The Data-Driven Enterprise: Transforming Business in the Digital Age

(DESCRIPTION) 
Joan Woodward in a video window in the upper right corner. Text, Wednesdays with Woodward (registered trademark) A Webinar Series. The Data-Driven Enterprise: Transforming Business in a Digital Age. Logos under the text, MIT Sloan CIO Symposium, Travelers Institute, Travelers, Georgia Tech College of Computing, Northeastern University Khoury College of Computer Sciences 

(SPEECH) 
Good afternoon, everyone, and thank you for joining us. I'm Joan Woodward, and I'm honored to lead the Travelers Institute, which is the public policy and educational arm of Travelers. Today's program, which is going to be terrific, is part of our Wednesdays with Woodward, a series we started last year to explore issues impacting your personal lives, your professional careers, and our businesses in these very uncertain times. 

(DESCRIPTION) 
Join our mailing list: institute@travelers.com. Indeed Connect: Joan Kois Woodward. Watch Replays: travelersinstitute.org. hashtag WednesdayswithWoodard 

(SPEECH) 
So we're pleased you're here with us today. We hope you'll stay engaged with us, and you can do that by joining our mailing list, connect with me personally on LinkedIn, we'd love to have you, or watch our replays of the past 25 webinars we've had last year on TravelersInstitute.org. 

(DESCRIPTION) 
Wednesdays with Woodard is an educational webinar series, presented by the Travelers Institute, the public policy division of Travelers. This program is offered for informational and educational purposes only. You should consult with your financial, legal, insurance or other advisers about any practices suggested by this program. Please note that this session is being recorded and may be used as Travelers deems appropriate. 

(SPEECH) 
So before we get started, I'd like to share our disclaimer about today's program. We sit here today over Zoom more than a year after the pandemic did begin, which has challenged all of us in countless ways. From a business perspective, employees went remote, storefronts became e-commerce platform, retail sales went through the roof, doctor's visits went virtual, and consumer and business expectations changed in really profound ways. It's really incredible to think about the acceleration in data and technology that we've seen over this past year. 

(DESCRIPTION) 
Photos and text, Joan Woodward, Executive Vice President, Public Policy; President, Travelers Institute, Travelers. Mojgan Lefebvre, Executive Vice President; Chief Technology & Operations Officer; Travelers. Dr. Tom Davenport, Author, The AI Advantage; President’s Distinguished Professor, Information Technology and Management, Babson College; Co-Founder, International Institute for Analytics; Fellow, MIT Initiative for the Digital Economy; Senior Advisor, Deloitte Analytics 

(SPEECH) 
So today, we're going to hear from two leading authorities and thinkers about the power of data and how organizations can harness the explosion of data-driven capabilities to help meet our business priorities and customer preferences. We've been flooded with requests to cover this topic on our Wednesday session so we're really thrilled that you're here today to lean into it with us. So now, it's my honor to introduce our terrific speakers. First is my colleague and friend, Mojgan Lefebvre, who serves as our Executive Vice president and chief technology and operations officer here at Travelers. It's a very big job, folks. 

Mojgan joined the company in 2018 and leads the combined technology, data, and analytics, cybersecurity, and insurance operation functions globally. Under her leadership, Travelers has won the Gartner Eye on Innovation and the CIO 100 awards for innovation and transformation two years in a row. She was recently also honored as Forbes CIO Next 50 Innovative Technology Leaders. And I think, Mojgan, your picture was actually on the cover of that magazine this time around. 

Mojgan Lefebvre:  Much to my surprise. 

It's a great pick. Mojgan earned her computer science degree from the Georgia Institute of Technology and her MBA from Harvard University. 

So next is Dr. Tom Davenport. Tom is the President's Distinguished Professor of Information Technology and Management at Babson College. He's also the co-founder of the International Institute for Analytics, also a fellow at the MIT Initiative for Digital Economy, and a senior advisor to Deloitte Analytics. He's written or edited over 20 books and over 250 print or digital articles for the Harvard Business Review, Sloan Management Review, and the Financial Times, and other publications. He earned his PhD from Harvard and his latest book, The AI Advantage, How to Put the Artificial Intelligence Revolution to Work. 

(DESCRIPTION) 
Book cover, The AI Advantage, How to Put the Artificial Intelligence Revolution to Work. Thomas H. Davenport. Text, Complimentary Copy. Request a complimentary copy of The AI Advantage. Click the link in the chat to enter your shipping information. Limited supplies are available on a first come, first served basis to attendees of the live webinar. Due to applicable gift rules, Travelers may be restricted from providing a complimentary book to certain government officials/employees. 

(SPEECH) 
So Tom and Mojgan, welcome to you both. There's another special treat for all of you tuning in today. I'm going to pause and share that we're thrilled to offer every one of you who joined us today a copy of Tom's new book, The AI Advantage. So if you're interested, click on the link we just added to the chat feature there at the bottom of your screen and fill out your contact information and shipping details, and we'll ship you his new book. 

So we'll start with opening presentations today, followed by moderated discussion and all of your Q&A. So feel free to put your Q&As in the chat feature there-- I'm sorry, under the Q&A, not the chat feature. And click anonymously if you don't want me to read your name. We'll try to get to as many questions as we can at the end of the presentation. So with that, Mojgan, please take it away. 

(DESCRIPTION) 
Lefebvre in a video window in the upper right corner. Text, The Data-Driven Enterprise. Transforming Business in the Digital Age. Mojgan Lefebvre, EVP, Chief Technology & Operations Officer 

(SPEECH) 
Thank you so much, Joan, and a pleasure to have you with us, Tom as well. It's an honor to be here with so many of our customers, agents, brokers, employees, hopefully prospective employees as well to talk about the role of data and technology both in the world of insurance and at Travelers. What I wanted to do over the next few minutes was to talk at a high level about some technology trends, and then from there, hone in on a few key technology enablers that are really making a difference in the world of insurance, and then showcase just a couple of examples from our teams. 

(DESCRIPTION) 
The products identified by these ten well-known trademarks did not exist before 2007: IPhone, Twitter, iPad, Instagram, Uber, Ring Doorbell, Zoom, Amazon Echo, Fitbit, Apple Watch 

(SPEECH) 
It's incredible when you look at the list on this page to think that a little over 15 years ago, none of these services were actually available. Personally, when I think of the Apple Watch and how much I use it, not only for fitness but really in my day to day life, I can't believe that it was only introduced in 2016. And then as Joan said, of course, the last year has accelerated so many things, Zoom being among them, both in our professional and personal lives. Incredibly enough, the number of people I actually connected with and saw their faces on video whom I hadn't talked to for many years, many from in far flung parts of the world, was incredible and I'm sure I'm not alone in that experience. 

Now, what's behind this? 

(DESCRIPTION) 
A graphic depicts the growth of computing power from 1900 to 2050. Text, Exponential Growth Of Computing Power. Calculations per second for a $1,000 computer. Analytic Engine, 0.0001 calculations per second in 1900. IBM PC, 100,000 calculations per second in 1981. 10 to the 16th power calculations per second in 2000. 10 to the 26th power calculations per second in 2023. A graphic depicts All Human Brains 

(SPEECH) 
I wanted to look at a couple of things. First, it's really the power of compute. And let's think of that as the number of calculations that a computer does a $1,000 computer, personal computer, per second. If you think of 1981, the IBM personal computer that was around $1,000 or $2,000, depending on which version you bought, could only do 100,000 calculations per second. Today, a similar computer can do up to 10 to the 16th and don't ask me about the name of that number is, but it can do 10 to the 16th calculations per second. That's the number of calculations that the human brain does when it's doing pattern recognition. 

And there are sources that say that in the next 30 years or so, that similar computer can actually do calculations that are up to the capability of the entire human race. So just think about that. $1,000 will buy you the power of the human race. 

(DESCRIPTION) 
Two boys with smart phones. Text, Almost anyone, anywhere can access information and build new products and services. 

(SPEECH) 
And what's amazing is that this is not really limited to a part of the world. We'll talk about cloud computing and how it's making everything available to everyone. So think of a kid in some village and some part of the world where you don't think of technology. And with the access that they have to a smartphone, they actually have access to more computing power than a professor would have had 20 years ago at MIT. I'm sure Tom will correct me if I'm wrong about that. 

(DESCRIPTION) 
Convergence of Technologies. Physical; 3-D printing, Nanotechnology, Graphene, Conductive polymers, Magnetic levitation, Reusable rockets, Quantum dots, Metal foam, Programmable matter. Digital; AI and Machine Learning, Robotics/Process Automation, Internet of Things, Augmented reality, Autonomous vehicles, Blockchain, Conversational interfaces, Power from the air, 5G, LiFi. Biological; Genetic engineering, gene editing, Biofabrication, Epigenome Mapping, DNA App Store, Plantibody, Virotherapy, mRNA. Gartner studies 2000+ technologies 

(SPEECH) 
And it's really the coming together of these hundreds or thousands of technologies that we're talking about from the physical, you know, 3D printing, nanotechnology, a whole set of them, to the biological, genetic engineering, to mRNA, what's behind the vaccines, to the digital, artificial intelligence, machine learning, and so much, and really, the blurring of the boundaries. So it's using computational algorithms and cloud-based catalogs that a company like Moderna or Pfizer-- and Tom is going to talk to us about some of these specific companies more-- that they identified mRNA sequences that really teach our cells in the body to develop a protein that creates the immune response that's in the vaccine. So just think about the power of all of these coming together. 

(DESCRIPTION) 
Key Enablers. Cloud, API, Sensors, AI, Agile 

(SPEECH) 
Now, let's focus a little bit more on some of the ones that have started to make and are making a big difference in the world of insurance. What I wanted to do was to bring us onto the same page, provide a high level definition of each these, some of the benefits, and then bring it all together with Agile, which is really more a way of working and culture and thinking, and I'll talk more about that, and then a couple of use cases at Travelers. 

(DESCRIPTION) 
Cloud - A Vast Network Of Remote Servers Around The Globe. On Demand Compute 

(SPEECH) 
So let's start with cloud. Really, think of cloud as a vast network of servers that were created and have been created in the form of mega data centers by companies like Amazon, Google, Microsoft. And they're really making on-demand compute available to anyone. 

(DESCRIPTION) 
Lower Cost of Ownership 

(SPEECH) 
So whether you're a consumer or you're a company or you're a startup, you no longer need to have access to hundreds of thousands or millions of dollars to go buy all the hardware that it takes to start creating and selling your services. 

(DESCRIPTION) 
Flexibility & Scalability. Rapid Development & Innovation. 

(SPEECH) 
Really, having availability, scalability to spin up or spin down as needed, and really do innovation the way that you need to. And then more importantly, I'd say, make innovative services that these companies like Amazon and others are creating in the cloud available to the technologists in other companies or to the business people in the other companies. 

(DESCRIPTION) 
Business Resiliency 

(SPEECH) 
And then the inherent disaster recovery and business resiliency that comes with that, with the availability of all of this compute across different parts of the world. So that if one goes down, it's available elsewhere. And we'll talk a little bit more about the power of cloud and our use of Travelers which we've leveraged quite a bit, both for internal capabilities, whether in the form of software as a service, whether it's salesforce, workday, or other things, or capabilities that we have built and continue to build for our customers, partners, and with many of our partners. 

(DESCRIPTION) 
APIs In Today's Economy 

(SPEECH) 
Next, I want to talk about APIs or Application Program Interfaces. I'm sure all of you read a lot and talk a lot about it. 

(DESCRIPTION) 
The way software talks to other software 

(SPEECH) 
You what might have thought that was just limited to the world of computer science given that it's really the way software talks to another software. 

(DESCRIPTION) 
Repeatedly leverage data and functionality. Mechanism to exchange value in today’s economy 

(SPEECH) 
But it has become the way, the best practice for how you build software so that you're exposing functionality of data within the software in a way that it can be reused by others and really creating a mechanism to exchange value in today's economy. The platform companies like Amazon, Google, and others, that's what they've done. 

So if you think of the ride-sharing app Uber or another one, the navigation capabilities in that app are really leveraging the Google Map APIs that were created by other engineers, or the Payments APIs were created by another software company, and that concept has started to come into the world of insurance as well. We certainly at Travelers have leveraged APIs for integration with a lot of the agency management systems that many of our agents and brokers use, and are now driving towards creating integration with our agents and brokers, and I'll talk a little bit more about that. 

(DESCRIPTION) 
The Proliferation Of Sensors 

(SPEECH) 
We can't talk about technology and its impact in the world of insurance without talking about sensors and connected devices on physical robotics. 

(DESCRIPTION) 
Photos of Industry, Car, Wearables and Home. Text, The number of connected devices has grown 1,100% to 20 billion in the last 10 years. By 2025 there will be over One Trillion connected devices. 

(SPEECH) 
Sensors have been around in the world of industrials for a long time, but it's really their abundance and their introduction into the world of cars, homes, wearables, and so on that we see the explosion in the number of connected devices around us. And there are predictions from McKinsey and others that say by 2025, there will be $1 trillion connected devices in the world. Who knows if that number is accurate? But the scale certainly is. 

(DESCRIPTION) 
A bar chart depicts the Explosion Of Data from 2010 to 2024. Text, 90 percent of all data has been generated in the last 2 years. Source: Delta Net 

(SPEECH) 
And what does that lead to? It's the explosion of data that we see around us. And again, talking about numbers with names that we may not say a lot but measured in zettabytes which are billions of terabytes, what you see is that 90% of all of this data was generated in the last two years. So there comes again that exponential curve that's just going to continue. 

(DESCRIPTION) 
A graphic depicts circuits in a human head. Text, Artificial intelligence. Techniques used to teach computers to learn, reason, perceive, infer, communicate and make decisions similar to or better than humans. 

(SPEECH) 
And I say it's really the compute, the power of compute, and the explosion of data that are really the fuel to the next category of artificial intelligence. And I'll certainly leave this field for Tom to talk to us about, given his expertise and incredible experience. I'll just say think of it as the techniques that help machines do and do many of the capabilities that a human brain does, but probably do them at a much larger scale than a human brain can. Looking forward to hearing more about that, Tom. 

(DESCRIPTION) 
The Key Elements Of Agile 

(SPEECH) 
And then last, I'll talk about Agile. 

(DESCRIPTION) 
An arrow connects graphics and text, A light bulb, Idea. a clock, Fast, a shopping cart, In Market. 

(SPEECH) 
And I just touch on this because you probably hear a lot about it in terms of how is it that these digital native companies are introducing capabilities into our hands literally in minutes and seconds. But these are things that have gone from taking years to bring product to market to today being so much faster. 

(DESCRIPTION) 
Adjust, Learn. Optimized teams and skills, customer centricity, Continuous, iterative delivery. 

(SPEECH) 
And then taking the input from the customer in a very customer-centric manner and bringing it back into the product and continuing to iterate and make the products better. 

(DESCRIPTION) 
Industry Leaders In Agile Ways Of Working. Spotify, Capital One, ING, Netflix, Uber, Fidelity 

(SPEECH) 
So whether you're thinking of companies like Spotify and Uber and others where their capabilities seem to change without us noticing, there are also other larger companies which for the past several years, like Capital One, among many others, and ourselves, that are driving towards transforming our way of working and bringing capabilities to market much faster, much more effectively, and much more focused on the customer. 

(DESCRIPTION) 
Wildfire Loss Detector. A Deep Learning solution to assess property damage caused by wildfires. Text in blue ribbon graphics, 2019 Gartner, and 2020 CIO 100 

(SPEECH) 
And now, for a couple of examples that I just wanted to highlight very quickly, there's a lot behind all of these. But first, I'll start with our Wildfire Loss Detector. It's our award-winning solution that was created by our claim data and analytics and technology teams, which is really a deep learning model, again, a field of AI, that's focused on assessing property damage after a wildfire, and the ability to do this without any sort of in-person inspection. Really, leveraging a lot of the geospatial and AI capabilities that we've built over the years with a focus on convolutional neural networks. Again, an area of deep learning. 

So post-wildfire, it analyzes a set of aerial images and identifies the customer properties with near-perfect accuracy. So imagine the ability to drive towards delivering the Traveler's promise much faster for our customers, and at the same time, doing it far more safely for our employees, which is critical as well. 

(DESCRIPTION) 
The interface to myquote.travelers.com on a smartphone. Text, New mobile app. Get an instant quote. See a simplified version of a policy. Make policy changes, pay a bill, file or track a claim with a swipe or a tap. Request roadside assistance and track the driver when help is on the way. Directly contact an agent. Download auto ID cards 

(SPEECH) 
And the next, very quickly, our fully relaunched re-engineered mobile app that we launched just a few months ago, leveraging full native cloud capabilities among many other technologies. Again, working in agile methods and bringing together an experience that's highly rated, gives the user the ability to get an instant quote, a simplified version of their policy, or directly contact an agent or continuing to introduce new capabilities based on the fact that this is based on modern technology, modern architecture. 

(DESCRIPTION) 
BI eSubmission - Proposal. Return the proposal document for submissions that were sent digitally to Travelers Business Insurance. Learn More. BI eSubmission - Status Text, Let's build together. Start driving better digital experiences. API Developer Portal. The public face of Travelers APIs, used by third-party business partners to enable speed and agility. 

(SPEECH) 
And last but certainly not least, just a very quick screenshot of our API developer portal. Really, introducing the capability of making the APIs that we're creating available to the engineering and technology teams within our agents, brokers, and partners, and other partners as well. And much more to come here. 

(DESCRIPTION) 
Scaling & Accelerating Our Perform & Transform. Focused On Business Outcomes. 

(SPEECH) 
And certainly, I wanted to end with the topic that regardless of what technology you're talking about, having the strategy, a business strategy that's really focused on outcomes, knowing what problems you're going after certainly is important.  

(DESCRIPTION) 
Graphics and text, Talent As Advantage Multiplier, Integrated Agile Teams, Data & Technology 

(SPEECH) 
So regardless of technology, ensuring that you're providing the right capabilities to your talents, bringing your talents together in a way where you're blurring the lines between different roles, business technology, all in teams that are focused on creating these outcomes, and again, as we said, providing the right technologies, infrastructure, and data to the teams, and continuously improving on this. And 

(DESCRIPTION) 
Our Vision. Text, To be the undeniable choice for the customer and an indispensable partner for our agents and brokers 

(SPEECH) 
doing this all for us in a way to be the undeniable choice for the customer and an indispensable partner for our agents and brokers. 

And now, with that, it's my pleasure to turn it over to Tom Davenport, one of the great voices in AI. Tom is going to talk to us a whole lot more about AI and what they can do. Thank you. 

(DESCRIPTION) 
The State of Artificial Intelligence in The Enterprise. Thomas H. Davenport, Babson College/MIT/Oxford/Deloitte 

Davenport in the video window in the upper right corner. 

(SPEECH) 
Thanks Mojgan. That's very interesting presentation. And I think all of those technologies are having a big impact on the world these days. But if you believe survey data, it would suggest that people who are knowledgeable about technology think that artificial intelligence is the number one transformative technology that their organizations are adopting. Not a new thing for Travelers. I can remember more than 10 years ago, I was doing a research project on automated decision-making and talked to Travelers about that. It was using automated underwriting decisions. And then one of the companies that I advise, an automated machine learning company basically arose out of Travelers. So Travelers has been data-driven for a long time, I guess. 

So I'm going to talk about the state of AI in large enterprises. That's generally where I hang out from a research and consulting standpoint, and so on. So if we can move ahead, I'll get right into it. 

(DESCRIPTION) 
What Is Artificial Intelligence? No Perfect Definition. Generally defined as a group of technologies that can do tasks previously only performed by human brains. Reasoning, perception, decision-making, other tasks requiring knowledge. But some tasks (massive machine learning models) could never be performed by human brains. Debate about whether some technologies should be included or not. Robotic process automation, statistical machine learning. “When AI works, it's no longer defined as AI.” 

(SPEECH) 
Some of you may want a definition of AI. There's not a great one out there, I don't think, but the consensus view is that AI is a group of technologies, that's important, they can do tasks previously only human brains could perform. Things like reasoning, and perception, and making decisions, and any other kinds of tasks requiring knowledge. But as I say, it's not perfect because some tasks could never be performed by human brains that AI does today, a massive machine learning model, making very data-driven decisions in milliseconds. At least my brain is not going to be able to do that. 

And there's some debate about whether some technologies should be included or not. Robotic process automation, it's very good at automating back office structured tasks, but some people say it's not really smart enough to be AI. I generally include it because it's often being combined these days with machine learning. But even some people say, well, machine learning it's not really AI. It's been around for a long time. It's really just statistical analysis, which is correct, but it's kind of the hottest field in AI so I certainly wouldn't want to exclude it. But there's this old complaint by AI researchers. When AI works it's no longer viewed as AI, things like extracting data from documents or as Mojgan mentioned, some of these smart speakers and so on, we don't even really view them as AI anymore because we're so used to what they can do. 

(DESCRIPTION) 
What is at the Core of AI? Applications; Image recognition, Speech recognition, Intelligent agents/bots, Intelligent robots/co-robots, Predictions/Classification, Robotic process automation. AI methods; Neural networks, Deep learning, Machine learning, Rule engines, Natural language processing. Underlying tools; Statistics, Logic, Semantics 

(SPEECH) 
OK, so moving along. I always find it's very useful when talking to people about AI to say, well, there are a lot of applications, things like image recognition and speech recognition and various types of intelligent agents and bots. And if you're talking about physical robots, collaborative robots, those are an application of AI in itself. Certainly, prediction and classification of what machine learning do really well, I mentioned RPA, but there are only a few underlying methods that are all related as Mojgan mentioned, deep learning very popular today. Sounds like it's great at recognizing fire damage and I'm talking to you from California. Probably come in handy this year, sadly. 

Neural networks are somewhat more simple version of deep learning models. And those are all machine learning. All statistics-based, as I suggested earlier. The insurance industry has probably done more than anybody else with rule engines particularly in areas like underwriting. But now, I think there's a general feeling even though those were very useful, there's a general feeling that you can have more precise and data-based decisions with algorithmic or statistical decisions. 

And the natural language processing is kind of a hybrid. It can be based on machine learning or it can be based on semantics. So, but I just think it's very comforting to think, well, there are only three underlying tools that drive all of this. Statistics, which are the basis of machine learning. No matter how complex the deep learning model, it's still based on betting lines and curves to data points, although it could get very complex versions of that. Logic and things like rules and semantics or the way sentences and words get strung together. 

(DESCRIPTION) 
The State of AI in Large Companies in Mid-2021. About 40% are “AI-aware” globally and actively employed multiple technologies. A few are “AI first,” most less aggressive. Some challenges getting systems into production. Diverse objectives beyond automation, but 63% will automate “as many jobs as possible.” Less ambitious “low-hanging fruit” projects are often more successful than “moon shots.” The more experienced, the more bullish on AI. Pandemic economy wreaked data havoc on models. Increased focus on automation technologies. 

(SPEECH) 
So if you look at surveys around the world, about 40% of large companies are actually doing something with AI and employing multiple technologies. I'll say what they are in a second. A few of them are AI first, or you could also call it AI-fueled. But most are less aggressive, and I'll talk. What does it mean to be AI first in a moment. The big challenge with AI is not experimenting with it but getting it into production, and I'll discuss that further, too. 

I'll talk about the objectives that companies have for doing AI. If you're a human being, you may worry about one of automation. And indeed, in a Deloitte survey we did a couple of years ago, 63% of US managers said you know, we'd like to automate as many jobs as possible to cut costs. But in general, that's not happening. Travelers has assured me it's not their goal with their AI. It's much more augmentation than automation. Machines and people working alongside each other. 

In general, I believe that these less ambitious AI projects are more likely to be successful than the really dramatic moonshots, probably true of any technology. But I think it's particularly true of AI, and AI does small things really well. It doesn't do big, big tasks nearly as well. Good news is that the more experience the company is with AI, the more bullish its executives are on it. We had some issues in the pandemic, particularly with machine learning models because they're based on past data, and past data wasn't reflecting current behavior particularly of consumers. But that started to ease a bit. And also with the pandemic, I think there's been a big upsurge in automation technologies which compared to us humans don't get sick and don't have to stay home and help their kids do home-schooling and so on. 

(DESCRIPTION) 
What Technologies Are Businesses Using? Statistical machine learning, deep learning, and computer vision are the top three widely used AI/cognitive technologies. Statistical machine learning-now, 67%. Next year, 97%. Natural language processing, 58%. Next year, 94%. Deep learning neural networks, 54%. Next year, 96%. Computer vision, 56%. Next year, 94%. Robotic process automation (2018), 59%. Total (n equals 2737). Source: 2021 Deloitte “State of Enterprise AI” global survey. 

(SPEECH) 
So on the next page, I talk about the technologies that companies are actually using. And in these surveys, you get a high level of ambition, I would say. It doesn't always come true but this is the most recent Deloitte survey. And we asked people, what are you using now? And note, a lot of machine learning, a little less natural language processing, but still more than half, deep learning a form of machine learning as I said, pretty high. Computer vision, which probably also is a form of machine learning, pretty high. Robotic process automation, the people who worked on the survey this year didn't ask about it but a couple of years ago it was about the same, maybe a little more than some categories. But basically, everybody says they're going to be using it next year. So very ambitious plans for using AI in business, and about the same mid-90s percentages say that they think AI will transform their company, and a somewhat smaller group, their industry within the next three to five years. So great expectations for this technology. 

(DESCRIPTION) 
A graph depicts responses to the question, What Objectives Are Being Pursued? Making processes more efficient. Improving decision-making. Discovering new insights. Making employees more productive. Enhancing existing products and services. Enhancing relationships with clients / customers. Creating new products and services. Enabling new business models. Lowering costs. Reducing headcount. On the y-axis, Percentage who achieved outcome to a high degree. On the x-axis, Percentage who reported as top two potential benefits pursued. Source: 2021 Deloitte “State of Enterprise AI” global survey  

(SPEECH) 
So moving along to the next page. This is the one, the topic on what objectives do people want to accomplish with AI. And the upper right is where companies say we really want to accomplish this, this is one of their two benefits pursued from AI, and also what they've already accomplished. So making processes more efficient is the winner on both, and this is something we've been doing for a while with analytics and other automation technologies. But it continues whether you're talking about marketing or supply chain or even human resources. Improving decision-making also a focus for analytics. 

A little less, this idea of enhancing existing products and services. So as Mojgan was saying, you know, if you're selling property and casualty insurance and you can very quickly figure out whether someone has suffered damage to their home or their car or whatever, that has enhanced that particular service. Note, in the lower left, this objective of reducing headcount is the smallest, both sought and achieved, but about 31% said they've done some of it already. That was a little higher than I would have expected based on my conversations with companies. 

(DESCRIPTION) 
How Much AI in Companies? Vanguard, One major “robo advisor” application. Pfizer, About 150 AI projects, many in marketing and sales - not drug development. CapitalOne, About 1000 projects, mostly in machine learning for credit, risk, marketing. Alphabet, “AI First” – Several thousand projects and search, ads, autonomous vehicles, etc. 

(SPEECH) 
So moving along, the issue of how much AI to use comes up when you look across organizations. So Vanguard which is I think the largest consumer mutual funds company in the world has one big application. It works really well. I'm a customer of it. I'm very happy with it. They called me one day and said, you know, you've been complaining about not being able to get advice directly, having to talk to your advisor, and sometimes complain about what we charge you. What we're cutting--what we charge you in half and--we'd still like to talk to you occasionally but we're going to give you all that information directly in what's called personal advisor services. And they have, I think I looked a couple of days ago, over $175 billion in assets under management with this tool. And for me, it's worked quite well. It's not rocket science it's all just choosing among mutual funds and ETFs, but it's nice to have a lower cost. 

Pfizer is one of the more aggressive drug companies in this space. I don't believe, I think most of the development of that drug was done by--that vaccine for COVID was done by BioNTech and Germany, but Pfizer is known as a great marketing company and most of their projects are in marketing and sales now. They have experimented with it in drug development but right now, it's mostly startup firms that are doing aggressive work in that area. 

CapitalOne, when I was doing work on how companies use analytics, was one of my poster children and it's a poster child for AI as well with more than 1,000 projects, most of them machine learning oriented, credit risk, marketing. They have the obligatory chat bot as well and getting a lot of value from it, almost totally in the Cloud as Mojgan was referring to. 

And then if you had to pick, I guess, the grandfather or grandparent of AI first organizations, you'd have to choose Alphabet/Google, which I think in 2016 said they had 2,700 different machine learning projects. Alphabet, of course, the holding company. They have things like autonomous vehicles and health care research, and all of that uses AI quite extensively, too. 

(DESCRIPTION) 
How to be “AI First.” Use AI across the company, with multiple use cases and technologies. Many systems in production deployment. Use AI to reimagine and reengineer work processes. A high percentage of employees fluent in AI and how it can be applied. Voluminous, high-quality, and unique data. Long term commitment, large Investments. A framework for ethical, trustworthy AI in place. Well-defined governance structure and substantial talent. 

(SPEECH) 
Maybe you're wondering, and the next page addresses what it means to be AI first, and I'm doing a research project on this now. If you're in that category, I'd love to hear from you. But it's companies that use AI quite broadly with a lot of different use cases and technologies. As I mentioned, you know, in the thousands in some cases, many of them in production deployment as I say. Not just a bunch of pilots using AI to reimagine the way work gets done. I think it's very critical if we're going to get productivity and value out of the technology. 

Companies are increasingly educating their employees about how AI works and how it can be applied in their business, and that's a key component. If you want to do a lot of AI, you need a lot of great data. And companies in AI first organizations really have that. They made a long-term commitment and are spending a lot of money in many cases, approaching billions and a few very large companies, having a framework for ethical and trustworthy AI. And I believe I heard that Travelers has one of those as well, soon to be rolled out. And a well-defined governance structure and lots of talent. We still need human talent for making AI work. If you have all those things, I think you can justifiably be called AI first or AI-fueled. 

(DESCRIPTION) 
How Ambitious Are Projects? The University of Texas MD Anderson Cancer Center. A red dot next to Treating Cancer. Green dots next to “Care concierge,” Predicted no-pays, IT operations. DBS. A red dot next to “Robo-advisor.” Green dots next to India chatbot, ATM cash, Sales attrition. Amazon. A green dot next to Go stores; a yellow dot next to Drone delivery. Green dots next to Fraud detection, Product recs, Merchandising 

(SPEECH) 
On the next issue, of course, is just how ambitious should an individual project be. And I've just listed three organizations here that have both moonshot attempts in the center column, and low hanging fruit attempts in the right hand column. And you might guess from the colors of dots there that most of the moonshots have not been effective. MD Anderson was not successful in treating cancer with its big IBM Watson project. DBS, the largest bank in Southeast Asia based in Singapore, tried a robo advisor to recommend stocks and bonds. It didn't really work for them. 

Amazon has been, I think, successful with the Amazon Go Stores where you don't have to pay, or you think you're not having to pay, you get billed shortly after you walk out of the store on your smartphone. Drone delivery, they said they'd have it by 2018. I haven't had any land on my front porch yet. Who knows when that will actually happen. But all of these companies have had successes with low hanging fruit. And even Amazon, you know, arguably one of the more technically expert organizations on this planet, Jeff Bezos said in a letter to shareholders a couple of years ago. He said, the great majority of our work with machine learning is quietly but meaningfully improving core operations. And that to me is another word for low hanging fruit. 

(DESCRIPTION) 
A photo from Mission Control in Houston, Apollo 13. Text, Evidence of Deployment And “Return On AI” Problems. Seven out of ten companies in an MIT/ SMR/BCG survey report minimal or no impact from AI thus far. Only 15% of respondents to a NewVantage Partners survey report any production deployment of AI. VentureBeat AI: 87% of data science projects are never deployed. In a McKinsey global survey, only 21% of respondents have embedded AI into multiple business units or functions. The Gartner 2021 CIO Agenda survey report comments: Quote: "the reality is that most organizations struggle to scale the AI pilots into enterprise-wide production, which limits the ability to realize AI’s potential business value.” 

(SPEECH) 
So moving along to the issue of deployment that I mentioned was a problem. I won't go through all these surveys but they certainly suggest that a number of organizations have experimented with AI, but a relatively small percentage have said we put AI systems into production deployment. And you say, well, why is that the case? It's hard, it's expensive, you have to integrate it with your existing systems often using APIs as Mojgan suggested. You have to retrain your people. You have to probably change your business process to some degree. So it's far easier to experiment and do a pilot or proof of concept on the other hand, though, those don't provide any economic value typically. They provide learning but no dollars. So companies need to move more in the direction of pipeline toward production. 

(DESCRIPTION) 
In a graphic, four figures push an object. The figure in front pushes a ball, while the three others push boxes. Text, Creating an AI Advantage. Think big- strategize about how AI can transform your strategy, business model, or business processes. Start small - start with pilot projects and less ambitious goals, and scale them. Upskill - emphasize augmentation, offer skills training, and give employees job options and the time to transition to them. Get your data in shape - internal, external, big and small. Reprioritize your projects, but don't ease up. Put an ethical framework in place. Create AI leadership and governance. 

(SPEECH) 
On the next page, I talk about this whole set of issues and really summing things up--how do you create an AI advantage? I think it's important to think big about AI, and think about how it can transform your strategy, and your business model, and your key business processes. But you probably want to start small. As I said, pilot projects are fine but make sure you have plans to scale them and to deploy them. If you do low hanging fruit projects that aren't very ambitious, you can combine a number of them in a way that together they might have a transformational impact. 

I think upskilling, it's not too early to start. Although many organizations haven't started, some have. They're saying, we need our people to be able to work effectively with AI, they're offering skills training, they're giving employees some options for where they might migrate their roles over time and you know some time to make that transition. As I said, data is really critical. Internal transactional data, more and more companies are using external data for AI and other purposes. Big data, small data, it all matters. That's what makes AI work. 

In this economy, I think it's probably a good idea to reprioritize your projects, but the best companies are not easing up in this space. So you don't want to take too much of a breather. Most of the companies that have put these ethical frameworks in place have been vendors, but I think it's really time for user companies of AI to create some sort of ethical framework or an issue--around issues of algorithmic bias and transparency, and so on, particularly if you are an aggressive user. And this doesn't happen without AI leadership and governance so you need to put someone in charge. And with that, I hope that a number of you have questions that Mojgan and I can address. 

(DESCRIPTION) 
Davenport, Lefebvre and Woodard share the screen. 

(SPEECH) 
All right. Well, Tom, Mojgan, that was really terrific. That was a great overview of how you're thinking about AI and data analytics in the past year and going forward. So let's get right into digging into some of the stuff. First Mojgan, and again by the way, please put your questions in the Q&A function. And we've had a number of questions about will this be a replay. There will be replays, and we'll shoot out the link after the session. Probably tomorrow. And then you can request Tom's book by going into the chat feature. And the chat feature there's the LinkedIn profile for me and also the book request. So those are the two questions we got most coming in the feature. 

So Mojgan, let's go back. You talked about some of the ways Travelers is already leveraging data, technology, AI in our business right now. What are the future applications you're really most excited about for the insurance industry as a whole? 

Now, as I think of that, Joan, I think of it in like there are three categories. And the more of these categories that come together in a use case, the more that excites me. So something that brings together the impact that it has on the experience, so improving the customer experience and at the same time has the ability to either impact our expense ratio and reduce our costs, or alternatively, allow us to do better risk segmentation and continue to improve our risk expertise and improve our loss ratios. And so as I think, for example, the wildfire example solution that I gave, that's really doing both. It's improving the experience and it's also allowing us to do our job much more effectively and much faster and at much lower cost than if we had to do inspections. In some cases, it would not be possible. 

So I agree with Tom, by the way, that AI absolutely is one of the most exciting, but I would say it's the coming together of the availability of third-party data sources, the availability of live data feeds that's now becoming more and more possible with Cloud technologies and others. And then the advancements in the AI itself that are making it possible to drive those hyper-personalized customer experiences, where, you know, the customer any time that they contact us, they feel like we know them. We know all the interactions that we've had with them before. And then at the same time, depending on the actions that they've taken, we take those elements into account in terms of how we think of them as the risk and the pricing. So it's really bringing it all together, Joan. 

OK, so let's follow up on that. So we have, you know, business units at Travelers, then we have technology, and we have operations, and a lot of other agents and brokers on the phone kind of have these different, if you, will silos in the past. But how does the partnership work between the business units and their prioritization and the technology solutions? Because and is that relationship evolving? 

So it is, and you know, that's really what I was talking about when I talked about Agile. I mean, again, the way you tend to want to think about this is not in silos, but really a holistic think of it as an initiative or a way of introducing a product or a service to market. And think of everything that it takes from who are the technologists and who are the business subject matter experts that need to be on the team. And the teams should be small, they should have--they should start with your top level strategy and how you define the outcomes and the key results you're shooting for, how you define the metrics that are going to help measure. And then each of the teams kind of drive towards that and have the ability to do the work, prioritize, and reprioritize, learn from it, adjust. 

And then, again, of course, we do strategic planning every few months and ensuring that all of these things are connected to each other. We've set up the big bodies of work as what we call value streams that are very aligned to our business objectives, and think about it that way. And again, with everything that we do, there are infrastructure costs, there are data costs, there are business people costs, and we look at these holistically and try to think of them, you know, as having a long-term view but then thinking of them in small increments and adjusting in small increments. 

Can I address that question, too, Joan? 

Oh, sure. 

So I think last year, I was on a panel with Mojgan for the MIT-CIO symposium. And this year for that, I've been interviewing a number of CIOs. And one of the striking things to me, I've seen this in some of my other research as well, is that the historical gap between business people and technology people is definitely closing. I mean, several CIOs said sometimes you go in some of these Agile meetings, I don't know if you had this experience, Mojgan. It's hard to tell who's an IT person and who's a, quote, business person because the IT people are getting more business-oriented, the business people are getting more tech-oriented, and they can collaborate much more effectively. I mean, we've been wrestling with this gap for as long as I've been in this field. It was for a long time. But I do think there's a lot of hope for closing it now. 

And that's a great point, Tom, and you know, I think like it's in the world of engineering, we talk about it as like full stack engineers, but it's almost like becoming full stack employees or being t-shaped where you've got depth in one area, but you absolutely need to have capabilities in all the kind of areas that are right beside you that enable you to do it. And I think if you look out five years from now, that division that might exist more in terms of defining myself as I'm an underwriter, I'm an actuary, I'm an IT person, that's going to start to go away, and that's what we need to enable with the upskilling and reskilling that we talk about. 

And that's really exciting the way you both kind of positioned that these are converging more and more now between the business and technology. OK, Tom, we're going to go to you because you really did touch on it and it's just fascinating area of study. The pandemic really has changed the consumer and business behaviors in countless ways right, over the last year or so. How can we use the past datasets to predict the future given what we just went through in the last year? And how does that decision-making count for such an unprecedented year? 

Yeah. Well, you know, that's a very important message for anybody who's not that familiar with analytics. God did not see fit to provide us with data from the future. If you think you have some, you may want to talk to your psychiatrist. So we have to rely on past data to predict the future, and if it's no longer a very good predictor, as was true in the early stages at least of the pandemic--one of the retailers that I work with said, it's amazing what eight weeks of zeros will do to your demand forecast. You really have to, A, look for other sources of data. And in many cases, the analytics and AI people I work with were just being asked, can you at least tell me what happened last week? Forget about prediction, just give us some up to the minute data. But also, they were using external data, trying to look at other measures of consumer activity, what they were looking at online. Auto company was looking at smog you know to see--that's apparently a fairly good indicator of how many cars are on the road. 

So you look for other types of data. And then you--I think this is important for a business, like Travelers, or any company that's becoming dependent on these analytical and AI assets. You've got to constantly be checking, does it still predict well? And if not, you retrain it, get some more recent data and retrain the model. I think there's a human tendency to sort of, once you have something automated, to sort of let it ride. But that's gotten us into big trouble. It was a big factor in the Great Recession of 2008-2009 in terms of mortgage models. They just didn't predict well whether you'd pay back your loans or not anymore. So we, that's I think the responsibility of every manager to constantly be checking, are these models still doing a good job? 

Great. Thanks. So on that point, everyone's coming off this difficult year. How should people think about the ROI, return on investment, in AI? 

Yeah, I think we haven't seen a lot yet. There is a fair amount of faith out there that it will be coming at some point. But as I said, if we're only doing pilot projects, one company said to me, we've got more pilots around here than there are at O'Hare Airport, we need to bring some of these things in for a landing, so to speak, and get some value out of it. So I think it's a pipeline toward production, as I suggested. It's thinking about redesigning your work processes because typically, that's going to be required to get productivity and value. Some companies are getting their CFO organization involved in kind of certifying, yes, this really pays off. And as Mojgan suggested, I think we have to be a little less-oriented to the hype of AI and more to what business value is it going to achieve for us. 

Right. Thank you for that, Tom. OK, a question coming in from the audience several different times so we're going to just hit both of you with it. We're going to try to do these more rapid fire. So can you talk about the ethical considerations here of this data-driven transformation, the bias in datasets, and how you think about it, Tom, and then Mojgan on how do we think about it at Travelers. 

Sure. Well, I think it's important to remember that humans are very biased in our decisions as well. And so if we're training our machine learning models on the decisions that we humans have made in the past, chances are pretty good they're going to come out biased. So right now, it's hard to see bias until you see outcomes. So I think you need to monitor early outcomes and see is there a bias in the predictions or the categorizations that you're making. And then see, do I need to find some new data and retrain my models to prevent that in the future? 

And then, Joan, what I did is we--first of all, I think it's ethical AI but it's also ethical computing generally, right, because you can leverage technology the wrong way. And any reason it goes to the core of values of a company and, and whatever technology you use has to be aligned to your company values. We think of it as making sure that it's still people-centric that we've got human oversight and judgment involved, that we're ensuring privacy and security of the data. That's critical, our customers trust us. We have a lot of data that we're going to use carefully in the right way. That we have diverse perspectives, so to Tom's point, ensuring that the human unconscious bias doesn't get introduced, that you've got as complete, as unbiased, and as high quality data as possible, and that it's fair, that it's responsible. 

And then of course, you've got to make sure that it's trustworthy in both externally and internally. So those are some of the things that we've had in place. And you know, I know Tom, you talk about the fact that AI hasn't been used at scale. I'd say it depends on how you define it. Right, like one could argument that the insurers with what the actuaries do in the use of statistics and things like that, like there are some things that we have in place at scale, and we've used a lot of these ethical values and frameworks in how we've done our modeling. And so it's really the extension of that. Certainly, it becomes much more complex with some of the newer technologies, no doubt about it, but it's a continuation. 

And then, you know, it's also the industry and the fact that the National Association of Insurance also is very focused on this and has just--I think it was August of last year that they said that absolutely these are things that they're looking at as well. So I think it'll depend on all of us to ensure that that's how we're leveraging the technology. 

OK, great. A couple of questions coming in on labor force and labor shortages and workforce. This from Celia Via. We're hearing a lot about onshore labor shortages as newer technologies such as AI move front and center. Has Travelers run into any challenges in acquiring the talent necessary to fulfill on AI and other key technology initiatives? And what approaches are you taking to resource planning that may have not worked previously? 

Yeah, I'm sure that Travelers is running into the same issues that many other companies. And you know, when it comes to talent like this, you're really competing with anyone in any industry right. Because again, as Tom said, there is the hype around it and then there is also the real use case. I say that generally, there's a shortage of deep engineering and software engineering capabilities. If you focus just in your locale, and so that's why I think you need to think of talent globally. Even with that, I say there's probably a shortage, especially with some of what's going on in these Eastern parts of the world where a lot of this talent is, with our friends in India and the pandemic. But we're doing many things to address this. We certainly have continuous hiring pipelines. We've been lucky enough that we've been focused on data science for a long time, so certainly, leveraging those capabilities to ensure that we're bringing some of that learning to other parts of our organization. And we just need to keep at it and we're partnering very closely with our talent acquisition organization, posting our roles nationally, and ensuring that we're focused on potential and, you know, doing all of those things. But yeah, I mean, we at any point in time, have hundreds of open positions across AI and other technologies. 

OK, great. Two questions coming in. We're going to do rapid fire. I'm going to ask this question and let you think about it and then I might ask you another question because Andrew Zacha wants to know, as we head towards the summer, what is on your reading list for AI digital transformation? So what are on your reading list? But while you're thinking about what-- of course, other than Tom's book, which you should go request in the chat feature, think about this question. So for both of you, what would you suggest is a first step to implement an AI strategy for a small insurance agency? So maybe a mom and pop out there, this seems so overwhelming. You know, your Harvard, your Oxford, your MIT, just incredible minds on AI. But imagine a small mom and pop shop out there in the insurance space, what can they take is first, second, and third steps to test an AI strategy? 

So is that me or Tom or both of us? 

Go ahead, Mojgan. 

So I mean, look, I was going to shoot and say The AI Advantage, for sure. Look, I think it's about books but it's also as much about a lot of--I mean, there's a lot of great work that's being done by organizations and readings that are out there. I'd say McKinsey has a quarterly and actually monthly review that comes out that I try to read diligently. We've got tech deep dive sessions, three hours a month that we do and in many of those, we bring some of our experts and actually sit around and they educate us. So I get a lot of value out of that beyond reading. 

And then attending a lot of the immersion sessions that our teams do as they're sharing some of the work that they're doing. I say a lot of the education comes from that. So that's the first answer. I don't know, Tom, if that's a question for you as well or you're the writer of books and that doesn't apply to you. 

Well, I do. 

What are you reading, Tom? 

I do find that I can either read a lot or write a lot, and I'm trying to finish three books. So I'm mostly writing a lot, but there is a--I really like to read fiction and there's a good fiction book about AI that I'm looking forward to reading. I've just put it onto my Kindle. Kazuo Ishiguro whose work I've liked a lot in the past, who remains of the day, and some other books I like a lot. There's his book called Klara and the Sun, about an AI-based companion that a little girl gets assigned. And I'm really looking forward to reading that. 

And then to the next question, Joan. If you're a small company or a small insurance agency, where to start? I think, again, a lot of it does start with education and really having alignment in the leadership of the organization in terms of what you're doing and what you want to do and having a strategy around it ensuring that you're aware of the data that you have. You've got experts in the area, I would say, you want to make sure you've got some experts. And then leveraging a lot of the capabilities that are out there. There are a lot of solutions, again, from the large providers to smaller providers that are out there. Again, understanding what you want to focus on and probably as a small company where you're the user of AI, I'd say more than anything, leveraging a lot of the capabilities that are out there where you can buy and not necessarily have to build. Because, again, as Tom says, investing and building AI capabilities is not a small undertaking. So those would be a few of the steps that I would suggest. 

OK. Tom? 

Yeah, if I had to pick one technology for a small business to start exploring, it might be robotic process automation, which is relatively inexpensive, relatively easy to use. You don't need a lot of data science talent to do it. An insurance agency has a lot of administrative tasks that need to be done. Sending out emails, getting information out of emails, making relatively straightforward decisions and RPA is good at that. And as you advance, you can start making some more complex decisions with machine learning in that same robotic process automation context. So some people have described RPA as the gateway drug for AI, and I think that makes it well-suited for smaller businesses. 

OK. That's very helpful and useful. Thank you both. One more question here. How is AI being used to facilitate straight-through processing and claims? And how should we expect straight-through processing will be used? I guess that's for you, Mojgan. 

Look, I find it easier to talk about these things in terms of like thinking of it several years out because I think we tend to probably overestimate how quickly some things happen and then how long it might take for other things. Our claim organization has absolutely leveraged robotic process automation quite a bit for automation of tasks from a routing and triage and things like that, again, depending on the complexity. And there's a whole lot more that I would say claim organizations can envision doing. We are also leveraging-- we have leveraged and have big capabilities in drone technology, and again, geospatial and computer vision capabilities. Those are some of the things that I see. 

I think as you think of it, again, personal type insurance, a smaller commercial, will probably, again, thinking several years out, be leveraging automation and probably be somewhere around 90% plus straight-through. And then the larger, more complex ones will regardless of how much AI they're leveraging will probably still absolutely need that human judgment and intervention. And if you think of other industries that have evolved then, if you think of the financial services industry and trading, like regardless of how much automation they've had, you know, they still have the use of those experienced traders for some of the transactions that they need. So that's how I think about it. 

OK, one last question for you, Tom, and then we're going to wrap up and thank our wonderful speakers. How do we solve AI's last mile gap? 

Well, I assume that person is addressing the issue of how do we get AI deployed in companies. And I was thinking about that as Mojgan was talking. I don't--I'm not familiar with what in detail what Travelers is doing with claims. But some of the other insurance companies I've worked with have said, well, doing all of claims is really quite difficult and complex. But we can make some early decisions pretty easily, like is the car totaled or not? And even moderately good machine learning model can help make that decision, it may not be able to say exactly what needs to be done but it can say you need to throw that car away and start from scratch. So pick off a piece of the process and do that first, I think is one way. I guess you would say it's solving the last mile problem, you know, hundreds of yards at a time. 

OK, great. Well, listen, we've run out of time. And I want to thank both of our really terrific speakers and would love to have you back on our Wednesday sessions in the future. Terrific work here. 

(DESCRIPTION) 
Book cover, The AI Advantage, How to Put the Artificial Intelligence Revolution to Work. Thomas H. Davenport. Text, Complimentary Copy. Request a complimentary copy of The AI Advantage. Click the link in the chat to enter your shipping information. Limited quantities are available on a first-come, first-served basis to attendees of the live webinar. Due to applicable gift rules, Travelers may be restricted from providing a complimentary book to certain government officials/employees. 

(SPEECH) 
I'm also going to put in the chat feature right now. If you're interested in receiving a complimentary copy of Dr. Davenport's book, The AI Advantage, please fill out this at this link in the chat feature. 

(DESCRIPTION) 
Wednesdays with Woodward, A Webinar Series. Upcoming Webinars. May 26 - The Future of Cities. Partnership for New York City's Kathryn Wylde and Bay Area Council’s Jim Wunderman. June 9 - A Bright Future Tackling a Global Pandemic. Former FDA Commissioner Mark McClellan M.D., Ph.D. June 16 - Behind the Scenes at the NYSE: Everything You Want to Know about SPACs, IPOs and Direct Listings. NYSE Head of Capital Markets Amanda Hindlian & NYSE Northeast Regional Head Joe Tama. Register at travelersinstitute.org. 

(SPEECH) 
And also, we have a number of really interesting webinars coming up next Wednesday. We're going to have one on The Future of Cities. And so we're going to talk to Kathy Wylde about how New York is going to pivot after the pandemic, what's going on in San Francisco with Jim Wunderman. And then we have a former FDA Commissioner, Mark McClellan he's going to join me on June 9th, talking about a bright future after the pandemic with all of us vaccinated out there. So Mark McClellan is coming back. We had him in January and he was just terrific. We've had a number of requests for him. 

So we really appreciate everyone's time today and especially to Dr. Davenport and Mojgan Lefebvre. That was a really terrific session. We're going to have a replay out to everyone shortly. And then lastly, again, June 16th, Behind the Scenes of New York Stock Exchange. I'm going to interview the head of capital markets there. Everything you need to know about stocks, IPOs, and we're going to take a little tour behind the NYSE and ringing the opening bell. So again, my speakers, we're very grateful for your time with us today. 

Thank you, Joan. 

Thank you all. We'll see you next Wednesday. Take care.