Here. Our next guest is very exciting to have him join us on the stage. Uri and i talked a little bit about a. I. Elons not a fan, uri is. Either way, the future is here. So please welcome to the stage dr. John kelly from ibm and our moderator, alex wilhelm. Okay. So were going to try to be different today. Were going to do a live demo which i props you works this morning promise you works this morning. Well see. What were about to see and why do we care . Well, alex, what youre about to see is something that no one else has ever seen other than people deep in my research labs. Many of you remember watson, the Artificial Intelligence machine, on the game of jeopardy. At the time it was an open domain, question and answer system at the forefront of Artificial Intelligence. Fast forward to today, what youre going to see is watson has ingested all of wikipedia and not only answers questions, but reasons over that and form opinions on any subject. Okay. Lets, so the tech guys can make it turn on for us. There we go. And i believe it has an introduction going on first. Watson . Is it going to do it . Is it going to happen . Lets go watson . Please start. Hello, and welcome to the ibm demonstration. Today we will demonstrate [inaudible conversations] capabilities in generating arguments for or against specific [inaudible] please select a topic. Okay. Lets go ahead and do wikipedia is reliable pro speech. Scanned approximately four million wikipedia articles. Scanned all 3,000 sentences in top ten articles. Assessed pro and con [inaudible] ready to deliver. Thank you for the opportunity to share my view about this interesting topic. It is the always good to start with some background. The reliability of wikipedia compared to other encyclopedias and more specialized sources has been assessed in many ways including statistically, through comparative review, analysis of the historical patterns and strengths and weaknesses inherent in the process unique to wikipedia. My opinion is that wikipedia is reliable, and i would like to state a few simple reasons for that. Its editors and contributors provide identification. Another reason is that wikipedia provided better coverage and longer articles in general. These claims are supported by the scientific literature. For example, an early study conducted by ibm researchers in 2003 two years following wikipedias establishment found that vandalism is usually repaired extremely quickly, so quickly that most users will never see its effect and concluded that wikipedia has surprisingly effective selfhealing capabilities. Im sure that my time here was not wasted, and i hope i have convinced you to support my view. Wikipedia is reliable. [laughter] thank you for listening. [applause] thatll make everybody so happy. Its reliable, we can use it on papers. So thats good. Im very curious, when you ininterest that much data into watson ingest that much data into watson, do you say i want you to answer this, do you train it . Or does it automatically understand how to the pull out different topics . All you coto is give it a topic all you do is give it a topic, alex. Watson was not trained in that domain. We gave it a topic and said form an opinion on this topic. It went through and organized all the information, it understood the context of all of the passages in wikipedia, took apart all the syntax, reassembled it and then machine generated the language. How does it understand the concept like authenticity or accuracy or whats good about wikipedia . How does it understand those concepts so it can actually pick how to argue its own position . Yeah. So it looks for multiple scenarios and reinforcing information. For instance, i was surprised to hear reference ibm research working in that area. [inaudible] yeah. So what it did was it found some passage in wikipedia, and then it went off and validated that that was a legitimate scientist from ibm research before it would say it. So its constantly looking for verification of what its about to say is correct. So is that piece of technology, is that currently something people can use on the market, or is it more of a forwardlooking thing you havent released yet . Thats a forward looking. Let me remind you all at the time of jeopardy, watson was a large Computer System in a room. Since that time we have taken watson apart, if you will, lifted it up to the cloud sure. Its available as a set of services and apis so that everyone in this room can now go out and start to compose their own mini watsons, if you will, in the application space. So how long until thats not forward looking but actually in the market . Is it a six month question . Is it a three year question . How far away are we from having that out . This is probably 1218 months or so away from being in the market. Okay. Imagine, though, thats a trivial example of wikipedia. Remember that in the area of health care, watson has ingested almost all of the worlds medical domain. So if you give watson a disease topic, it will go into all published medical journals and form opinions on diagnoses or drug discoveries. So i was doing some research, and i found a really fun story. Nikita khrushchev actually came to an ibm facility down in san jose thats right. This was back in the 50s. They are rigged up a computer he could ask questions of back in 56, 57. So you have been on this topic for decades and decades and decades. Have you solved it . Is it done now . Or are you still making progress on the issue . Yeah, thats a great question. So prior to watson, everyone in the Artificial Intelligence community had tried to solve open domain question and answer by writing a bunch of rules, rulesbased learning. Or organizing the information for search. Sure, sure. That had very limited success, and it really wasnt until we took a full open domain statistical approach and taught watt natural language watson natural language that it was able to answer questions in open domain. Okay. So on that point, ive heard more about a. I. In the last six months than in the last six years. It does seem to be a moment going on. Yes. Why are we seeing innovation grow so quickly right now . Whats changed in our approach maybe . Thats allowed us to do this . Yes. I think were sort of at the sort of proverbial perfect storm here. Much of the worlds information is now digitized including natural language in areas like health care, law, etc. So computers can now access it. We now have the compute power to do Something Like you just saw. We did not have that a few years ago. So just the raw metal capability. Yeah. And what has really changed then in the area of Artificial Intelligence is areas like Machine Learning, Statistical Analysis are now advanced to to the point where we can do things like you saw there. And just to make sure i understand correctly, Machine Learning is one of the things watson can use, part of the technology stack, if you will . Its one of the Building Blocks of watson itself . Thats right. Think of watson as a whole series of statistical learning engines. Its not one thing. And its constantly computing and trying to learn over information. We do not program watson. We give watson data or we give watson better data, but we do not reprogram it. We give it better algorithms or better Machine Learning. So you give it a stronger brain over time, but you dont tell it what to think. It learns over the information. Okay. I brought some with us today, we have them here, when you go out to pitch developers they should use watson at mit and new york, whats your core pitch . What do you start with . What do you tell them . The interesting thing is its such a powerful technology, i dont have to give a sales pitch. They immediately go into how can i use this . Can you give me an api or a set of services to do this . I was up at mit a couple of weeks ago, and theyre blown away by it. Theyre just using the services now that are coming out of this. We have opened up, as i said, the platform, and there are hundreds of startup companies. Ibm, we formed a 100 million fund to starlet the fund and bring additional to start the fund and bring additional. Is if they want a check, they should find you when were off tabling. They should grab me. Been kind of an outreach push for you, so how has developer update been so far . Has it been slow . Its been exponential, alex. Okay. We are returning as fast as we can to keep up with demand in the investor commitment. Even Large Enterprises now are using it to transform many areas of their business. So ibm is a very, very venerable company, very old, very respected, big history, lots of work. Young, its not snap you know, its not snapchat, its not facebook. Do you think you have the mind share you need here in Silicon Valley to attract even more developers to your company, or are you guys still fighting off that stodgy ibm pocket protector kind of image . [laughter] no fence. No, no, no. Well, we do not have mind share here. Its gaining rapidly, its growing exponentially. But if you listen carefully later this week, youre going to hear some things from us that i think are going to reinforce our commit to the Developer Community out here in San Francisco and in the valley. I would also remind you that when you think about the eras of computing, the first was tabulating machines. That lasted through roughly the 40s. And then we we started programme computers. Every device is programmable. Watson is not a programmable system. Its a learning system. St the first of a third era of computing. And ill just remind you that ibm played a pretty Important Role in the first and second eras of computing. We will play a key role in the third era of computing. Thats a really good, actual, segway. Whos your main competition here . Are any of the big players out there trying to encroach on your territory, or are you a bit out ahead . Yeah, i think well, first of all, theres a lot of great people working in the area of a. I. And cog cognitive computin. Most of them, though, are developing single algorithms or solutions to solve some problem so a very narrow theyre very narrow solutions. Were the only company, alex, that has developed a complete platform, the ability to innovate on top of this. And, i mean, if i had to make a prediction, i would say for every innovation going forward, therell be a hundred or thousands of pieces of innovation that will occur on top of the platform. And thats what we want because, yes, were an old to company, but we think in terms of era computing. We dont hi in terms of thunderstorm we dont think in terms of tomorrow afternoon. [inaudible] is that correct . We do a revenue share, value share on the platform. Aws has a different sort of platform, charges by usage and not by rev share. So why did you pick yeah. Well, you know, the first decision we made we could have, and we still could sell lots of watson boxes. But we decided that we wanted to open the platform up and make it as a service knowing where cloud was going, knowing where as a service was going. So that was a very strategic decision for us. And then its just a matter of do you charge by the click or not . And because this is a learning system and its constantly becoming better and smarter, it was very important that we share the value with our partners and customers, and, therefore, the rev share, value share model. Have you had pushback from developers saying they want to pay for it more on a click basis . No, because its a shared risk. For developers, you come on for free, develop your business, and as you start to earn your revenue, then well share that. But we want an easy onramp. We want to get going, and we think its very fair. Is watson going to be a key revenue driver for ibm moving forward . We have a set of Strategic Initiatives in the company which is our highest growth areas. Watson is part of our Analytics Business which is a 17 billion doubledigit high growth rate business. And watson is the Fastest Growing component of our Analytics Business. Okay. We can talk about future. Oh, great. Thank god. Im going to go ahead and presume you dont think that a. I. Or cog anity [inaudible] is going to end human kind as we know it. Because theres been some talk among luminaries about what this looks like and should we be worried. We talked about this seems relatively benign and more industrial than aggressive. Yeah. You know, i think that when any new Technology Come along, people get worried. And, of course, technology can always do great things or not so great things. It depends on how we use it and how we control it, you know, whether that was the steam engine, locomotives, automobiles, xrays, you name it. This is a brand new technology. Its going to change the world. There is no question in my mind. But when you think about what this technology is do, i mean, fundamentally if you look at decisions that human beings make, we make decisions with a bias, we make decisions on incomplete information. Yes. I believe that nearly or perhaps every decision that we are make of any importance mt. Future will be made with a in the future will be made with a watson by our side. Its giving us options that we can select from, so so its more of an addition. Thats right. So think of the simple demonstration we did, although complex technology around wikipedia. Think about watson where it has ingested all of the worlds information on drug therapies, cocktails for oncology, for cancer. I dont know about you, alex, but if i had cancer, god forbid, and the cancer board was meeting to decide what chemotherapy they were going to give me, i would want watson to have gone through to be sitting at that cancer board to augment the decision and say, hey, you know, guess what . You may want to rethink that. Or heres a new con to that procedure. Right. Okay, but as watson or platforms like it get smarter over time, will they less and less be help mates and more leaders in our lives, or are they always going to maintain that secondary status . Handsoff in a way stuff. Yeah. I think therell always be a set of decisions or moral decisions where itll be a joint decision, but we as humans will always have the control position. But i think as time goes on more and more decisions will be able to allow the system to make. Think about internet of things where we need constant response in milliseconds. Making those decisions in real time versus waiting for me to be able to digest something in seconds or minutes. So for all of our daily lives, does watson power apps and experiences we already control or is it a standalone product that i would talk to . Is it more baked in or standout . Yeah. So i predict its going to be literally everywhere. It will be powering things that you just take routine in your daily life. If youre applying or buying something, watson will be behind that in enhancing your experience. But it will also be visible in of your applications where youre being creative or you want information that you just dont have time. So i think its going to be invisible and highly visible, basically ubiquitous in everything that we do. But to be fair, you want watson to be a brand name in technology. Am i going to have a powered by watson sticker on my pone . Incredibly, alec, watson is already the premiere brand in cognitive computing and Artificial Intelligence. Not something we set out to do, but through the demonstration in jeopardy and what weve already applied it to, it already has very broad brand recognition. So as we move forward, well decide whether we want to do powered by watson. We already, for instance, opened up a new division of ibm around watson health. Sure. Thats your take on the health care vertical thats our health care vertical. Youve seen us recently stand up new divisions around internet of things, around education. So as watson then becomes capable of operating in those domains, well bring watson into those areas. Okay. Well, before we go, humor me with a last question. Was will a moment today where you and your team were like, oh, wow, it actually works . Howd that moment come to be . Yeah. So in the journey to jeopardy which was all the way back in 2011, there was a point many time in in time in 2009 when watson had made a quantum jump forward in it ability to learn and answer open domain questions. At that point i remember sitting in my Conference Room looking at the demonstration, and i said we have something thats to going to change the world. It was shocking. Did people believe you back then when you said that . Were they like, youre craze i or, yeah, we agree . Pardon me . Did people agree with you at the time . Yes. Obviously, my Research Team had high creation visions, but they were very much focused on just a q and a machine. When i saw that system could reason and learn, i said we have something thats going to change the world. This is not just another Computer System. This is a new era of computing. But, again, no zombies, no apocalypse, no taking over the world, no matrix . Only going to help us, alex, with really complex decisions. Good. Thanks for coming out, man. Thanks. [applause] [inaudible conversations] ground rules over there. Man, this room is filling up. Youre here to here the investor talk, arent you . Here hear the investor talk. How many entrepreneurs are in the room . So all of you. Listen up then. This is time that even like shut off your laptops and just learn something, because these folks are the ones that are going to give you money. Please welcome to the stage aileen lee, jeremy liew and dana sid t and settle and connie loizos. Big round of applause. Pleasure. Ooh, i like your shoes. All right. Thank you. Well, im so glad to have great panel here. Guys, im going to just roll right boo things and ask you a hypothetical right into things and ask you a hypothetical. Say i am this super smart intrup neuro. Ive entrepreneur. I live in austin, and everyone keeps telling me i have to move to San Francisco. Is that true . Dana . Not if youre raising money from greycroft. Why is that . [laughter] you know, i think 82 percent of our investments have actually been outside of ill con valley. Silicon valley. So were finding opportunities everywhere, and i think there are great growth pockets in austin, in new york, in l. A. , in seattle, in chicago. Sort of all over the country and all ove