptions we recommend using one or more\u00a0experimental\u00a0tools and methodologies that communicate\u00a0the\u00a0answers explicitly, visually, and precisely. They may not easy to adopt but small investments can yield very large returns. Alternative forms of specifications can describe software function and structure clearly and precisely, and automated verification can effectively prevent the majority of bugs.[\/vc_cta][\/vc_column_inner][vc_column_inner width=\u201d7\/12\u2033][vc_single_image image=\u201d1164\u2033 img_size=\u201dfull\u201d add_caption=\u201dyes\u201d style=\u201dvc_box_shadow_3d\u201d onclick=\u201dlink_image\u201d][\/vc_column_inner][\/vc_row_inner][vc_row_inner equal_height=\u201dyes\u201d content_placement=\u201dmiddle\u201d][vc_column_inner][vc_column_text]<\/p>\n<blockquote><p>The design of computing systems can only properly succeed if it is well grounded in theory, and the important concepts in a theory can only emerge through protracted exposure to application.<\/p>\n<footer><cite>\u2014 Robin Milner (1986)<\/cite><\/footer>\n<\/blockquote>\n<p>[\/vc_column_text][\/vc_column_inner][\/vc_row_inner][\/vc_tta_section][\/vc_tta_tabs][\/vc_column][\/vc_row]<\/p>\n","protected":false},"excerpt":{"rendered":"<p>[vc_row][vc_column][vc_column_text] Advice Dr Eden provides expert advice on strategic decisions which require both of scientifically-informed academic research and hands-on experience. [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":1,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-855","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/eden-study.org\/index.php?rest_route=\/wp\/v2\/pages\/855","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/eden-study.org\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/eden-study.org\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/eden-study.org\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/eden-study.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=855"}],"version-history":[{"count":26,"href":"https:\/\/eden-study.org\/index.php?rest_route=\/wp\/v2\/pages\/855\/revisions"}],"predecessor-version":[{"id":2297,"href":"https:\/\/eden-study.org\/index.php?rest_route=\/wp\/v2\/pages\/855\/revisions\/2297"}],"wp:attachment":[{"href":"https:\/\/eden-study.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=855"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}