AI REPORT GOLDMAN SACHS FT-Artificial-Intelligence | Artificial Intelligence | Robot

Please download to get full document.

View again

of 24
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report



Views: 8 | Pages: 24

Extension: PDF | Download: 0

Related documents
Where the Artificial Inteligence will be used in the near future
    The Goldman Sachs Group, Inc. Goldman Sachs Global Investment Research The real consequences of artificial intelligence From the editor:  In this edition we cogitate on the increasing transition of artificial intelligence into the mainstream, commercial world and its potentially disruptive implications and uses. We interview three experts on this subject and include our global analysts’ views. The substantial progress made over the last decade in the capabilities and cost of parallel computing, algorithms, big data and the move to the cloud is set to bring artificial intelligence out of labs and into the real, mainstream world. This has implications for information-intensive sectors as well as businesses that rely on highly skilled personnel. AI is already driving changes in advertising (programmatic ad buying), parts of retail (customised recommendations) and investing. With other sectors becoming increasingly data-intensive we see its tentacles reaching healthcare and manufacturing as well as logistics and energy consumption. The ability to think and learn enables AI-aided technologies to constantly improve and refine and when applied to decision making in businesses we believe this can lead to better cost and capital allocation, lower error rates and accelerated innovation. AI is aimed at augmenting human decision making, but in many areas it could replace humans or resolve the shortage of human skills (think collaborative robots in warehouses and hospitals).  AI can also instigate disruption in industries much like the advent of the internet did. While we do not foresee the indiscriminate demise of industry incumbents, we expect  AI to act as a necessary stay-in-the-game cost to incumbents and also open the door to capital and labor-light new entrants. However AI can reinforce dominance in industry leaders that have a pre-existing edge in terms of access to proprietary data.  Apart from them, we see opportunity in AI enablers such as NEC, Nidec, Verint and Criteo, as well as users of collaborative robots and AI like Amazon and Baidu.   The age of innovation Total patent grants by the USPTO by field of technology   Note:all ‘*’marked categories relate to data processing; Financial, Business Practice includes things like cost/price determination Source: USPTO. Hugo Scott-Gall [email protected] +1 (212) 902 0159 Goldman, Sachs & Co. Sumana Manohar, CFA [email protected] +44 (20) 7051 9677 Goldman Sachs International What’s inside The real consequences:  Our lead article on AI   2 Interview with…Raj Rajkumar:  Prof., Carnegie Mellon University 7 The road to automation: Takafumi Hara on the future of connected vehicles 9 Interview with…Rodney Brooks:  Founder, chairman and CTO of Rethink Robotics 11 The start of the collaboration era:  Yuichiro   Isayama on collaborative robots 13 Interview with…Manoj Saxena:  Founding General Partner, The Entrepreneur’s Fund   15 AI – is this time different?:  Greg Dunham on the use of big data in AI applications 17 The age of “robo” advisors:   Alexander Blostein on automated investing   18 0100020003000400050006000    2   0   0   0   2   0   0   1   2   0   0   2   2   0   0   3   2   0   0   4   2   0   0   5   2   0   0   6   2   0   0   7   2   0   0   8   2   0   0   9   2   0   1   0   2   0   1   1   2   0   1   2   2   0   1   3 Computer Graphics ProcessingImage AnalysisDatabase Management*Financial, Business Practice*Vehicles, Navigation*Speech Processing,Translation* Artificial Intelligence *Virtual Machine Task   Fortnightly Thoughts February 18, 2015 Issue 85 Goldman Sachs does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. For Reg AC certification and other important disclosures, see the Disclosure Appendix, or go to Analysts employed by non-US affiliates are not registered/qualified as research analysts with FINRA in the U.S.    Goldman Sachs Global Investment Research 2 Fortnightly Thoughts Issue 85 A broad way to define artificial intelligence or AI is that it is any intelligence exhibited by machines or software. By this we mean machines which can think, which are able to not just process data into information, but also derive knowledge from that information to augment human decision making or to act independently. It is difficult to draw a line between AI and a smart piece of code or smart connected devices, but the key difference here is that AI refers to technology that can be taught or is capable of self-learning and so can continually improve itself; capabilities once thought to be the forte of human beings. Take the case of Google search, which Nick Bostrom, Professor at Oxford University and a leading thinker in this area, highlights as the best example of AI to date in his book Superintelligence. What makes the srcinal Google algorithm truly powerful is that it is capable of learning from millions of users searching, ignoring and clicking through results each day, teaching it to yield better results every time. AI can similarly be applied to continually refine decision making in any information-driven business model to streamline costs, enable more efficient allocation of resources, improve product quality and accelerate innovation. This means that AI can be deployed in sectors that rely on skilled employees (manufacturing, software, engineering), skilled intermediaries (doctors for pharma, mechanics for cars) or skilled users (autos, agriculture), and even in operations and processes that are simple and mundane. We see AI as the next leg of the technological revolution that digitises decision making and as a result threatens swathes of human expertise and pattern recognition skills, lowers entry barriers for new competitors and adds to the disinflationary forces in the world. The stars are aligned  Artificial intelligence is by no means a new concept; however, even almost 60 years after the term was first coined it continues to conjure a perception that it is something that will be realised in the far future. This is the paradox of the AI effect – as soon as artificial intelligence achieves a commercial application, it ceases to be viewed as AI and is instead considered to be a really clever piece of code or a smart device. But in reality, AI is alive and kicking, and its influence is already evident in many industries, which we delve into in the next few paragraphs. However, we are writing about AI now because we think that it is on the cusp of a period of more rapid growth in its use and applications. The reasons are multi-pronged, but the improvements in the capability and economics of hardware is a good place to start. On one hand, many different device components such as sensors, cameras, radars, lasers etc. have become much cheaper in recent years thanks mostly to the smartphone revolution, which has driven innovation and economies of scale at an unprecedented pace. This has resulted in more connected devices and in turn, a lot more data being available for analysis. On the processing side, significant progress made in the field of neural networks and parallel computing in the last decade has meant that computers are now better able to understand even unstructured data, like oral conversations and pictures. Deep mines   At the same time, the advent of deep learning, only about a decade old in its current form, has enabled faster and more accurate reasoning and processing algorithms in most of today’s AI tools. Plus, as Manoj Saxena argues on page 15, the move to the cloud has meant that massive computing capacity and near infinite processing power is now available at very low prices and requires no upfront installation or capital spend, making AI more accessible to businesses than it has ever been. This should contribute to the network effect that accelerates the capability and adoption of AI – the more it is used, the better it becomes which should result in greater usage and so on. Taken together, the improved capability of machines to collate and comprehend more information and better optimise results is driving the broader and more significant potential for AI. Stores of value Stored data in the US by industry, petabytes   Source: IDC; BLS; McKinsey Global Institute analysis Where is AI now and where can it go?  AI going mainstream has broad implications, but it is currently most evident in data-heavy sectors. Algorithms behind Spotify, Netflix and Amazon are all aimed at driving greater customisation and user engagement. Retailers, both online and offline, are increasingly deploying AI to leverage their customer interaction data to boost revenue at lower selling costs. i.e. customised loyalty card-based promotions could become even more sophisticated. Smarter algorithms are also automating the process of buying and bidding for online ad space; programmatic ad buying, enabled by companies like Rocket Fuel, The Rubicon Project and Criteo, is reshaping digital advertising, driving cost efficiency for advertisers and publishers. And in the world of investing, algorithmic trading has been a force for the past few years, thanks mostly to its speed advantage. But more recent improvements in parallel computing technology have resulted in more powerful tools that are helping analysts digest reams of unstructured data, helping them replicate years of experience and acquired pattern recognition skills (see page 18 for Alex Blostein on robo-advisors). Private firm Kensho is an example here and so is BlackRock’s Aladdin (on understanding investment risks).  Apple’s Siri, Amazon’s Alexa (inside Echo), Microsoft’s Cortana and Google Now are also examples of AI-aided personal assistants on our devices. And this is perhaps one use case of AI which has grabbed the most mindshare of everyday consumers, by making that inconspicuous, but huge shift between text and voice commands and between reactive and pre-emptive search results and recommendations. This is the difference between searching for weather forecasts on smartphones versus them pre-emptively advising users to take an umbrella to their appointment. As they learn more and improve further, it is easy to imagine devices answering queries before they are asked. It is interesting to note here that last year, Amazon was granted a patent for ‘anticipatory shipping’, which aims to prepare a package for delivery before the customer actually makes the purchase, based on his or her browsing and purchasing behaviour. 01002003004005006007008009001000ConstructionConsumer/ recreationResourcesUtilitiesWholesaleTransportationInsuranceEducationRetailProfessional svcs.Securities/investment svcs.Healthcare providersBankingProcess manufacturingComm & mediaGovt.Discrete manufacturing The real consequences of AI    Goldman Sachs Global Investment Research 3 Fortnightly Thoughts Issue 85   Patently important   Patent grants by the USPTO office by technology class and owner    Note:all ‘*’marked categories relate to data processing; Financial, Business Practice includes things like Cost/Price Determination Source: USPTO. Which industries are becoming more information-intensive? Healthcare, and particularly pharmaceuticals, stands out here. One of the most well-known examples of cognitive computing of course is IBM’s Watson, which won Jeopardy! in 2011. Its first commercial foray was in healthcare in 2013, helping doctors lower the error rate in cancer diagnoses. As a result of the significant progress made in gene sequencing over the past decade, healthcare providers now have access to a substantial and rapidly growing amount of genetic data; and given that the true value in genomics lies in the ability to digest and understand this information to translate it into more personalised and accurate treatment, this is another area where AI can play a huge role. Education, energy-consumption, logistics and manufacturing are all becoming more information-driven as well and hence present opportunities for AI deployment.  Automatic genomics   Data output measured by number of terabases   Source: The Broad Institute  AI can also lower the reliance on high-skilled personnel, especially in sectors that face a shortage of talent, thus helping companies counter wage inflation and make their product or service available to a wider audience. Again healthcare is an area where this is relevant; AI could go some way in resolving the shortage of care workers, nurses and physicians. Other sectors that are both labour-intensive (wages as a % of costs) and talent-intensive (wages per employee) include media, financial services and technology. All said, artificial intelligence should augment human intelligence and lead to better-informed decisions, and depending on the nature of the industry it is deployed in, it can be disruptive to incumbents or at the very least it can re-allocate industry profits among them. It is useful here to draw parallels between AI and other technological waves – the use of plastic money, automation and even the advent of e-commerce. In the latter case, the investment opportunities were quite varied in the early years of online adoption. Whether it was in media, apparel or grocers, new entrants were disruptive to existing market shares, while amongst the incumbents, there was a marked difference in the performance of those who were willing and able to adapt early and those that were less capable of making the shift (especially those tied down by legacy fixed assets). Today, most large retailers have established an online presence and competition in some of these sectors is reverting back to the product quality, price or range (“old school” competition), rather than the convenience of the channel, but that has followed a costly and painful transition period. The longer the transition period, the more persistent the shifts in market share are (think online impact on food retail versus classifieds), even though the technology ultimately becomes part and parcel of the business.  AI we think is headed in the same direction. In other words, AI can introduce new pure-play entrants and re-shape the revenue pool, until it is eventually adopted more broadly in the industry (much like Netflix vs. other cable companies adopting OTT, Tesla vs. other OEMs building EVs). It will become critical for competition but irrelevant for strategy, which means that the advantages of adopting AI are easily outweighed by the risks of not using it. And so, investment opportunities are more likely to be found by identifying AI tool providers, new entrants and incumbents who are best suited to adapt or respond. Who provides the tools?  Within industries which have been early adopters of AI, there are few listed companies that provide specific tools like advertising and programmatic ad tech vendors or financial analytics. Japan’s NEC is dominant in face recognition and text analysis, which has broad applications in areas such as security and marketing. Tech firms that provide advanced data analytical and visualisation tools, like Verint Systems and Marketo are similarly exposed to this space (see longer list on page 6). In terms of general artificial intelligence, IBM’s Watson remains one of the most prevalent providers of AI solutions. It now offers 13 different services including speech-to-text translation and trade-off analytics on Bluemix, its cloud platform for developers. IBM’s CEO has previously mentioned that she hopes that the company will generate $10 bn in revenues from Watson in the next decade (versus $100 mn as of October 2014). Having said that, our analysts don’t foresee a significant revenue contribution from Watson in the near-term. But apart from these, most solution providers remain private and relatively small, reflecting the nascent nature of this space. Greg Dunham has more on this on page 17. Having said that, the list of AI start-ups that have been acquired by established companies in tech and other sectors has been growing rapidly in recent years. Google has been particularly acquisitive, buying deep learning, image recognition and neural network technologies. Yahoo, Microsoft, Intel, LinkedIn, Walmart, and Infosys have similarly invested in this field. This is of course one 01,0002,0003,0004,0005,0006,0007,000    I   B   M   M   i  c  r  o  s  o   f   t   I  n   d   i  v   i   d  u  a   l   G  o  o  g   l  e   S  o  n  y   S  a  m  s  u  n  g   E   l  e  c .   C  a  n  o  n   S   A   P   A  p  p   l  e   H   P   O  r  a  c   l  e   P  a  n  a  s  o  n   i  c  Artificial Intelligence *Computer Graphics ProcessingImage AnalysisVehicles, Navigation*Design & Simulation*Speech Processing, Translation*Financial, Business Practice*Database Management*Virtual Machine Task05001,0001,5002,0002,500200920102011201220132014e    Goldman Sachs Global Investment Research 4 Fortnightly Thoughts Issue 85   M&A activities related to Artificial Intelligence   Source: Goldman Sachs Global Investment Research route that incumbents can take to adopt AI, the other being that they develop AI capabilities in-house (like social media companies for instance). But more broadly, incumbents that are best suited to respond to any AI-led disruption are likely to be those that have access to proprietary data. With AI, data is only set to become a stronger entry barrier. A large and growing population dataset allows the machine or the software to learn faster and deeper. In other words, more data can make a clever algorithm cleverer and so an early mover with a unique or large dataset can build a huge advantage. AI can thus reinforce dominance in industries where some existing players have a pre-existing data advantage, while for the others, it can prove to be quite disruptive. Let’s start something special    Artificial Intelligence start-up funding Source: Company data and News Sources Baxter, is that you? So far we haven’t spent a lot of time on the hardware aspect of machine intelligence output because it is not paramount for AI. AI is the brain inside a robot, a robot is the container for AI and each can exist without the other. In previous issues of Fortnightly Thoughts, we have written at length about smart, flexible robots becoming more adept and expanding their scope beyond traditionally automated end-markets (like machine tool, auto manufacturing), into previously labour-intensive sectors like consumer electronics and food production. Robots equipped with AI can further negate the need for human instruction and intervention and, counter-intuitively, this is likely to have the greatest impact in tasks that are mundane, simple and time-consuming (like cleaning tools, moving waste and applying glue). On page 11, we interview Rodney Brooks of Rethink Robotics, which builds the Baxter robot and he notes that collaborative robots, which are smaller, mobile and often interactive, are better suited for small manufacturers with limited budgets, staff and space, rather than being disruptive to industrial robots that perform heavy and complex tasks. Another application for collaborative robotics is in logistics and warehouses; Amazon acquired Kiva Systems in 2013 which provides automation solutions in its fulfilment centres, using mobile robots and sophisticated control technology to identify and retrieve packages that have been ordered, allowing for faster cycle times, longer operational hours and reduced labour requirements. Again, these are simple, but cumbersome and tedious jobs.  At your service Sale of service robots by type Note: 2013 data not available for PR robots, only handicap assistance sales considered for 2013 in the elderly and handicap assistance category Source: IFR World Robotics Helping the age old problem Outside of factories and warehouses, the potential for collaborative and service robots is perhaps highest in the field of healthcare – ageing demographics and a shortage of care workers is an emerging issue in many developed markets and particularly in Japan. This is one of the reasons that in 2013, Prime Minister Shinzo Abe's government announced subsidies to encourage companies to develop robotic care for the elderly (the government forecasts the nursing care robotic market to grow from $140 mn in 2015 to $3.4 bn by 2035). These are robots that can understand human conversations, avoid collisions, lift heavy objects (including people) and in general, make life incrementally easier for users. Acquirer Target Business   description Acquirer Target Business   descriptionTechnology Socia   media Google DeepMind   ($400m)  ‐ 2014 Deep   learning   to   understand   images,   text   and   videos Facebook  ‐ 2015 Siri ‐ like   API   voice   interfaceDark   Blue  ‐ 2014 Deep   learning   to   understand   natural   language Twitter Madbits  ‐ 2014 Image   understanding   using   deep   learningVision   Factory  ‐ 2014 Visual   recognition   using   deep   learning Pinterest VisualGraph  ‐ 2014 Machine   vision,   image   recognition,   visual   searchJetpac  ‐ 2014 Image   recognition   and   neural   network   technology LinkedIn Bright   ($120m)  ‐ 2014 AI   and   big   data   algorithims   to   connect   usersDNNresearch  ‐ 2013 Deep   neural   networks,   language   processing  E ‐ commerce Yahoo IQ    Engines  ‐ 2013 Image   recognition   to   tag   and   organise   photos Ebay Apptek  ‐ 2014 Hybrid   machine   translation   using   machine   learningLookFlow  ‐ 2013 Enhanced   image   recognition Amazon Evi   Technology  ‐ 2013 Internet   search   and   voice   recognitionSkyPhrase  ‐ 2013 Natural   language   processing   technology Kiva   ($775m)  ‐ 2012 Manufacturing   mobile   robotic   fulfillment   systemIBM Kenexa   ($1.3bn)  ‐ 2012 Data   analysis   to   help   recruit   and   retain   workers Deutsche   Tel Magisto  ‐ 2014 AI   video   story   telling   platform   (partnership)Microsoft Equivio  ‐ 2015 Machine   learning   technology   for   info   governance  Retail Revolution   Analytics  ‐ 2015 Statistical   computing,   predictive   analytics Staples Runa  ‐ 2013 Specialist   in   e ‐ commerce   personalization   techInfosys Panaya   ($200m)  ‐ 2015 Automated   cloud   based   quality   management   services Walmart Luvocracy  ‐ 2014 Discovering   recommendations   by   friendsDropbox Anchovi   Labs  ‐ 2012 Image   classification   using   AI Inkiru   Inc.  ‐ 2013 Predictive   analytics   applicationsIntel Indisys  ‐ 2013 Natural   language   recognition Home   Depot Black   Locus  ‐ 2012 Data   analytics   innovation   lab   based   in   Austin   TXOmek   ($40m)  ‐ 2013 Maker   of    gesture ‐ based   interfaces  Capgoods Stryker MAKO   Surgical   ($1.65b)  ‐ 2013 Advance   robotic   assisted   surgery   in   orthopedics GE Pivotal   ($100m+)  ‐ 2013 Develops   data   analytics   offerings   Monsanto Climate   Corp.   ($930m)  ‐ 2013 Underwriting   weather   insurance   in   real ‐ time Schneider InStep   Software  ‐ 2014 Real ‐ time   performance   mgt,   predictive   analytics   Artificial   Intelligence   Start ‐ upTotal   raised   ($m) Start ‐ up   Description Sentient   Technologie 144 Data   analysis   through   massively   scaled   AIRethink   Robotics 127 Robots   for   production   and   researchVicarious 72 Advanced   image   recognitionModernizing   Medicin 55 Cloud ‐ based   Electronic   Medical   AssistantContext   Relevant 44 Using   machine   learning   to   analyse   big   dataNarrative   Science 32 Generate   written   naratives   from   dataScaled   Inference 27 Platform   for   general ‐ purpose   AIKensho 26 Automation   of    financial   researchBuildingIQ 21 Energy   management   software   platformExpect   Labs 15 Voice ‐ activated   discoveryBlue   River   Technolog 13 Lettuce   thinning   using   machine   learning   Nara   Logics 13 Big   data   analysisTempo   AI 13 Mobile   productivity   app   organises   user's   daySentrian 12 Remote   biosensors   to   prevent   hospitalisationPrecisionHawk 11 Aerial   data   analysis,   AI   software   for   UAVsMetaMind 8 Natural   language   &   image   recognition 0200040006000800010000120001400016000PR (guides in supermarkets, museums etc.)Rescue and securityUnderwater systemsInspection, maintenanceProfessional cleaningConstruction, demolitionElderly and handicap assistanceMobile platformsMedical robots Automated Guided VehiclesLogistic systemsMilking RobotsDefence applications2013Average annual sales, 2014-17E
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks