We are going to welcome uh Lisel Yearsley. She's our next speaker. It's a very big honor to have her. She is the CEO at Akin and has done so much work in A I. This is uh a brainiac. She's an amazing woman. She's accomplished.She got to work on lead Watson life, uh which is just really historically, something that changed the game for artificial intelligence early on. So we're really eager to hear about her talk on frontier tech and artificial intelligence. So, Lisa, thank you so much for coming.

How are you doing today tonight? Wherever you are.

I'm doing good. I'm actually in Sydney, Australia. I was meant to um go to the US in March for a couple of months because I have a home there. And um yeah, the borders shut down before I could get there. It turns out this is not a bad place to be stuck for a few months.

Yes. Yes, it isn't. And um you know, we're just hoping everyone stays safe everywhere. There's a definitely, maybe there's a way we can use A I to help some of these global challenges, uh which is what we all hope, uh, we could hear you, we could see you beautifully. Uh, and we want to make sure you can share your slides if you have them. Uh,

so if I thought given it's a virtual space, I'll just talk if excellent. That's the way to go. You have to of it. I love it. I've been a

lot of back so it's full screen to you. And

I hi everyone. I thought a lot about what I was gonna talk about today. You know, usually if I go to big conferences where there's a whole lot of middle managers, I'll talk about how frontier technology is gonna change society. But what I realized is that I'm probably speaking to people who are creators themselves, people who are out at the forefront of what they're doing or want to be. So, what I'd rather talk about is the process I've gone through of building and distributing frontier technology.

So just to give you a little bit more, I've run three tech companies and they've all um been very successful. My most recent company was an A I company. It was a platform A I, I started way before the current A I boom. So I started in about 2007, um 2005. Sorry, we won a federal government research grant to build cutting edge A I. And we were the first company in the world to put out human quality, virtual humans. So we were doing a lot of front line work with avatars that could um work as tier one bankers for customer support. We did high level diabetes coaches. We we had a platform that had about six fortune 100 companies using it, including big companies like State Farm and HP. And we had about 26,000 external developers using the um by the time IBM acquired it, so IBM acquired it and it became part of the backbone of, of Watson. And as um Nicole mentioned, I worked at IBM for about two years after that. But before that, I built another uh frontier technology company in about 2001, I decided that search engines were gonna become important.

And at the time, um pretty much everyone I spoke to said, but we've already got Ulta Vista, we don't need more search technology and Yahoo was the dominant player in the market. Um But I dove straight into search, um got a federal government research grant bootstrapped away right through the tech crash double value every year of the company and that company listed publicly um was a big player in the Asia Pacific market. Um And right now I'm running a Kim which is a deep, deep tech A I company and I'll talk to you a little bit later about what we're building there. So basically, what I've done is three times in a row. I've picked broad, broad shifts um in society and technology. I've built technology. That's about five years ahead of its time. Then I've scaled it successfully and had a great exit. So I thought that's probably quite interesting. Um And II, I took some time to unpack my process, which I haven't done before. I just do it. So there's a couple of stages and this applies, whether you work inside an organization trying to bring in frontier technology or whether you're a founder of a start up, whether you're a coder or someone who wants to get more, more significance in an organization you work in.

So the first step in anything you build is actually about yourself. I go through a very, very deep and immersive process of creative visualization where I really think about what do I want in my life. Um You know, a couple of years from now, I'm not just talking, oh I want a big shiny house and this much money, but what kind of work do I want to be doing when I walk in a room? Um You know, how do I want people to be looking at me? How much influence do I wanna have? How do I wanna spend my time? And I build a very rich complex visual map of what I want my life to look like to the, the level of, you know, pictures of the house I'm gonna own, you know, pictures of the team I'm gonna run. Um And this is quite personal. OK. And then you have to invest that with emotion, you have to start to believe it. And that's the most important step. There's no point in visualizing your future. If you don't believe it, when you start doing this, that, that step of investing in it, it's, it's easier than it sounds like we have a lot of fear and um um imposter syndrome and stuff to overcome, to truly believe that.

But every great company you see today or every great division that you see today was started by someone who believed something. And then enough people come along in the journey that it becomes real. It's like a standing wave form of energy that initially exists just in your mind.

So giving time and space to that and committing to that every day um is probably the most important thing you're ever gonna do in building frontier technology because the biggest asset we have is this mind and this energy that we give to the world, the sense that we have some important work to do in the world.

So I'll move through because we've got a whole frontier tech to build in the next 10 minutes. The next thing I do is I I really step back and look at broad trends. Now. Again, this is very important that you commit time to this process. I look at where is society going? Where is technology going and where are they gonna converge 2 to 5 years from now? And then you build to that point. Now again, it's not easy to give this time. I know what it's like to run a start up. I know what it's like to have not enough people to do all the work that needs to be done. I know what it's like to have not enough money to pay salaries for my team at the end of the month. But if you don't commit time and, and emotional and mental energy to the step, you can chase the wrong goal for five years and waste a decade of your life. So it's very important. Now, I did this with every company that I started, I basically committed in my mind that we were not gonna commit to a, a user application inside of the first six months. It's hard. Usually it's sort of three months we're already pushing towards something. But you spend that time working, you read, you meet with any venture capitalist who will talk to you, you meet with people 10 steps up in your organization, you reach out to them and say, what do you think? Where's the world going?

Anyone you respect or admire, listen, listen, read, read, think, think. Um So this really helped me, this is how I landed on search engines in 2001. Um I saw this massive proliferation of data on the web. Um I saw this massive increase of people getting into, you know, um logging into the internet and, um, you know, getting their own personal computers. So it was, you know, finding information is a fundamental human need and it wasn't being done well. Um, so I thought, well, we need search technology. Um, and it was a desert for a couple of years, you know, it was really rough out there trying to sell innovative search tech before we even had AAA revenue model like pay per click. Um, but I was absolutely convinced that that's where we were going in 2006. Um After I exited my search company, I was looking at 22 major trends. Um Basically, you had the societal shift where people were getting more and more publishing themselves online. Um More and more profoundly lonely, you know, this hyper isolation, we feel even though we're so connected and yet we're demanding more and more personal attention out of companies we work with and out of things we buy. So I was a bit torn. I, I thought social media was gonna be important. We didn't even call it social media then it was called social network services. It was myspace was the, in fact, it was friends that was the king of the universe then.

Um And I also thought artificial intelligence was gonna become important. So in my motion time, I wrote a couple of patents around social media and around A I and then the A I started picking up and we won a big ride. So I had to drop the social media side and, and focus on a I. Um so, you know, this step is more important than you can realize and it's about honoring yourself. Um If you think of your mind and your energy as your biggest asset, not what comes out of your hands with code, um You have to value that. You have to honor that because no one's gonna give you that time. But yourself, even if you have to steal it and pretend you're doing something else, um Even if it's just one day a week, you need to get that time. Um I better move on. So once you've got this big idea of where the world's going, where society is heading, what sort of technology started to get on what will converge. So I, to me, it was obvious it would have a rise in social media, they would have a rise in artificial intelligence because everyone wants this individual experience and this massive data and this huge volume of people as our population grows.

So then the next step is to find your mountain. I can't emphasize this enough too often what we do um When we're leading a creative projects or frontier technologies, we go, I'm gonna build an app that does XYZ for ABC population because that's what every investor and every project manager and whatever we do say, what are you building?

Who's the addressable market? What's your KPIS? That's bullshit. Excuse my language. I've been in Australia for a couple of months now. Um You know, essentially you can do that. But if you bet wrong your whole company or your whole project will fall over. So when I say pick your mountain, I'm not saying you don't commit to something, but I'm saying commit to something big. Think is what I'm aiming towards able to reach a billion lives. In theory, if it's not, you're playing too small. OK? And what that allows you to do is iterate towards that goal more effectively. So for example, when I started my A I company, I wanted to democratize A I, I wanted, I knew A I was gonna be influential, be making more decisions for us. And I wanted it to be tools that anyone could use to build their own A I. So that was the broad mountain mass uptake of A I. Um And along the journey, the market changed and it didn't matter. We, we could shift our products and we weren't wedded to one product. So we started out in 2007, launching social media A I bots that you could put, this was like 10 years before it became mainstream bots that you could throw out yourself in two hours flat. They could talk on your behalf, could represent corporation, corporations then 2008 hit massive um economic downturn and we then shifted our focus to um big corporations needing to cut costs and doing customer support and we were rocking it.

We were charging like 20 cents a chat and our volumes were just going through the roof in the downturn. So what I'm saying is the mountain was not bots that talk to you in. So for you in social media or customer support, virtual agents or predictive A I on the back end, the mountain was mass uptake of interaction with intelligent A I and that was the mountain and that is a worthy goal. OK? It's worthy of your time. You think that a project can be done well inside a year or two. But if you really do something, well, it's going to take five years of your life, it will um it will take a year or two to get it up and running another two or three years to get it really scaled and then you either grow with it or you exit and have to make sure it's handed over.

Well, so make sure what you're committing to is a beautiful mountain with shining snow on the top and something big that you really want to put your life into next stage and I better watch my time. I don't see a clock anywhere here. Um product um paradigm ideation. So now that you've figured out your broad trends, what your mountain is, um you now think about, OK, what's the core of what we're gonna build? You don't look at how other people have done it. Ah, someone else has done a ecommerce site with this twist. I'm gonna do that twist. No, you go right back to the ground. Go imagine yourself in a cave with a bunch of other humans sitting around a fire, scratching your fleas and think. How do, how do we do these things? How do we talk to other people? How do we make purchase decisions? How do we fulfill our entertainment needs? Whatever your mountain is? And what that allows you to do is hire differently and come at things from an orthogonal paradigm. I find as a female executive, I need to do this more. I um even though I've got significant credibility, I have four patents in my field, I've got, you know, rock solid history, I still know that I stand in line behind 50 other blokes for venture funding because we get 2.8% of the funding. So if I stand up and go, oh we're doing, you know, evolutional algorithms on, you know, data corpuses for blah, blah, blah. I'm just another iteration of an iteration. Um Plus we need more of a moat as females, we need more runway, we need more buffer.

So you need a deep technical or market moat um so that you can gain dominance and before others even realize what you're doing and they come after you, but they'll be two years behind you. So it's very important that you, you come at the problem differently in Ortho agony and have the courage to do that. So just as an example, in the early days, we were building, uh I don't even want to call them chat bots, cos today's chat bots are done, you know, single turn task performance. We were building something that people could talk to for 20 hours a week and not realize it was not a human. We built companion characters for media companies. We were building, you know, diabetes coaches that were outperforming human coaches, that sort of thing. So for example, in the early days, we thought, ok, everyone's doing natural language processing, they're doing, you know, either a rules engine or machine learning to classify inputs. And we thought, what if it's not about question answer, question answer, question answer. And we hung out with humans who were in the field as front line workers helping customers with their needs, being a companion to someone, you know, um being a health coach.

And we realize for example, that what they do is they look for these deeper variables, these little fulcrums like if someone, you know, is really depressed and having a terrible day and their blood sugar is low, then this is gonna happen. So our whole model got flipped. We we focused on this kind of underneath model that humans build and construct to make rational decisions. And that step came out of um psychology, I have a psychology degree, not a coding degree. Um Just a little anecdote. I grew up in Zambia in Central Africa and then we migrated down Africa as it crumbled. And I ended up um trying to study software in South Africa, right in the middle of the apartheid era. It wasn't just racist, it was sexist. I was an anti apartheid protester. So I really feel for you guys in America with all the protests going on. Um And I try to um get into um uh engineering to study software and I was, you can apply if you like, but there's no woman here. Um So I applied and I, I didn't get in, there was not a single female. Um It wasn't stated anywhere but I basically couldn't get into engineering. I still learned to code on my, on my own. But what I'm saying is I use what I know. Um I have a psychology background. We come at things Ortho agony. I'm not scared to use theory, developmental theory for early childhood or, you know, you know, personality theory and it gives you interesting stuff.

Um You know, my search engine, we were outperforming googled in click through rates in 2004. Um That meant we got to sign a deal with Yahoo to deliver their search results across Asia Pacific. Um Our, our A I was outperforming um all the major, major um companies A I systems in, in 2014 and that's why IBM acquired us. We, we, we rocked at work but our approach was different. I better move on. Um Next thing you do is you build the core. OK? So you gotta make sure your core is extensible. You're building something that you you got to think. I gotta apply to this scenario, that scenario, that scenario because remember the market will move and it might take you a while to build product market fit. So if your core substrate is not right, you're not gonna be able to move and iterate with the market, you've got to be prepared to throw away half of what you build. Otherwise, it's not pure innovation. It's building an app for a known market using known technology. That's hard. You've also got to be prepared to be ruthless with people. I get rid of about a third of the people I recruit really fast because you can't afford um to nurture, I don't want to say this in the wrong way.

Um You can't afford to babysit people while they figure out if they really want to be there or not. Or while you figure out if their skills can step up, you're running fast, you're running hard. People either hit the ground running and work with you or you let them go fast because we spend too much of our time looking after people who are not adding value next. As you think about your embodiments, your embodiments of your core have to be strong metaphors. So for example, when we launched our search engine, it was a, a Starburst cluster of visual maps and it was people could get straight away what we were doing with this dynamic personalization. When we launched the A I, it was really a platform, but we had to put avatars talking avatars because if people got, oh you can talk to it. It's like a human. Um Right now I'm building an advanced A I that does deep relationships, complex problems and I'm applying it to running a human home. So we're building robots for the space industry. Um You know, the robots are called what they do. We've got a helper. Um We've got a companion, we've got a manager. Um So it's important to build strong metaphors if you're doing. Now, what all of this does is gives you room. Um It gives you space, it gives you a year or two to then go and find your product market.

Fit. Um Nicole, I just need to ask I'm doing for time because I'm not watching it very well. And I'm conscious we don't want to run behind. You're

fine. I would say just a few minutes and start to wrap up. It's about you. So just wrap up. That'd be

great. OK, cool. OK. So now finally you're getting to grow, you've got all this great foundation. Um I don't grow companies on a or, or product lines on a traditional uh business plan, you know, people, product market, whatever the II I spend time thinking about what I call valuation drivers. So what is it one of the three key things we need to be doing as an, as a group or an entity to grow the value of what we do in the world. It might manifest in your share value growing or in your product uptake. But these evaluation drivers are how you frame all of your decisions. They're critical. So for example, with my A I company that IBM acquired, we did a one week emotion workshop with a core founding team and we came up with our evaluation drivers. So it was early days of A I. So it was defensible A I tech really, really cutting edge defensible good patents, that sort of thing. Next, next was fix real problems, solve, solve things for people. And the third was mass uptake. So every decision we made, someone came along and went, oh, you know, I really want to use your A I for this or can you add this widget onto it? And we go, is it gonna get us mass uptake? Is it gonna fix the real problem?

Is it gonna advance our defensible A I? No, someone wants to spend their time building something. Is it gonna meet those requirements? It also allows you to shift as the market shifted. So for, as I mentioned, we started with mass uptake, um then the market crashed and we fixed, we, we, we, we, we switched to solving problems, you know, big corporations had to pour a lot of crashes and had not enough people. Um So those valuation drivers are key. They've allowed me to basically double or treble value of what I've done every single year. Um And market downturns by the way, are freaking fantastic for females. We are capital efficient. We are resource efficient cause we're forced to be. It doesn't matter how good you are.

We all know this, you don't get as much resource dished out to you as blokes. Um But that means we're also extremely good at taking whatever resource we have and leveraging it up to value. And here's the thing. If you really, really that skill that you have and embrace it and use it on the other side of a downturn, you will be standing and 10 blokes will have fallen down and that means you will be valuable. So you think, oh my gosh, this is so terrible. How am I gonna do this? We're doing such a shoddy job over here. I don't even have people but everyone else is doing it way worse than you and they don't have the skills. I'm, I'm getting a bit gender biased here, but it's, you know, there is data about how much resource we get relative to the, the main. Um So you focus on your valuation drivers. The other thing we really have to overcome is an obstacle is we need to this billion person mindset don't do small projects. Everyone says, oh, you got to prove it first rubbish. It's as much effort to sell or scale AAA project that gets 100 million people using it as 10 people using it. So be ambitious.

Um You know, say, well, is there a way, is there somewhere where I can take this work I'm doing and apply it where it reaches 10 times as many people or 100 times as many people. Shoot for that because your most precious resources, your, your energy, your passion, put it in the right place. If you get big enough partners or a big enough goal, it's a, it's a better use of your energy. So those are really the steps um really be ruthless about the valuation drivers, particularly in, in downturns. Um Just a quick note on where I think the future is going. Um If you wanna hear that, I think um you know, that the big social drivers, big social shifts and technical shifts that I'm seeing is, you know, human beings are we just proliferating this double humans that they were, you know, just a hand a couple of decades ago. Um We are AAA planet that is resource constrained. I personally think a lot of the, the massive societal shifts we're seeing are a reflection of we, we kind of like bacteria in a Petri dish and we've outgrown the feeding trough and so your rise in nationalism, um you know, your, your pursuit of um you know, stuff to fill this kind of gap.

I, I think it's all, all a symbol of, of these, these um you know, us kind of taxing our, our, our resource base. Um We have a values vacuum, you know, our, our parents would have had AAA fairly rigid framework of values and, and uh you know, an afterlife or whatever that they believed in, you know, our society tends to value um um consumerism and, and commercialization. So I think what we're gonna see, there's gonna be more, we're gonna have to do more work hours and have less fun. Um So those are the big social trends, obviously, technical trends, you've got a i more immersive technology. Um I think that's gonna keep growing. Um I'm very interested in quantum um and, and neuromorphic chips. Um where I think that's gonna go in terms of uh how business will change is we're gonna, I see, I think we're gonna see actually more consumption. Um But we won't have money so it's gonna be cheap stuff. So any technology that's helping people get their consumption fixed, I don't admire this. That's not what I'm building now, but it's hyper personalized to make them feel special, that's gonna get somewhere immersive entertainment to help people escape. Um There's gonna be a rise in geopolitical tension. I, I would a war between um China and America.

Um So I think, you know, anyone in defense and energy will, will be doing well. Um Energy and, and renewables, what I'm building for is I've seen front line that A I can have a massive impact in human behavior. Every single bill we did, we brought between a 3200% shift in behavior. If a bank wanted more debt, we doubled the debt uptake, doubled the credit rate of credit cards or loans. If a health insurance company wanted people to be more healthy, we would get AAA doubling of, of healthy behavior. So when you take a really compelling front end A I that looks and talks like a human and gazes in someone, someone's eyes and tells them how wonderful they are and how much they care and you marry that with a deep predictive A I that can track behavior and patterns and, and really manipulate decisions.

You have one of the most compelling and persuasive technology forms in the universe. So I've actually set up a public benefit corporation. We are building A I um to run the human home home is where we make all of our decisions. It's where we live our lives. It's where we manifest our values. It takes over 25 hours a week to run a home. It's a complex, un, unappreciated job and that we're doing, it's a complex system. So you can't just use question answer A I to run it. I basically have been, um, a single mom for a lot of my career. Um, and I've had to hire someone 30 hours a week to run my home. So it's a problem space. I get to work on now because I've had a lot of success. It's not one that's easy to pitch. Um, but basically we're building a I to help you organize your life and meet your goals and free up time. We're starting out in the space industry because space is a beautiful habitat. It's deep tech science. So we, we perfecting um We're moving into vulnerable homes, people living with isolation and disability and aging. Um And then in about two or three years time, we'll, we'll launch one for every home and I want one myself because I'm so sick of getting everyone to do the chores and eat their vegetables. Um That's me.

Great. Uh Thanks again, Lisa. We're waiting on the next speaker. So that's why we're giving you a little bit of extended time. Uh It sounds like you're very successful. You mentioned that you're a mom. Um So you've talked about a lot of technical things that our audience really loved about the future of tech technology if you believe in aliens, but we'll get to that later. But as a side note, how do you balance like the aggressive success that you had with being a mom? Um I know it's not your topic, but you're so insightful and heart driven that I'd love to hear your answer.

It's back to that being ruthless with my time. Um, you know, I, I have a great husband he took about, he might be listening to two years of training before he was actually, you know, equal sharing in household task. Look up the, the US department of Labor data. It's frightening. Um, it's actually a 37 hour a week task, running a home about 27 of that falls on women. What I do is I'm really ruthless with my time. So at the beginning of the day, I don't look at emails. Um, I, I take a posted note and I go, right. Let's say I only had four hours today. What's the most critical thing I should be doing or the most critical thing my team should be doing. Then I think about that for the week and the month. Um, and I used to not switch off. Um, I used to wait till my kids went to bed and then do emails till midnight. Um, I've done, I've done this in cycles. So when I do those cycles they last for about a year and then I start making bad decisions because I get so burnt out and I'm trying to do everything and prove everything to everybody that subconsciously, you start self sabotaging your company because you want to break or your project.

You go, ah, you know, maybe I need to hire a strong blah blah, blah to lead the project and you just drop that and next thing you're out, you know, so you need to be careful. So, being ruthless with your time also means being ruthless with giving yourself downtime. So when I'm in cycles where I'm, I'm trying to do everything within a year, I start subconsciously sabotaging. So, what I do is I carve out, for example, I'm, I'm older now, you know. So, um, I take an hour and a half in the middle of every day and I go exercise. Well, actually I haven't for a couple of months. So I'm lying. But, you know, um, it feels like I'm burning nanny time or, you know, um, business time, but I'm, I'm more productive. I'm more effective. I, I switch off the computer at five o'clock and I go hang out with my family. Sometimes it means I only get a five or six hour work day. But that also means I have to hire an E A early, an executive assistant. Even if I don't think the company can afford one engineer, someone to support me, I'll hire someone to support me. And these are hard decisions for us to make because the market tells us build stuff and prove yourself.

And I'm saying, look after your energy, you're a massive asset, preserve it and cherish it because people will follow a leader, but you can't lead if you're in the trenches, doing everything,

great wisdom and we're waiting for Sunita to find her way into the session. We had a session, we had a question for you around abusing uh A I where do you think the biggest vulnerabilities are? And what could go wrong there from what you've,

oh, look, the market talks ad nauseam about bias in data. And you know, that's a known thing. What I think is way more sinister is how influential this is, is an interaction form. So just as an example, pretty much all the front line avatars, you see Alexa Siri, they're all female. Um We found we started doing large scale deployments and enterprises would say, oh, you know, the customer is always right. Um So we would build a bot that was pretty submissive and people shift their behavior negatively really rapidly. OK? So here's the thing if you have II I have no doubt. As I said that one third of our interaction in the next five years is gonna be with A I and the bulk of them are female and submissive. And I watched the data as audiences shifted their behavior, they became ruder. Um They expected um You know, I'm not just talking um adult abuse. Um but you know, you bet you're taking too long. Um You're stupid and we would watch them if you handed them over to a human, they would transfer that behavior. Now, in normal society, we have a constant modulation of this kind of behavior. You know, someone will slap you down or someone will call you out for it. If you spend five hours a day talking to A I and they say, oh, you're so awesome.

You're so amazing here. Load the abuse on you will shift society at scale. Um The, the biggest problem in all of this though is that A I it's not about data bias because we, we can, we can select data. It's about what they optimize for A I are increasingly self learning and self training towards a goal and the goal we're giving them is Amazon saying, consume more, buy more shit even if it's gonna break your budget and put you in debt. Um Facebook saying spend 10 hours, 20 hours, 30 hours a week on my platform, even if it means you don't take your dog for a walk or you know, do some exercise. We are training A I to bring about that shift in society and we should be thinking about um if you use an A I, your life should be better and this is a lonely journey. I don't know how many companies apart from ourselves trying to do this. We are going our core metric if you use this A I to run your home, are you gonna have more money in the bank? Are you gonna be more socially connected? And are you gonna be healthier? Are you gonna not watch Netflix and go for a walk? Um So that is not an easy journey because there's not a lot of money, there's not a transaction of any of the time. So I, I think those are more serious problems we have with A I mass shifting of and bad social behavior, particularly to women.

What we see in the gamer world against female um gaming coders is coming into every, every home with the kind of A I systems we're putting out

and that's why we're so grateful to have women like you leading women, like you leading with balance, uh mastering the technology we need is key. And how can artificial intelligence allow us to take what we need to be productive because you could consume internet, you could consume data all day to your point to remember why we're here. And I think what is it uh live to work or work to live either way, enjoying our life and our work balance. I love your talk. I love hearing about breakthrough goals going big, thinking about making a big impact. That's what you have done, what you will continue to do. And it is an amazing um experience to feel your energy because I know that you're behind this really making the human race better. Uh You've done that um in more ways than one and I believe in your sticky note because I have mine here for the day. I'm a sticky girl myself. So we are gonna give you a huge round of applause. We got a little extra time squeezed out, which is amazing. Uh Follow her work, Lisa is just like I said, a rock star as we come around the corner here to our last speaker that we have on the main stage. Thank you, Lisa. We'll have you exit off.