Experts from Accenture and AWS on AI in the Enterprise

hi I'm James Maguire and on today's e Peaks we're talking about cloud computing and artificial intelligence two very Dynamic emerging technologies that are impacting each other in many ways we'll talk about the relationship between cloud and Ai and how companies can make the most of these powerful Technologies to discuss that I'm joined by two major industry thought Leaders with me is Chris neerman managing director Global Systems integrators at Amazon web services Chris happy Friday to you happy Friday good to see you James uh also with me is Chris Wagman global technology lead of the AWS business group at Accenture Chris thrilled to have you with us today James it's awesome to be here thanks for inviting me so let's talk about cloud and AI in in January Accenture labeled 2023 as the year of cloud value which I certainly wouldn't dispute that and while cloud has continued its major growth generative a AI has really been the star of the show it's dominated the the conversation creating huge interest among Enterprise decision makers Chris W your thoughts on this what's the balance between cloud and AI these days what's going on here in late 2023 yeah I mean obviously you can't do uh AI without cloud right uh that that's you know a fundamental uh situation that we have so it's going to be bound together you know now and going forward uh you know you mentioned you mentioned the comment that uh you know we said it was the year of value um I've seen that continue right um and for a couple reasons one um you know customers are always trying to drive value and no matter what's going on in the economy and and and the macro economic situation that's going on you know being able to drive more value being able to take that that uh value and invest it back in your business and what you know with generative AI you know bursting onto the scene this year that value is being pulled and being invested in geni right there's just no doubt about it and you know I think also AI overall right not just geni right I think we've seen this Resurgence of AI or interest in Ai and the importance of what that's going to do and then obviously leading uh to the generative AI so value definitely is still is still key for our customers but where they're putting that money is has changed I think over the last year you know I think that one thought you made is not always understood by everyone the fact that we really don't have ai without cloud to support it I mean because AI gets all the headlines and we think about some of the Buzzy consumer applications but in fact without a huge robust Computing platform enabled by Cloud AI doesn't go too far um Chris and what would you add about about that where where are we in the generative AI maturity models growth yeah sure so so the other Chris and I Chris Wegman and I have talked a lot about how there's there's a lot of cost cutting that is the initial part of the conversation and I don't think that's and that's been for the last decade that I've been at Amazon and I don't think that's changed I think that uh it starts with cost cutting but our experience is that customer conversations quickly turn to how we can help them innovate and transform their businesses in an accelerated pace and as such we're seeing the next wave of of widespread ml adoption with the opportunity for every customer experience an application to be reinvented with generative Ai and so AWS is making it super easy and practical and secure and cost-effective for customers of all sizes and developers of all skill levels to use gener of AI in their businesses across all three of the layers the ml stack including infrastructure and ml tools and purpose built AI services and what a lot of people don't know or have forgotten is that Amazon's invested heavily in the development and deployment of AI and machine learning for for decades for more than 25 years and so for both customer facing services and internal operations um from things like the the recommendation engine that that personalizes the shopping experience on amazon.com to the AI powered robots that optimize order fulfillment in our warehouses so you know there's there's a lot to this as you know gener of AI maturity models continue to evolve you know I think it's also a thought that people don't realize that AI is not actually new at all and that you know chat GPT came in the scene in November 2022 and everyone realized oh there's this thing called AI it's going to change our world it's really not news you say aw has been you know investing in a long time uh let's talk about real world experiences uh can the TUV provide a real world experience of how companies can boost their Enterprise Readiness for new and evolving Technologies like today's Cloud Solutions are J of AI chrisen what what do you see here I I really think it's industry focused at least that's the way that I'm I'm working with my teams and internally talking to and and to customers you know based upon customer feedback and observations we're investing in developing in Industry specialized gener AI practices and solutions and that enables our customers and we always start with the customer work backwards but it enables our customers and neous teams know who to proactively engage for specific use cases and capabilities and so it starts with the industry and then the use cases in those Industries then how do we have the capabilities to actually help our customers achieve that the outcomes that they want and so generative AI powered Solutions have implications in every industry Healthcare manufacturing uh Financial Services media entertainment countless more applications that support speed to Market uh a reduction and cost savings so some specific examples of Industry applications of gener of AI we're seeing I'll start with advertising and marketing it's blog writing uh advertising copy creation stock image generation Market content creation Healthcare Life Sciences is a passion of mine so in particular there we're seeing medical take medical notetaking and virtual assistance solutions for doctors we had a great release back in April with 3m and ads with a partnership where 3M using uh ads machine learning and gener services like Bedrock Amazon bedro Amazon comprehended medical and Amazon transcribed medical to really help with 3m's ambient clinical documentation and their virtual assistant Solutions which essentially transforms the patient physician experience placing the focus back on the patient where it should be and reducing administrative burden for Physicians and and I can go on and on there's Financial Services with personalized conversational agents and synthetic data generation for broad modeling and and as Chris Swagman knows I can talk for hours so I'll pause there but we can talk about in depth on examples in every imaginable industry Ju Just One followup Chris and that I think you it sounds like you would push back on those people who say oh generative AI is really cool but it's really not ready for prime time from what you're saying or AI in general it's it's it's it's not only ready for prime time it's actually used in the field it it's it's and it is it is interesting I hear that quite often it's like okay so it's not right for prime time and then I go well it really is well it's not actually in use today it's going to be you know in the fullness of time I go no we have numerous examples where it is absolutely in use today with a number of customers real world experiences and we're continuing to evolve that but it's far more than I think people realize Chris W your sense Far More Than People realize what real world experiences would you point to yeah James you're going to realize when you get Chris and me together uh you know we're gonna we're GNA go on for hours on this um these are complicated topics so it's twe a nutshell is difficult they they are and you know you know Chris Chris new I I think about you talking about this it reminds me of the early days of cloud when you and I were together you know you know 10 15 years ago in in this industry starting to come up it was completely it's not ready for Prime Time Cloud's never going to be ready for prime time you know and we show you know use case after use case where our customers were already using it and it it took a while right it took a while for for other uh customers and other Industries to realize that it is real and it is ready for prime time so you know I I could not agree with Chris nman more um specifically around the industry Focus definitely industry the industry use cases are different they're going to you know they're meaningful they are in production today you know whether it be um you know um taking letters that come in and and you know synthesizing them and and and helping you know helping uh the people that need to respond to those letters right uh and and help them build that or find the letters that need the fastest response um you know over other ones we get customers to get you know still you know thousands and thousands of letters a day uh from you know from people that you know aren't used to having you know going online or or sending you know doing a chat or something like that but what I'm seeing is you know if I I take a step back and look at every every one of our customers um there's two areas that I see that all of them are are really going deep in in in Ai and generative Ai and one that is in software development right um almost every one of our customers has some type of development organizations they've got Engineers writing code you know they're they're trying to um do that faster with less errors less problems you know one to get gener of AI Solutions out there but just just in their day-to-day business as new pressures come to transform um so you know we definitely you know we're seeing customers uh get a lot of power and a lot of Leverage out of code whisper from AWS uh right to help that you know help with the you know software development uh be that second set of eyes what we call peer programming right having that two sets of eyes on that code they're developing at any given time you know and that's helping you know obviously generate the code faster better quality right uh better productivity and what we're seeing in that better productivity from the more senior develop right uh who are the people you want right really being very effective in in their in their work right because they've got the experience they know the applications you know they they can just they're just better at what they do and faster uh so that's great you know the other thing is just pure you know defects and looking for things like you know security flaws and stuff like that that they they introduce right code whisper is helping them um you know avoid that and which is taking time out you know deeper into the you know code supply chain so to speak um the secondary and Chris mentioned on it you know all almost all of our customers or all customers some have some type of customer interaction right whether it's you know internal customers right from a a help desk or or HR or something like that to external customers right and whether that's B2B or B Toc type customers and you know generative Ai and making better chat booot is what I call it right because everyone's got some type of chatbot having a a customer uh service rep be able to have the next best action have that call synthesized for them you know you know a lot of times if you're if I'm a call center agent right I'm I'm typing as fast as I can type to to you know look up what I think the answer is and you know because of generative AI because of AI right that's now being searched and given to me much faster on my screen giving me the response right now as a as a CSR I got to decide whether or not that's the right response um you know so you know those are the two use cases we're seeing almost every customer starting with uh in their in their G of AI Journey yeah Jam Just Add to Chris you made me think of a couple things like we've been at this a long time and I'm G to steal something from Andy Jazzy Andy but from the first day I joined in 2014 he'd say there's no compression algorithm for experience and I was thinking about what you know we've been at this for decades AI ml Amazon has we've been at this for the last 10 years with Accentra with the with the accent business group and that that just you just can't turn the ship and build that overnight so everything that we're delivering today you think about the velocity platform that extension was announced last year reinvent that's the culmination of the last decade and longer of work that actually helps us deliver Solutions faster than ever before and you just can't do that overnight again no compression algorithm for experience you know one one followup about the the software development piece uh I this Chris W said this you checked about the development are you seeing AI with a no code low code software development programs what do you see in that in that world or not much necess s yet no both I mean you know um you know even with no code low code right you still you still have to um you know take your inputs and bring it together like take the you know take the requirements from the customer uh and get them into some type of format so you can do low no uh no code low code uh development and you know that's one thing we're seeing you know for an example you know we do a lot of sap uh configuration right uh for our customers right um so how do you do that how do you go and and prepopulate those those documents that do that configuration right uh with generative AI right so pulling that data you know pulling the requirements picking the right the right uh configuration Flags to to set right that that that's huge that's takes a lot of people's time uh and there's a lot of crosschecking and that that can be can be eliminated so yeah we're seeing it there there as well as uh you know what I would call traditional uh you know code development you know one thing I'm seeing is um you know it's it's where we're seeing our Enterprise customers get the most value isn't generating from scratch right or having it hey let me you know go create a new homepage for me or whatever it might be but it it's truly helping them you know support the code they have today or add new features to that code uh which is where I think I I'm I'm a huge fan of of code whisper because you know I've seen it in my development team we use it uh in our platform the velocity platform that the C talked about you know we've used it quite extensively in that and I just see the productivity Improvement and the quality improvement that our developers have gotten uh by using it interesting all right well we've talked about this a little bit but let's drill down into this how are companies like AWS and Accenture addressing the today's emerging Tech like cloud and generative AI Chris W additional details on this yes so together is the key to me is the is the key start to that to that question um you know you know Chris talked about you know 14 15 plus years of partnership um you we were one of the first users of S3 uh Accenture was uh you know storing uh of all things uh you know receipts for our uh you know for our time reports uh that was an early early early uh use of S3 and you know now I look at that how that's evolved right the analysis that goes on on that unstructured data right on that just internally for us so you know first and foremost together um you know we bring together you know as you talked about that industry experience um you know we've you know both developed uh these incubation um groups to help customers take these use cases and what I find is there's no there's no um um there's no limitation on the number of innovative use cases customers can come up with but what we're finding is very quickly thosea use cases need to be looked at and value needs to be find be needs to be found in them right and the only way to find that value is to very quickly analyze them and start testing them and I think we're seeing that we're seeing a lot of customers started you know developing these um you know these different use cases and testing and stuff like that and now we're starting to see the ones that truly are driving value um you know starting to show up in production right and the ones that weren't going to drive value okay maybe generative AI wasn't the answer to that to that problem that I have as a customer those are starting to fall out so together you know together we're taking this industry Focus we talked about together uh we're uh you know we have this incubation team that is taking those using Bedrock using Titan using you know using the models that we can get access to using you know using the Amazon chips right to do those those trainings of that model and so on um so we've been very close in in that regard as well uh and you know together working with our clients um and and trying to make sure that they're putting their focus in the right places um so it's been a big Focus over ours last you know since we started but especially in gen over the last several months Chris what would you what what would you add here what's important to know sure I I think I I just want to continue off of Chris W's point about value creation I think with respect to Value creation that Chris mentioned I think it is all about value creation for our uh our customers and you know I think back to a a paper that mcken McKenzie published a year ago really great research that estimated that the total iida opportunity from cloud for the Forbes Global 2000 our customers is estimated at 3 trillion by 2030 that that's value that we can work together to help them unlock and I think it's critical that we we kind of shift it to the customer that help them understand there's there's a ton of value now let's figure out and dive deep like you said James how do we dive deep to actually help them unlock that and uh it is absolutely in the details so at AWS we want to make every company an AI company and there are four components to our approach to help customers leverage the full potential of generative AI applications to unlock that value we know first and foremost we no model choice is Paramount there's no one model that will rule them all it's about choosing the right model for The Right Use case then the ability for customers to securely customize these models with their own data is key and next easy to use tools are critical for democratizing gener of AI within a customers organization and increasing employee productivity that that's super important about increasing the employee productivity there's there's stats out there that say that it could impact 40% of the hours our employees uh our employees worked and so if it's on the mind of our customers that they have to figure out how to get the most uh uh most efficiency out of their employees we can absolutely do that here and then in underpinning all of this is the need to deliver responses that are that are low cost and low latency which is absolutely made possible with purpose-built ML infrastructure yeah the question I'm sorry go ahead yeah no Chris you know you you make a good point you know and I think that stats from the the department Department of Labor uh which I think is is is interesting because it's you know it's not it's not us coming up with that Stat or you coming up with that stat you know it is the government looking at how what the impact of gener a could be on you know could be on the labor market but you I also think about look at the number of conversations that we've had you know well over 600 conversations with CEOs and boards around the impact right of of what this stuff can do so it's not just you know it's just not the you know the you know cios and the the CTO and is having this conversation it's not a technology conversation right this is a board level conversation and I you know um you know no matter what size company you are or what you know what the uh board structure is I mean this is the conversation that's happening there and there is you know I think we all believe that you know this is a once in a generation you know uh transformation that's going on and you know I reflect and I think about you know we talk about all the great stuff that has to happen you know can happen with with AI and generative AI but it's all going to come down to the core systems that our customers have right what we call the digital core right and I think we're seeing more and more as as customers start going through this journey you know they they start having to look at their data is their data ready to support this right yeah I can use chat GTP or I can use you know whatever to get you know to get um um you know generic kind of answers to stuff but if I truly wanted to have my data I've got to put my data into those models right and and using bedrock and others so um you know that's that's probably something that most of our customers are are spending more time thinking about now I want to Circle back to something that Chris and said you're talking about models and I I hear so much about large language models and and companies grappling with how are we supposed to approach this I mean do we need to build our own do we need a fully proprietary one can we rent one off the shelf you know is it's it's the big distinguishing characteristic Chris any any advice you might give to companies about handling the issue of large language models some yeah I I absolutely think the right like I said the right model for The Right Use is absolutely critical and as Chris W said it's it's really their data and so regardless of the model you choose and we can we can absolutely help them with the right model um using their own data is absolutely essential and making sure that that data is secure to their environments and it's not you know not open for others and so I think that you know owning their own data using their own data is absolutely essential I'll add on the models I'm with you right I'm I'm a firm believer that it's not going to be a single model for our customers right there's going be multiple models there's going to be industry specific models uh some of those models are going to be quote unquote off the shelf or come with you know come with uh you know come with the solution that they're built into other ones you're going to build yourself right and you're going to you know you're going to like you said you're none of them I don't think anyone's going to use them just completely generic like they're going to have to put their data into them or they are putting their data into them to get the right the right value out of them and um you know I think again that's I like Bedrock a lot because of that right because it gives you that choice of those models you can use and and takes a lot of that heavy lifting away from the training which most of our customers are going to do right I mean there's going to be you know um they're gonna they're going to want that um that full Mo that full environment to build the stuff versus having to stitch it all together themselves well let's let's uh talk about what is the bottom line here what are your top takeaways for Clans to close the year 2023 in particular how do you see today's core Technologies like cloud and AI evolving in the years to come Chris double you your your forecast yeah um you know I think between now and then of the year it's it's about customers getting uh and clients and companies getting uh gener of AI models into production right getting the the use cases that they've been testing and uh getting those into production and starting to show the value of them right I think um starting to show you know we we didn't touch on responsible AI around this right making sure that those responsible AI um guard rails are in place for their companies right because they're starting to see this stuff become go into production and they've got to make sure you know they can they can uh Trace back the algorithms you know truly you know show show how the the uh answers are coming about but you know it's one getting getting those into production if we think P if we think further right we're early on like I think we always talk about you know we're in you know I'm a baseball guy it's the playoff season you know we're in we're in maybe inning one maybe inning two um you know I think in Cloud competing we're probably only in inning three right so you know it's still a very early game um so to speak and you know I think cost is going to become important right the cost uh the cost of training these models you know we all know that it takes a lot of energy it takes a lot of power it takes a lot of water to you know to do to generate these um these models so I think that's going to come very important to our customers as they try to drive more value out of this they're going to look for the you know the best cost uh performance uh indicators on this stuff I think um I think we're going to continue to see massive uh Innovation right we're going to continue to see you know everyone is investing heavily in this market and I think you're going to see new innovations that going to make it cheaper make it faster and may able help our customers get these uh use cases to Market faster Chris and a key takeaway and most important where are we going what's the future here sure as we're as we're talking about this one thing came to mind for me and we have our leadership principles Amazon does and one of them just struck out to me as we're talking through this which is uh success in scale bring broad responsibility and I think that when you think about artificial intelligence and applied through machine learning it's going to be one it is or will be one of the most transformational Technologies around generation tackling some of Humanity's most challenging problems and augmenting Human Performance and maximizing productivity so there there's a they a bigger picture here as we look to the Future and I think responsible use of these Technologies is key to fostering continued Innovation which is absolutely what our customers are asking us for is continued Innovation is committed to developing fair and accurate AI nml services and providing customers with those tools and guidance needed to build AI ml applications responsibly and and I think that again we owe that to the world success and scaving Broad responsibility especially with the Decades of our experience and what we've developed to date and it of us is focused on democratizing ML and making it accessible to anyone who wants it anybody want want to use it and currently we have over 100,000 ads customers leveraging our ml Solutions so together with the 8S partner Community we're now taking the same approach with generative AI you know the two of you I've got to thank you for sharing your expertise today a lot of good stuff and I learned a ton uh please come back and talk with us again sometime absolutely Thanks James thanks James anytime thanks Chris good see you Chris you too

This transcript was generated automatically from the video's captions and may contain errors.