At the end of 2017 if it didn't become abundantly clear that artificial intelligence is well on its way to changing everything that matters about the way we work, then you weren't paying attention. The examples were too numerous to list, but here's a few: In 2017 AI systems beat human doctors at detecting irregular heartbeats, tracked player statistics for NFL football fans, and out-bluffed the world's best poker players at Texas Hold'Em. When it comes to solving problems, and doing jobs that are inherently based on patterns, the machines have overwhelmingly won the race against the human brain. So as 2018 gets underway, it's worth asking this out loud: If AI is changing everything about our jobs, what does that say about how we as humans prepare for those jobs? I got into a conversation about this last month with Gordon Ritter, founder and general partner at Emergence Capital.
It wasn't that long ago that the National Football League (NFL) was considered infallible. The league had a lock on the most coveted demographics in the country, and the revenues that poured in where only outdone by the mega ratings that accompanied them. The fact that they could achieve this as the rest of the world moved away from real-time, "destination TV" viewing made the NFL all the more desirable to advertisers. But as the 2017 Playoffs kick off, that era seems like a distant memory. Ratings have been in a free fall.
Video: IBM provides ultimate fan experience at Atlanta's Mercedes-Benz stadium You have more choices for your football-watching pleasure on the internet than ever, and that's a problem. I promised you that this year, now that my book is more or less out of the way beyond what is mostly cleanup, I would provide you with some fresh thinking and some great content from me and a host of guests -- plus some exciting announcements. Because I'm preparing a post that includes announcements about the CRM Watchlist changes for 2018 to 2019 and the rules of the Emergence Maturity Index battle royale, I'm starting the new year with a guest post from a really interesting guy who is positing a really interesting idea -- one that as a set of principals and practices I have been a firm adherent of for many years. Alex Slawsby goes by the title of director of innovation for Embraer, which, as all the road warriors among you know, is one of the largest producers of jets in the world. But prior to this job, Alex worked for the Sports Innovation Lab, putting forward the ideas of Clayton Christensen (who he also worked for) on disruption and innovation.
Intel is taking heat after it was revealed CEO Brian Krzanich privately learned of two vulnerabilities in its semiconductors before selling millions of dollars in company shares. LAS VEGAS -- Intel CEO Brian Krzanich wants to talk about the future -- the really cool if somewhat creepy future of drone swarms and chips that track your every movement. But the past just won't let go. The CEO has planned an elaborate, artificial-intelligence driven extravaganza during his Monday night keynote address at CES, the world's largest tech trade show. The company best known for chips that power PCs wants to show off its updated vision of how its smarter, tiny hardware components will look, with a drone light show and musicians creating sound virtually -- they'll play no instruments, but location tech will crunch the data created by their physical movements to power guitars and drums.
In 2017, my team powered chatbots and voice skills for leading brands like Nike, Vice, Jameson, Marriott Rewards, Simon, Gatorade, and more. We witnessed new user behaviors and uncovered an evolved set of best practices to build a chatbot. Here are four actionable learnings from our work that you should consider when launching your own chatbot in 2018. Bots that are designed to segment and engage customers throughout the entire conversation drive higher metrics than chatbots that do not personalize the conversation. For example, in our testing, personalized results yielded the highest click-through to website, up to 74 percent in some cases.
At least 80% of enterprise data is unstructured, contained in the myriad text-based social conversations that are happening every day. Unlocking the hidden value of text through predictive analytics is imperative to the understanding of customers' opinions and needs, to make better, more informed business decisions. A whopping 90% of this data is actually completely underutilized when it comes to data strategies and data analytics techniques. It's very easy for humans to consume and make sense of unstructured data, but machines don't find it as easy. It's not like other data sources, it's not staying in the table or a database, and it's not easily referenceable.
The Defense Advanced Research Projects Agency (DARPA) explored the application of transfer -- a notion well studied in psychology -- to machine learning. This article discusses the formal measure of transfer and how it evolved. We discuss lessons learned, progress made at the formal and algorithmic levels, and thoughts about current and future prospects for the practical application of this technology. The aims of TLP were to understand and formally frame how this intuitively compelling psychological idea might apply in the computational context, build computational models of transfer learning (TL), and explore how these models might apply to practical learning tasks. TLP and the field as a whole made great strides in each of these dimensions.
In practice however, there are many barriers to achieving this goal. In this article, we present a prototype system for the real-world context of transferring knowledge of American football from video observation to control in a game simulator. We trace an example play from the raw video through execution and adaptation in the simulator, highlighting the system's component algorithms along with issues of complexity, generality, and scale. We then conclude with a discussion of the implications of this work for other applications, along with several possible improvements. Broadly speaking, the goal is to apply knowledge acquired in the context of one task to a second task in the hopes of reducing the overhead associated with training and knowledge engineering in the second task.
Information extraction (IE) can identify a set of relations from free text to support question answering (QA). Until recently, IE systems were domain specific and needed a combination of manual engineering and supervised learning to adapt to each target domain. A new paradigm, Open IE, operates on large text corpora without any manual tagging of relations, and indeed without any prespecified relations. Due to its open-domain and open-relation nature, Open IE is purely textual and is unable to relate the surface forms to an ontology, if known in advance. We explore the steps needed to adapt Open IE to a domain-specific ontology and demonstrate our approach of mapping domainindependent tuples to an ontology using domains from the DARPA Machine Reading Project.
The NFL is joining Major League Baseball as an AWS customer, announcing a deal today to provide real-time statistics running on AWS. The tool is part of the NFL's Next Gen Stats program, which will take advantage of AWS machine learning and data analytics tools to enhance its current offering. MLB has had a similar deal in place with its StatCast tool. The NFL uses RFID tags in player equipment and the ball to capture real-time location, speed, and acceleration data. Much like the MLB product, this data can be used to heighten the NFL broadcast experience by showing viewers a unique data-driven view of the play on the field.