Business
Highlights:(0:00): You can't just throw an AI at it and go tell me the best price. Yeah, you have to bring in your human intelligence(0:08): Don't fall for hype. Start with an actual business problem that you have a reasonable understanding of and where automating would make a better experience. Or a better product.(0:21): Because we're not replacing humans, we're actually just giving them another tool to make them better at their job, because you still need a human.(1:00)Introduction: Colin is a sought after public speaker, thought leader and a firm believer in database decision making and applying automation to improve customer experience. With more than three decades of experience in financial services, Colin has held a number of C suite roles where he has championed data science initiatives in banking, FinTech insurance, risk management and marketing. Colin is the VP AI strategy for DataRobot, where he advises businesses on how to optimise their business operations using AI and manage data science projects. DataRobot is headquartered in Boston, Massachusetts, and has office locations across six countries. DataRobot enterprise platforms are used by Kroger, Humana, US Bank, Ben Kister, Australian Red Cross Blood Service, Boston Red Sox, lending tree, University of Michigan and a whole bunch more. Their platform helps steward health save $18 million. And their platform also helped a global bank financial services firm realise a 304% return on investment that amounted to about $33 million. Colin is passionate about how humans and AI can work together and is an advisor to the Singapore government on AI ethics, AI governance and the future of work. Colin is passionate about the science of healthcare and does pro bono work to support cancer research.(2:28)First of all, let's get to know you. How did you get into this? What was your journey like in your career up to up to this point.(2:42): Colin started in this field when it wasn't called AI or data science. It didn't have much hype back then. He shares his experience on how marketing professionals were more convincing than him. He learnt an important lesson from the book Sunsoo: The Art of War and he changed the conversations in the meetings to facts rather than who is more convincing. And step by step he did projects in marketing, operations and then healthcare and banking. Colin says he hasn't looked back since then. (5:06)Like, you know, AI, and there are so many terms machine learning AI data science. Can you help us unpack these terms for a layperson who's not a geek like you and I? What do these terms mean and how do they apply in different business scenarios.(5:43): Colin Priest defines buzzwords in simple and understandable definitions. Colin starts with data science. He says, Data Science is how you manipulate data to make better decisions. The next term he defines is machine learning. He defines it as- You give the computer some examples, with the correct answer, and you get it to figure out for itself, how to get from the inputs to the outputs. And the last term Colin defines is artificial intelligence. He says that there's a lot of people that haven't separated science fiction from fact. Artificial intelligence is merely a computer, doing something that historically required human intelligence. But that doesn't mean it has human intelligence. (8:20)Pricing is something, you know, we have learned in basic economics , like supply and demand. So, can you tell us how we can apply data science to optimise the price such that we get the best supply? Sorry, that's demand for the product or service, and we get the best price so that we have more profitability?(8:59): Colin says that we can't just throw a computer program at a problem and expect optimal results. Human intelligence is often needed as well. The best results often come from combining human intelligence with AI. You need to use experimentation to see how humans respond to different prices. Colin says that marketers are not always looking after your best interests. In fact, they're often trying to learn things about you so that they can sell things to you. He gives insights about a pricing strategy that supermarkets use. Supermarkets change their prices not only to satisfy customers at different times of the week, but also to learn about them as they can study how customers react to different prices. He says that you can enhance AI by once again allowing it to access and integrate with human knowledge. (10:46) And one of the major sort of, you can call it hurdles, or you can call it the first prerequisite of data science or AI is collecting huge amounts of data. Right? That is where, you know, I've gotten into some discussions with business folks. And, oh, can you not just use algorithms to figure it out? I'm like, okay, you need historical data for the machine to learn what the heck is going on, Right? So what is your take on things? People, specifically, C suite executives and decision makers, how much do they understand about how this works? And how do we collect historical data? And then we make decisions for the future? (11:41): Instead of getting bogged down in technical details and tuning hyperparameters, focus on the end results. Colin provides them with a framework for how they can estimate the benefit and estimate the complexity. Colin says that don't focus on the technology, focus on what problem you're solving. Don't get bogged down in complexity, especially if you're only starting out. Colin says that COVID has unveiled the frictions in the customer experience. It's a win-win. You are helping customers more. And they're buying from you more because you're giving them a better product. Colin advises- Don't fall for hype. Begin small, and start by solving an actual problem you have using AI and automation where you can measure the results.(15:55) I learned by going through these projects, most of the time is spent in AI projects in collecting data, cleaning data, collecting data. And, you know, the executives, they’re like what the hell is going on? Like, you know, it's been like weeks, you guys have not done anything but collect data? Like, what are you looking at all these, you know, numbers floating around? What is going on? So what is your take on that?(16:33): Colin shares that people love to jump straight to the complex, but it's important to lay a solid foundation first.So rather than just rushing to the result, it would be better if you explained to the executives on how you could modify the data collection process to improve the results obtained next time. Colin recommends a validation step to data scientists. Colin shares a project he worked on in the past. He found more than a dozen different ways they'd spelled Toyota. He shares that persuading the business people and executives to invest in boring stuff is hard because they want to spend money on grand stuff, and we have to make it clear to them that they link together. He explains this further by giving an example and linking it to AI and data. Colin recommends to start with a simple problem and use a simpler solution as it will need less data.19:33: MIDROLL (20:25)Most people think AI is more about numbers, and, you know, just just crunching numbers, but AI is getting into more humanistic fields like computer vision and recognising the audio and in some cases, AI is even composing music and art and whatnot. So tell us a little bit about what are the possibilities of AI actually breaking into these fields, which are closer to humans that they can understand, they can talk to AI, they can feel it and touch it and all that?(21:00): Colin explains with examples that with AI, people are never "replaced"; they are always empowered.The goal of AI is not to replace humans but to make humans more efficient at what they need to do. Don't be worried about machines taking over your work. Instead, think of what you can do with AI to make yourself better at what you're already doing. Colin is interested in AI because it can free up humans to do more meaningful things. It can also allow humans to do better at what they're already doing, so that's pretty cool. He says we are improving the world, simply by allowing computers to do what they do best, which is mathematical calculation and logical deduction.(23:40) So, you know, bringing it back now to do once again, you know, this notion that AI can take over it, because it can become like a terminator kind of a thing where, you know, it can destroy humans, or basically just take over the human race. What are the possibilities of that, in your opinion?(24:00): Colin explains that banks use more computers than they did before, so now they have ATMs. Banks still have more staff than they used to, though. And so, Colin doesn’t think we're in any kind of danger from killer robots taking over the world.Colin defines AI as just a computer system, a tool to do what you tell it to do. It is not autonomous. AI is a collection of algorithms, not a super-intelligence. It is a tool that you need to control, and it's important to know what those algorithms are doing. Humans control AI. Colin explains with an example that until we don't give AI a goal of destruction, it's not going to do it. So it is important that we use AI for good things rather than destruction. (27:20)We are under lockdown, the work environment is changing, and we are working from home. How do you think the role of AI will accelerate the adoption of digitization or remote work or anything like that? How has this experience in 2020 changed the trajectory of AI and its adoption?(28:00): Colin explains that a lot of companies have been struggling during the pandemic because they didn’t modernize their businesses. Some are now finding a guiding solution and are starting to adopt it. They gave their business a few weeks to get ready for the next revolution, but some timeless experts were prepared for any eventualities. Colin says he loves the software that changes the background which helps him hide the mess in his house. He explains that because of the pandemic, organizations have realized that face-to-face meetings and paper contracts are not enough when it comes to doing business. It is a natural fit that AI plus digital can solve most problems with digital contracts and AI monitoring.(30:06) In your opinion, what C suite executives and people in general may have to unlearn about the future of work about how they use technology or how they even look upon technology? What kind of new paradigms the C suite executives will need to learn? Now that you know, we have adopted this change. What is the future of work according to you but for now, what do we have to unlearn as a result of this? (31:20): Colin says that one of the things we have to change is the belief that AI is there to replace human staff. He explains further why human face and interaction is important for brand value.Artificial intelligence won't do everything for us. As humans, we are hardwired to connect with other people when we're emotional or confused. It is impossible for us to have any decision making process that is not biased. But, it is easier to reduce bias in an AI than in a human.Another thing to unlearn that you can get rid of bias. No, you can't. What you can do is prioritize what to get rid off. Colin believes that you cannot develop AI in a vacuum away from humans, it is there to help us.(35:46) Now, one of the last questions, oh, what do you think will be the future of work here now, like, you know, remote work, obviously, I think it'll stay? Well, what kind of other things you can foresee, as you know, that will change how we think about work these days.(36:04): Colin Priest uses the concept of comparative advantage and explains that humans are better than robots at things like empathy, creative problem-solving, general knowledge and common sense. Those things are our strong suits. Things that require that sort of human touch. Colin is optimistic about the future of work. Machines will make humans more valuable, and we'll be able to spend more time doing the things we love. Machines can't be taught ethics; it's just not in their programming. Human beings must determine what the machine will do. Colin believes that the impact of AI on work will be positive, but we need to be careful not to let it make our work lives less human.(39:20)And so if somebody wants to reach out to you to connect with you, how can they connect with you?(39:25): Colin shares some ways the audience can connect with him. He says, “You can find me on LinkedIn, there aren't that many Colin priests in the world and my picture is there. So just remember to look for the Colin priests with no hair. Find me there. You can read some of my blogs on the data robot website. You could read some of my personal blogs on my personal website www.colinpriest.com.”Resources:Website: www.colinpriest.comLinkedIn: https://www.linkedin.com/in/colinpriest/Data Robot’s website: https://www.datarobot.com/ ツ CONNECT WITH ME ツLeave a comment on this video and it'll get a response. Or you can connect with me on different social platforms too:• LinkedIn: https://www.linkedin.com/in/manujaggarwal/ • Facebook: https://www.facebook.com/manujagro/ • Website: https://manujaggarwal.com/ See acast.com/privacy for privacy and opt-out information.