 Welcome back to theCUBE, live here at Google Next 23. We're so excited to be here. I'm Rob Strecce along with Dustin Kirkland. We have some other hosts that are floating around here as well, Lisa Martin and John Furrier and the SiliconANGLE gang getting all of the stories and bringing it live to you from the floor. We're here back on our set. We have a really exciting customer and partner of Googles that are here with us. We have Dave Sy who is the VP of Enterprise AI for Toyota and we have Gopal Srinivasan who is the alphabet Google Alliance Generative AI Leader and Technology Industry Gen AI Leader at Deloitte. Welcome to theCUBE and glad to have you here. So why don't we, before we get into some of the questions and get talking about things because it's super interesting. Why don't you guys introduce yourselves first? I'm gonna go with you Dave first. Great, thank you. I'm a pleasure to be here first of all. So my name's Dave Sy. I am the VP of Enterprise AI at Toyota Motor North America. I also serve as a VP of Engineering at Toyota Connected North America. My two roles are vastly different but they all tackle the AI function for Toyota both in the enterprise space where it's gonna impact and touch every single pillar of our business as well as our connected vehicle side which we've been in the AI space since 2017. So really exciting in the recent developments as well. Oh, I'm sure. I'm sure this is just, I mean AI, AI, AI every day. So but what are you doing? What does it mean to be the lead for the alphabet? Alphabet Google Alliance Generative AI? No, certainly great question. And we've been a long standing partner of Google Cloud and have been on the journey as they've grown. And as Generative AI has taken off, we've had the opportunity to work with them over the past year to shape some of their solutions. And now we are working with them as part of our alliance to bring Generative AI solutions to market and work with our mutual customers like Toyota and others. So my role is to make that business opportunity between Deloitte and Google come to life for both sides to ensure that we are bringing value to our customers and apply Generative AI solutions. So I'm super excited about this interview as a multiple time Toyota customer myself, two Toyota vehicles in the garage, proud to say. And I know you guys have bet early and often on Generative AI. Tell us a little bit about how and where you're placing those bets and how that ties through Deloitte back to Google. Yeah, great. Thanks for sharing that, Rob. Thank you. First off, thanks for being a customer. Secondly, my two roles are again vastly different in terms of Generative AI and AI in general. On the connected side, we deliver advanced AI solutions through the custom voice agent that we built and run in-house. And we're gonna continue to use AI and Generative AI to advance that product for customer personalization and advancing customer experiences in the vehicle. We do multimedia design as well and we provide safety and safe routing for our customers and convenience routing for our customers as well. In addition to that, from a Toyota motor perspective where the enterprise AI function comes in, we're partnering with Deloitte to help us with our supply chain transformation. It's a really, really big effort. I think supply chain has been in national conversations throughout COVID as we move out of COVID. It still is a big transformational conversation for us and we feel that Generative AI will play a very, very important role within that space. Yeah, I think there's been things bubbling up out of this and how do you see it from the Deloitte side? I think broadly we see three categories of opportunities. One is in the core of the product where you're enhancing the product's features and capabilities with AI. And I think we hear a lot more about AI in digital products but AI in physical products are now going to come to the fore and automotive is actually a domain which is probably the furthest ahead when it comes to AI in a physical product. The second is customer engagement. So everything from hyper-personalized marketing to delivering AI-driven customer experience to predictive and proactive service. These are aspects where there are significant use cases and again as you think about where mobility goes next in terms of personalized in-vehicle experiences these types of use cases are going to come to the fore. And then the third are largely productivity centric within the enterprise and business operations as you're looking at spaces from finance to HR to supply chain. There are significant opportunities where you can apply AI and Gen AI to improve productivity. And in these three domains, we've at least I think encountered over 200 use cases just in the short period of time. And we're seeing a ton of both experimentation as well as use cases starting to go into production. That's fascinating. Another question for you on the Toyota side and I'm a product manager. I know that science and product management has a very distinct history tied to Toyota product management. I'm curious how product manager at Toyota is thinking about that next generation of driver experience and reinventing that experience with AI being a cornerstone new feature to be added. We tied that back to our core value, right? One of our core values, Toyota values is customer first and understanding the customer and collecting the data signals from surveys, market surveys, market data, research in addition to that, helps us inform from a product perspective. Can AI help with that? We just saw a product that was announced today regarding Jira integration and Confluence integration. Our research gets fed into tools like that and we should be able to do summarization and research based on all the data that we collect. In addition to that, we have vehicles on the road as well where we're understanding how the customers are interacting with our vehicles on a day-to-day basis as well. Those are key informants to making product decisions on whether to put in a knob or to go digital or to do a little bit of both, right? Which is what we're seeing as well, right? Yeah, that makes total sense and Bill Paul, what are you seeing from your vantage point and why Google, I guess? Let's bring it back, we're all here, what? No, I think that's a great question and I think if you look at the companies who've really brought AI into everyday products and services that we use, Google has been at the forefront. While generative AI is certainly the newest evolution of what AI is, AI as a discipline has existed for a long time and if you really think about what does it take for AI to be effectively put into use, one, it requires more than just a generative AI or a large language model, it's going to have to work with other forms of AI and Google has multiple AI solutions from doc AI to vision AI that have existed for a long time and continue to improve. Second, to string together all of the various workflows, applications, data that you have in an enterprise, you need some sort of a platform where you can actually build the application and then bring the generative AI models to work with those data within the applications and Vertex as a platform really offers all the tools for anyone who's trying to build an AI application to do it within the environment. And the third is really about doing AI in a responsible manner and that requires your ability to put in governance, controls, have traceability and these again are things that Google is leading with, your data is going to be private, the base models are not going to get trained on your enterprise data. So I think these three things are what distinguish how Google is approaching it and really make them stand out from other players. Now Dave, you're an engineer, you measure everything, I'm sure. How are you measuring the ROI where you've applied generative AI or other AI applications into your workflows? Are you able to tangibly see the benefits? Yeah, Gopal likes to go in threes, I'm going to go in twos, right? So I'll give you two examples of how we're measuring productivity from an engineering perspective. I have several hundred engineers that Toyota connected. We piloted code assistant capabilities such as Kodi, co-pilot and other tools and we are experimenting with that and we've seen a 30% increase in productivity from our engineers to date. Is that output, is that lines of code generator, is that features implemented? That is survey data from the engineers themselves on the number of hours safe per day for either generating automated tests, documentation or understanding the code basis out there. The other example that I'll give you is something that I think many of us here are grokking. The distance between an actual intelligence, business intelligence and decision report from an idea is pretty far and many companies out here, many enterprises actually are in an outsourced model. So what really happens there is business will talk to IT, IT will talk to an onshore manager, onshore will talk to offshore manager, then the engineer sees it, that's five steps. A lot of times that's maybe two weeks, right? And you rinse and repeat steps one through five, you're lucky if you get time saved with that. But AI actually, Gen AI actually gives us that capability for us to speak naturally in the English language to find business intelligence so they can make decisions. They're effectively translating SQL code or regular query code into something that the machine understands and then output some business results that they can act on. No, that makes sense. And that's, I think, a good productivity boost as well. What other productivity boosts are you seeing out there? What are other customers taking advantage of from a Gen AI? Yeah, I was going to give you three examples but dates are one of them, sorry. Now you're down to two. We're only going to give you two. Yeah, I think one very easy first out of the gate set of use cases are we have multiple studies that point to the fact that within any company, workers spend anywhere between 30 to 40% of their time just finding the information they need to do their job. And so everything you need in terms of answering a question directly, I'm trying to find which, what's the last time we use this configuration instead of having to go through 20 documents to find it, ask the question, get the answer. Summarize this document for me, I'm trying to find what a policy is actually saying, just summarize it for me so I can get to the answer quickly. So there's this whole set of find me the information faster, get me to my output quickly. That's one category of use cases we're seeing in virtually every function, be it in HR, in customer service, sales, finance across the board. The second category we're seeing also is document generation. I think some of it Dave spoke about in the context of code, but every organization also produces a ton of handbooks and other forms of policy documentation, contracts, all of which you can generate the first draft and get start with a 30 to 40% level of completion before a human even lays their hand on it. So these two kinds of use cases are seeing significant adoption in the enterprise today. So we'll step aside for a non-technical question for a second, maybe Gopal first and love for you to weigh in too, Dave. Ethical risks and challenges. As you advise your clients, Toyota and others, is this something that you think about at Deloitte and you've got opinionated views and advice on? No, absolutely. We think about it all the time. We wrote the book on it. We call it trustworthy AI and that's a framework we've defined to essentially outline how organizations bring AI to their customers and the ecosystem they operate in. One has to remember that inherently AI is predictive. So this is not about trying to turn AI predictable and that's not an outcome you're going to get, but how do we bound it with very low degrees of variation so you can manage the level of risk you're working with? And there are seven principles to what make up trustworthy AI. Not three. Not three, seven in this case. I didn't define this one. So the first is transparency, which is you are transparently communicating to the recipients and the users of any product or service you have out there that you are using AI and this is how you're using it, including how you're collecting customer data and what it's being used for. Second, reliability that the degrees of variance are low and you can reasonably expect that applying AI in any product or service will result in relatively reliable outcomes every time you're applying it. Third is traceability that when AI makes a decision, you can trace back to what were the inputs that went into making the decision and why the AI model made the choice. The third is fairness, which is you are being fair in how it's being applied to people of all backgrounds and in all situations and you're not getting biased because of certain characteristics. Then the fifth would be privacy, which is the data that you're using to train and run the AI models are private being used solely for that purpose and are not being misused or applied in other ways. And then the last two are really, one is responsibility and accountability, that there are clear lines of accountability and responsibility within the organization that when something doesn't go to expectations, there are people on point to identify, fix it and make sure that you don't repeat the same thing again. And then the last is this notion of, AI should not be applied to cause digital or physical harm. So those are the seven principles, but keep in mind that the principles alone will not let you do this scalably in an enterprise. Then has to be translated into how you train your people, what are the policies you put in place and how you use technology to enable this at scale. And when you think of this in the areas that we've applied, I believe that automotive is actually a domain that's very advanced and while we continue to have a lot of debate, including in this city in terms of whether autonomous cars should be allowed to ride around, I have personally experienced it and honestly, I think this is one of the domains where we are the furthest along. And Dave, I would love for our audience to hear from you on what within Toyota but broadly in the automotive industry you're doing to ensure that we're bringing a responsible approach. Yeah, absolutely. Absolutely, Gopal. I think very well said, right? Let's see, I think Gopal you very well highlighted those principles and we follow all of those and what I want to say there is great minds think alike. Our enterprise AI group in partnership with our data group and our security group has set some of these standards internal to the company, very, very similar themes around trust, safety, transparency, low bias or no bias. And our thoughts around that from an enterprise applicability perspective is really focused on compliance and governance. You can have these great policies if you don't follow them, right? They're really no good at the end of the day. So we have a responsible AI board and what I want to share from a Toyota perspective, this is very, very top of mind for us at the executive and the executive committee level and our leaders are very keen on moving fast with AI but moving fast very safely as well, all right? And we're seeing this pretty pervasively across the automotive industry as well. As Gopal said, the autonomous driving use cases and some of the moral dilemmas that are there as well are very challenging, right? No, it totally makes sense. And I think that is probably a really good place for us to leave off for this and I want to thank you both for being on here. It's been great, very insightful. I think that you brought a lot of knowledge to this practical uses of it. And like you said, the car industry has been way ahead of this for quite some time with language models and all that, especially voice recognition and everything else that's been in the cars for quite a long time. So I want to say thank you for being on here. You both were great. Thank you, I really appreciate it, thanks for having us. For theCUBE and for myself and Dustin, we'll be back here from Google Next 23 with more content, closing out to afternoon really after a really short break here. Happy to have you here, hang in there. Take care.