Genius of Things: IoT in the Cognitive Era with Dr John E Kelly



every time Harriet introduces me as the Godfather or Watson I said it must be an age joke Perry I'm not that old well good morning I am thrilled to be here with all of you I have to just tell a quick story how did we end up here in Munich with our IOT Center it was about a year and a half ago after we had located our centres for Watson health and in Boston for Watson developers in San Francisco for what for Watson platform in New York that we we decided that we were going to move big in IOT in the industrial area as a company and we took out a global map and said okay where are we going to put our Center for IOT on this globe being a large company in 170 countries we could have planted any place on that map but it was a very quick decision we said we're going to put it in Germany we're going to put it in Munich right in the heart of the industrial sector because we knew that we needed to partner with our clients and those who really had the domain expertise and what a motive in industrial and robotics etc now I have to tell you as sure as we were of that bet you always worry when you throw a party are people going to show up for the party well I think we found out today Harriet's that people showed up for the party in fact Harriet said there's 600 executives here there's another 300 that wanted to get in this room and we had to politely say no thank you next time but we're just thrilled that all of you came so let me give you a very quick perspective on how we're looking at the evolution of IOT and the role of artificial intelligence and cognitive in that I don't need to tell you about the numbers you're all experts in this but the amount of data we're generating and the amount of opportunity that there is for all of us to drive the world's economies is absolutely off the charts in terms of size and while there's the number of bits and bytes the data is is growing to almost ten with 24 zeros after it the real opportunity lies in the fact that 80% of that data is unstructured and dark and the majority of it is underutilized or not being utilized at all in fact in most cases it might even just be being stored and put away on a shelf that is huge lost opportunities for all of us in our businesses to uncover now there's a lot of talk about machine learning and artificial intelligence and cognitive and I thought I'd just spend one quick minute to sort of level-set everyone on what are the differences between these terms and the analogy I like to use it's a fairly simple analogy it's our own sort of human brain when you think about machine learning it's basically taking data data streams and looking for patterns and quickly reacting based on what the algorithms are learning over time that's very analogous to the bottom of your brain your brain stem where if somebody if you touch something sharp you immediately react to it because it's going quickly back and forth to the base of your brain that's the lowest level of sort of cognitive capability above that levels and it's artificial intelligence where we start to put together different forms of machine learning different forms of artificial intelligence that now start to work together in a more complex way more complex set of algorithms and those are you know more typical human functions and functions of all the other animal creatures on the earth what cognitive is the upper frontal lobe of the brain that is where humans differentiate and that's where we start to bring together different concepts and start to reason and to me that's where the enormous value in IOT and cognitive come together so it's not just about reading a simple image or processing a simple signal from a piece of equipment it's bringing together the human factors the context and all of the data sources into a more complex that are reasoning and that's what cognitive is and that to me is where exponential value can be unlocked in IOT the size of the opportunity in cost all industries is enormous and this is one reason why iBM has chosen to make an enormous investment in this area because as we look at the opportunities the largest of the opportunities for decision support is in IOT the industrial sector there's other enormous opportunities in healthcare and financial services and we are also going after those opportunities but this is the industrial area where the opportunity is just off the charts in terms of size but we also know that we have to partner to do this we are great at AI in cognitive we are great at handling large amounts of data processing data cloud technology but we know that we must partner for the industry expertise to really bring that capability to life and that's what this Center is all about that's what our partnership with all of you is about is partnering to unlock that value now our strategy in a in a nutshell is very very simple its number one the base platform for all that we do will be cloud cloud is the platform for IT technology for the next set of several decades it is global it's secure it's highly resilient it's elastic it's all the things that you need as the core platform for an IOT cognitive solution infinite compute power infinite storage capability burst capability etc secondly is lead in cognitive lead in AI we have the best algorithms the best cognitive capabilities the best platforms to do that all resident within our clouds and of course our brand for that cognitive capability is watching world leadership in technology not only in speech but in image recognition and signal processing all of the things that are are important to you and then thirdly and I think this highly differentiates us is that we do it through an industry lens we believe that it's decades from now before anyone will have a general purpose all-knowing cognitive system these systems must be trained in the domain that you're in if it's automotive it's robotics with health care if it's financial services they need to know these domains a dr. Watson analyzing cancer at a Medical Center is a different cognitive system than what's going to be in a autonomous autonomous car and so we're going at this through industry lenses and as soon as we do that of course we have to partner because we are not and what a motive manufacturer or parts manufacturer a robotics manufacturer we want to bring this capability to the market but we want to do it with you and through you very distinguished strategy our competitors will have won a couple of them have two of these capabilities but none of our competitors have these three capabilities and I hope that provides huge value for you in partnering with us so we have chosen three industries initially to make enormous investments as a company healthcare where we acquired four companies 7,000 employees in less than 12 months and we now have one of the largest repositories of health care data in the world through these acquired companies it's us data but we have a hundred million individual lives of medical records over decades which is being used routinely now to Train Watson in areas like diagnosis of cancer or other fatal diseases in the area of financial services this is another huge opportunity for cognitive everything from fraud detection the regulatory compliance to financial services individuals that are serving their clients and here again we're using Watson to augment human capability in the US we just formed a partnership you might have seen or heard with H&R Block which is one of the big tax consulting firms where Watson is helping the client and the tax advisor maximize the consumers tax deductions aka pay less taxes so what is becoming very popular in the u.s. in that in that regard and then IOT where we said we will invest three billion dollars when we chose this Center that you'll see hopefully everyone will move over to the center after this meeting you'll see that enormous investment not just on our behalf but on your behalf and what I'm most excited about is that it's not just our Center it's your Center we invite you to partner with us there more and more companies large and small are now coming to co-locate with us in that building I remember the discussion Harry when we said well what building are we going to take and I said this is going to be big just get the biggest building you can get in Munich having no idea how big you think Harriet see as you'll see we've got the biggest building in Munich but I think we're think we're going to fill it I think like this room we're not going to have room for all of the companies and people that want to come and partner and use this as their collaboratory going forward so space is at a premium get get lined quickly before it fills up I want to leave you with a picture of the way we're thinking about the architecture of IOT and first and foremost it starts with the clouds yes we can do private instances of analytics computing machines and whatnot but if you truly want global scale and you want the elasticity to be able to handle bursts of data you have to start with a cloud and that has to be a highly secure cloud the kinds of data that you're going to be putting on it secondly you need that that's lost in that that incredible AI cognate at the heart and center of your capability but underneath that the foundation the oil whatever you want to describe it is the data I always remind people that these AI cognitive systems are no longer programmable system you don't make them smarter by reprogramming them you make them smarter by giving them better and better and more relevant data to whatever problem you're trying to solve so underlying data whether it's your private data data that we have public data bringing these sources of data together is absolutely critical and this cloud is designed to do that at enormous scale any place on the planet then we're taking that capability and we're putting it into our own products and offerings things like Watson for cybersecurity that we announced early this week where Watson now is part of our security solution Watson basically sits on the sensors of our networks and systems looking for malware looking for attacks internal and external to enterprises and watch them not only through its machine learning finds those subtle subtle attacks that are coming in from a cyber security standpoint but it can react at Machine speed at the same time it's reading the entire web looking for any analysis of a recently released malware any plugs against that brings those two together and reacts in real time that is way beyond just low-level machine learning but again enormous capability and then the three industries that I told you about a minute ago are where we're investing the majority of our money to help companies such as all of yours and then lastly which I think also differentiates us is we are an open ecosystem we have api's that we expose for our developers your developers startup companies which again many of which will be joining us over at our Center as there resident but this allows companies to start up I flew from we had a enormous partner event in Las Vegas two days ago we had 1500 of our partners there all who are using the api's from Watson to do everything from environmental safety to health care to IOT in the United States an enormous ecosystem now being built on top of this platform so again I think what makes us different from a lot of others that are you know buzzing about machine learning or AI is number one we're taking an industry view and I think that will unlock a lot more value than just building some sort of smart box that sits in the middle of your your living room we want to improve the enterprise and unlock the value there secondly is the scale the enterprise scale I know of no IOT problem that we cannot solve at the scale we're building there's only three or four large cloud providers in the world we're one of them and we're the one that is the industrial-strength target platform for IOT and then lastly is next generation capability we invest six billion dollars a year in R&D it's a lot of money maybe not as much as some in the room but it's a lot of money six billion of R&D a huge portion of that is focused on this area we have 1,000 people just in our research division working on AI and machine learning and cognitive so let me just give you a glimpse of three things that I think are huge differentiators one is around weather I know a year and a half ago when IBM when we acquired the weather company people scratch their head and said why is IBM buying the weather company well true the data is incredible but we also with the weather company acquired the largest IOT data ingestion engine on the planet we we take in from every place in the world weather related data we carve the planet up to three billion elements and every ten minutes we run five weather models from first physics principles on those three billion elements to predict the weather in that element of the planet that is used for everything from insurance to drivers to aircraft planning to turbulence monitoring and avoidance in the planes that you use to get here an enormous capability and I think we're just scratching the surface of how we're going to use that weather data secondly it's very clear that not all IOT information is going to have to go up to the cloud in fact we don't want it to come up to the cloud more and more of the IOT intelligence cognitive capability will be at the edge device the device whether it's machines a machine on factory floors or automobiles or automobile or automobiles a sensor on the highway system so we're building the technology to push the intelligence out to the edge versus having it to go back up to the Watson brain basically distribute Watson's brains locally in connected devices and then lastly a technology which i think is absolutely going to change almost every industry and that is blockchain where we can do highly secure distributed transactions on a distributed ledger that self-correct and self locate instantaneously with every transaction that's enormous capability whether it's the financial sector whether supply chain or whether its distribution of any product or service having that local instantaneous secure ledger takes out enormous friction from all of these systems that will be a game changer and again it can only be done in a cloud such as this so very unique capabilities now I'll sort of close with this there are I think a couple of principles which we all need to think carefully about as we make decisions in this space one is data and Trust so if all the value is locked up in that data then where it is who has it what they're doing with it is extremely important and I'm not talking just about individual privacy but if it's the design of your product the signals coming off of your your systems that is very very sensitive data we have taken a position as a company and put out a set of principles about two weeks ago that says for one thing if we use AI and cognitive in a solution for you or to you we will tell you the AI system we will also tell you what data we use to train it so if watch is going to diagnose your health issue we will tell you watch this being use we will tell you where it was trained it was trained at memorial sloan-kettering by these doctors and we will also keep that data of yours yours we will not use that data with anyone else we will not train it without you're not trained our systems without your permission to do that that is highly differentiated from many of our competitors who would like nothing more than to have your data to train their models for things they'd like to do for their own business that is not our model our model is we want to be that the trusted partner for that data and work with you on the other side is the security itself it is absolutely critical that this data is properly encrypted at the locality of the data particularly when it's person sensitive or device or health or financial services the location of that I will tell you that many of our cloud competitors in spite of what they tell you they cannot assure you exactly physically where that data is and we know this because we have tested them and you think your data is in Germany it might be sitting someplace in China or the Middle East they do not know exactly where that data is at every point in time or who's touching that data so we take these two principles very very importantly and we think that these are not only differentiator for us but you need to demand these two things from any partner particularly as you're getting into your most sensitive data so I'll end with this I think it's a it's a great quote from our Chairman that this is the era that will redefine the relationship between man and machine I think that's absolutely true but I think if I could respectfully augment our Chairman's comment I think it also is going to redefine machine to machine in some ways that are really fundamental and in redefining that machine to machine we ourselves are going to redefine how we interact with this physical world this is an exciting time I've been in this industry three-and-a-half decades I've seen some incredible technologies I've helped to build some of the biggest computers in the world but I have never been at a point where I've been so excited about the opportunity for technology and how we can change the world and now I think what you'll see from a lot of our clients and partners the way that we're changing the world together thanks for the invitation Harry for being here let's have a great time thank you [Applause]

3 Comments

  1. Audrey Mciver said:

    I want you to take my data and create a security software for end to end protection and to identify the financial transactions that someone has done by hacking pinging And or piggybacking my utube. I know someone has done something using my credentials if it is my mom or my son it’s ok but also my internet or data is being used because it runs out too quickly

    June 26, 2019
    Reply
  2. KROOL said:

    It is relatively easy for other companies to make their own Watson like supercomputers with NVIDIA DGX-1 units. Watson is not that special anymore.

    June 26, 2019
    Reply
  3. mohamed Gougam said:

    Very interesting

    June 26, 2019
    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *