Chris: Welcome to The Tech Lowdown show where each episode we’ll be discussing opportunities in the tech space. With entrepreneurs from the US and around the world I’m your host Chris Jones I’m excited today to go deep on an important area in the tech space. Artificial intelligence has been on the lips of the people in the know for a couple of years now, but we are starting to see it creep into products and services. AI plus natural language processing plus machine learning is impacting everything from how we control our homes to how healthcare is being delivered. If you’re serious about the future of tech then this is one area you need to be across. Our guest today has been at the forefront of technology for almost 20 years. Alex Tarasov is the COO of Scrum Logic, a development consulting company with the client list that includes Microsoft, T-Mobile, E-Trade and many more. They’re also a Microsoft venture partner, working with high potential startups. Welcome to the show Alex what’s the low down?
Alex : Hey Chris how are you?
Chris: Good thanks for your time today, we really appreciate it.
Chris: First Alex I’ll like to start with if you could share a bit about your background and how you’ve gotten the opportunity with so many interesting clients.
Alex: Well essentially we started working with Microsoft Ventures, which is interesting because everybody thinks that Microsoft Ventures was originally investment fund and really it wasn’t. Interesting story there is that about 6 years ago Microsoft was very interested in getting started Azure products cloud solutions and they went to Silicon Valley and Silicon Valley said we never heard of Microsoft here you’re not relevant to startups. You are Enterprise Company nobody knows you. If I’m a startup today how would I work with your company tomorrow. Do you have a ventures department that somebody can talk to or how are you on board? Is there a special program for startup that can help us kind of grow? What do you guys offer? And at that point they gave us serious consideration to what needs to happen next and the department of ventures was born. So the main point of ventures, Amazon has it pretty much every company today has ventures and CIOs, is an interesting story. It all started in 2000ish with the concept of destructive technology. Interesting fact that for the past 5 years I believe, there’s been a billion dollar company born every month that disrupts technology. On the startup world yea it’s pretty interesting stuff. So in early 2000 I remember when internet connection what was it. 1 megabit connection through your internet service provider and that was considered to be fast. Those are group of guys that were trying to watch YouTube, but they couldn’t get you know high definition YouTube videos uploaded and it took forever. So etc. so they developed an algorithm on video compression that essentially delivered high definition videos 720 1080p over really low bandwidth connection so the internet had one megabit was more than adequate to deliver that. They essentially turned into idea – why not go and deliver as a video podcast or streaming service kind of like you know Blockbuster they were the big player. The story there was they essentially came to blog buster and they said hey we have this we’re doing it we believe the Americans are lazy nobody really wants to get up go to a store get a DVD come back home. The CEO at Blog Buster at the time said no you’re crazy this is an American tradition where very old company would coin the phrase Blockbuster hit so we are mecca don’t tell us our business you know nothing of us and it’s an American tradition to go get a DVD. Popcorns on demand with a family kind of thing this is our cornerstone. What 8 years later Blockbuster went out of business right a billion dollar company that had rights to so many movies it was the only content distributer really. The major one at least and they disappeared literally in 8 years so this is what we called destructive technology. And Netflix were the ones that did it by the way they kicked Netflix out of there office and they laughed them. They said you’re crazy this isn’t going to work and the list goes on the concept with Uber right when was the last time you took a taxi? And some point some guy said, hey listen it’ll be interesting if person random person would come pick you up from your house and you give that person your address. Ten years ago you would have said no you’re crazy that would never work I’m calling a taxi. Today once again destructive technology and the list go on AirBnB and hotels in five years ten years from now you kids are not going to know what a DVD is. They’re not going what a taxi is or what a hotel is.
Chris: How’d you guys get your hands dirty in this interesting space?
Alex: We’ve been working with Microsoft quite a while on their what you call evangelist departments and we’re working with existing Microsoft partners and helping them grow so there wasn’t existing program for that. And when Ventures started we really started pouring into that because startups are the future. They are the ones working 10 years, 20 years ahead and it was a very interesting figure for us because we really wanted to work with next generation technology which kind of led us to AI, machine learning predictive analytics and things of that nature. So that’s where we ended up.
Chris: That makes a lot of sense…interesting, so you guys have built some great stuff and tell me about when you first came across AI and when you recognized that it’s something that could eventually be commercialized and why are we seeing it takeoff now as supposed to last year or 5 years from now?
Alex: When did Terminator come out, the first one with Arnold Schwarzenegger, that’s when I remembered that we’re almost there? And then with the AI world it was very interesting I’m a big fan of Ray Kirks well I don’t know if you’re familiar with him. He’s actually one of my idols and he was able to predict the future using his own calculative services which is his brain. Essentially he was able to predict that what a computer would beat a human being in chess in 1998. It happened in 1997 he predicted that in 1963. So using that kind of concept I kind of got involved with it deeper and deeper and AI today. The reason that Microsoft, Amazon, Google have all of these platform for open source, is because they want startups that are moving forward in the future or enterprise companies that are not that much different from startups. Because they have their own internal R&D labs that they spend millions of dollars on researching, I mean there’s so many things that’s happening simultaneously right now. I think there’s more things happening in the past couple years than they have been in the past century in technology. So the most amazing thing about this is that AI is now getting closer to the real Terminator kind of thing and I think Watson is closest thing out there today IBM. The reason all the companies like Cortana, Siri, Google and Watson they all exist today because they want startups too actually. Instead of building their own version of AI they can utilize there and eventually they’ll become a customer where they pay subscriptions to use their AI technology and that’s how companies grow. They don’t actually build them for no reason they build them for a reason that enterprise with SMB which is small medium and startup businesses will start utilizing their technology and pays like the subscription SaaS fee, software as a service.
Chris: Can you tell us how does AI fit together with natural language processing and machine learning? You mentioned predictive analytics how do those different areas fit together?
Alex: Well Alexa is the great example of that so AI today is that you want to be able to communicate with a machine instead of typing and waiting for an answer you can start naturally talking to it right. So natural language processing is actually very difficult because the way folks are very different and the audio quality has to be extremely high. And a lot of projects that we’ve worked on one of the major one’s we try to do with Microsoft is their conferences have a lot of speakers and the way they do it today and the biggest pain point today for them is that they record hours and hours of folks doing presentation and so and so forth. And somebody sits down with a headset and literally transcribes it right. They had to put on the head phones and open Microsoft word and just literally type up all the hour and hours of presentation.
And then what they do is they summarize it and then they take a picture of it and clip it. What we try to do is to replicate that same process. That typically takes well depending on the time on the conference or whatever it is takes about a week or two real times. We want the machine to record it, we want the machine to understand it and we want the machine to summarize it. Meaning highlight the most important points of the presentation and we want the machine to take picture simultaneously and post it live on all the social networks. That are relevant to the subject matter you know Microsoft build I don’t know if you’re familiar with that where Silicon open door, Silicon valley open doors conferences. So a lot of those takes place today and they spend a lot and they spend a lot of time and money on actually trying to replicate that. But the problem is the conference is loud there’s people talking all over each other and it’s with Bing services today or Google services today it’s really hard to capture. So I think the closest was Alex one of the closest ones that I’ve seen in the industry today and it’s backed by a company I forgot what their called. Truuy I think something like that.
Chris: So that particular application involves natural language processing you’ve got the AI element where it’s trying to pull out what’s important. And summarize that, you’ve got a machine learning aspect because it’s trying to figure out over time to get better, to understand both what’s coming in and deliver it out better. So you got a lot of pieces working through that are there any other industries or applications of Ai that you guy have worked on?
Alex: Yea one of the biggest things that well Ai is we’re working on many different aspect of Ai but the biggest one that we’re going to actually start making a shift towards is health care. So ability to learn and give an answer so to give an example of high level Ai that 3 years ago when hepatitis C was considered to be one of the most deadly diseases out there. Today it’s being cured as quickly as cold and this is in 3 years they couldn’t do that in 30 years. But using machine learning and predictive analytics and Ai technology they’ve been able to gather all the research that came about from literally all the labs around the world. Combined, analyzed and instantaneously gave an answer most likely to work and then they add focus and pushed forward in more solution to create a DNA alliterating. I think that’s how it works it will alter your DNA that it becomes resistant to the virus. So in other words those are the things we’re moving towards right now is helping combine. There’s so much data out there and medical research that if combined together we could find cure for literally almost deadliest diseases out there. This is what Google project life is kind of focusing on next generation of that using AI with records while writing it. Which I think health care should be our primary focus as human beings we have this lazing capability now to do things that we couldn’t do in the past hundred years, thousands of years. So we could move forward space exploration technology, building like EM drive technology that’s also amazing electro magnetics.
Chris: Interesting so there you mentioned some of the AI platform providers you’ve got Microsoft with their Cortana, Google, Amazon and Alexa and IBM Watson. Can you help us just at a very high level how they differ from one another?
Alex: Well Watson is number one that was here in the background she didn’t like the Watson answer. Right the way that Watson is the one that’s able to talk back intelligently right so it’s known to be the mecha of them all. Sery, Cortana their still kind of catching up just because Microsoft, Google all those guys really ripped into that Ai space. A bit later where IBM is has been on it from very start as soon as they had the capability of doing so, because Watson was existed 1997.
Be as part of the best dress player in the world and he ministry in chess. So that kind of technology has been around for a while but they’ve been there first so how they differ is that right now most the everything except Watson when you talk to it, it just gives you a solid answer. It’s just like a bot system it’s not so much Ai, whereas Watson is able to process and we can have a conversation with that person. It’s like almost talking to a person right the rest of them kind of fetch data.
Chris: Got it so if I’m an entrepreneur and I’m looking at the different plat forms the highest and also the most expensive right now is Watson, but depending on what my purpose for using it is. I might be able to utilize different platforms rather than IBM and still deliver out a service that I’m looking for is that accurate?
Alex: Yes and no, so when you starting rubbing out to the Ai point is typically you’re probably going to be on like phase 10 of your project. So beforehand you have to build out your X and Y you create your back hand. You have to start building some sort of machine right now for that analytic platform and then Ai essentially takes over. So Ai I think a lot of people don’t understand the difference between machine learning, and predictive analytics and Ai. Machine learning is you teaching a machine you do and you automate the process to do it for you. So you know life for instance the way I conduct research on certain things I go into Google I types this in and I look for these key words and then I record it and once I find it. Whenever I put it into this database and I record it and I associate it to that. So that’s like say you’re work load for the day. So instead of you doing it I can teach a machine to do it but what takes you 6 months to do I can get the machine to do in like 6 minutes simultaneously it’s speeding up the process. Predicted analytics concept comes from learning data that you have today matching it to historical data and then predicting what the outcome would be. By using the most unbelievable concepts you never even took into consideration. Like the weather whatever it is like oh 6 years ago where they was the same way and it finds that pater of how well did they do and can predict what the outcome most likely would be very close to where it should be. Then you have Ai now Ai this is where it gets really tricky Ai is a cognitive ability so it’s like a human brain, it’s an ability to learn but also start making its own learning process. A machine that teaches itself where’s machine learning you teach constantly it never learns on its own it repeats your steps. Ai will repeat those steps go back analyze it and say you know what there’s a much better way of doing this I’m going to figure it out. and like as a human being that starts innovating the thought of going ho instead of me moving a box from point A to point B. I can create a wheel and put it in a cart and pull it over and now I can carry 10 boxes over. Right so Ai works exactly in the concept and idea of it is to replicate that exact cognitive abilities for a computer to be able to start thinking on its own and creating solutions on its own. And putting things together and it’s like growing a baby, eventually by time it gets older you can have so many thought processes it’s unbelievable.
Chris: That’s interesting great well I want to move to what call the real low down segment of the show and here just a couple of fast rapid fire questions really want to dig in tell us the high the low. Give us the real deal; I want to start with how are businesses screwing up with Ai today where do they mess it up?
Alex: A lot of them need to start talking or seeking or creating CTOs or talking to consultants who understand it and also it’s not so much. Are you asking whether it’s on a technical or business side?
Chris: You know what it could be either, I think most of the mistakes are happen on the business side is the way they think about the technology at this point.
Alex: Well I don’t want to be the guy that tells the kids there’s no such thing as Santa clause but unfortunately at times I have to.
So eventually the concept with in startup is never drink your own cool aid first of all right. Don’t be convinced that you need this technology and everybody would love it and blah, because you would be very surprised what the real world would give you back. So that’s why you create kind of like what Facebook did, you create an MVP and minimum value proposition. Create 4 features 1, 2, 3 features grow out and create a proof of concept in other words ask people or your potential customers clients who ever they could be and say hey we’re about to build this technology would you use it. Yes would you it, would you pay for it? Yes, would you give me one dollar right now if I can deliver it to you? So you have to create a monetary value out of that product right. So a lot of startups assume that they are going to go out and this market they’re going to build this. They start fund raising seed money precede money convertible notes. Get themselves into debt and could even an amazing technology it’s just not browsing in the field or becomes too expensive for its potential customer to use. And it doesn’t deliver much value and so they spend a lot of time like next five years kind of like push it through I don’t if you ever seen shark tank Mr. Wonderful is right a lot of time when he says “Hey listen I don’t want to be the bad guy you’ve been at this for like 5 years and you have 0 traction I think it’s time to put the old yeller down”. So yes that’s where I think the biggest thing is Ai technology is extremely expensive but I do encourage a lot of folks to start taking advantage of programs that IBM Microsoft amazon do offer. Like for instance Microsoft has a Bismarck plus program that offers up to half a million dollars in Asian credits. Once you become high potential kind of startup which is what we deal with and you start getting traction in the market. The whole purpose of Microsoft is to help you grow and they won’t equity for it. IBM will not take equity for it the reason they don’t want to drain you what they want to do is help you grow in the industry. And once you start growing you start consuming and that’s how they get their rewards right so you start paying them monthly fees to use their software. And some of them like Uber racks up I think 12 million a month at this point.
Chris: So just a couple more questions for you are there any freedoms coming back to your terminator example. Are there any freedoms that we as consumers are giving up to cooperation’s or startups so employ this technology? I mean basically should we be afraid?
Alex: Absolutely not I think that a lot of folks who don’t understand it fear it it’s kind of one of those saturations Ai is designed and machine learning predict is design to help us move forward and look at it from medical perspective for instance. We’re using this technology to help humans live longer, better and more quality lives. We’re going to use this technology to help prolong and learn the larger scale things we have never been able to learn before like for instance your release FTA approved Company XYZ. Company XYZ has now let’s Tylano plus whatever that is right and a lot of folks all of a sudden starting to have allergic reactions to whereas before they didn’t. The tool can monitor social network that offers up all this information for free every single day, because a lot of folks go to Web MV and describe their experiences so and so forth. We can capture all of that information instantaneously and stop something that shouldn’t be out there. Instead of waiting until 1000s of people suffer or die our cause irreversible issues. We can stop it right away create a red flag and the DA can go out and say hey Tylo plus is no longer aloud on the shells. Instead of waiting for placation law suit because the law suit companies will benefit from it from it in one end maybe not so much another on the financial but it’ll at least save lives. So the whole concept of having all of this technology is to help us move forward and cure incurable diseases and do a lot faster. And when you’re waiting in a hospital emergency room and they do surgery and machines do the surgeries most of the time now. So I don’t see the down side at all, like the terminator thing will not happen.
Chris: Alright folks you heard it first from Alex you’re not going to have a mutant machine come after you with a machine gun that’s good to hear. Alex I want to thank you this has been super enlightening can you please tell the listeners how they can find out more about Scrum logic and possibly get in touch with you guys.
Alex: Sure Scrumlogic.com is the website you can just reach out to us through there and if you have any questions you can send me an email at firstname.lastname@example.org.
Chris: Awesome. Folks if you like the show please download all the episodes and leave a five star rating on Itunes. You can find show notes at techlowdownshow.com and of course you can follow me on Twitter @cjones2002.