In this episode of Startup Talk, we’re joined by Philippe Beaudoin, the CEO of Waverly, the innovative deep tech artificial intelligence company building a mobile reader that understands you based on your waves.
During the interview, Philippe shares his thoughts on:
- The inspiration behind the creation of Waverly and its mission to cut through the noise and provide a personalized content experience.
- How Waverly’s Empathetic AI and its recommendation algorithm set it apart from other AI-powered news and content platforms.
- The process of how Waverly personalizes the user experience and the sources it uses to gather content.
- The future plans for Waverly and where he sees the company heading in the next 5 years.
- The challenges Waverly has faced and how it overcame them.
- The role of artificial intelligence in the future of the news and content industry.
- Waverly’s commitment to user privacy and data security.
- The response from users since the launch of Waverly and plans for expanding its reach globally.
- The future updates and features that Waverly has in the works.
This episode is packed with valuable insights and information for anyone interested in the intersection of artificial intelligence and the news and content industry. Don’t miss this opportunity to gain a deeper understanding of Waverly and its mission to revolutionize the way we consume content.
Article The perils of a huge pre-PMF Series A: My story at Element AI
https://philbeaudoin.com/2022/12/05/the-perils-of-a-huge-pre-pmf-series-a-my-story-at-element-ai/
Know someone on pain? Try The Pain Removers Pain Relief Balm a product we love and use regularly!
Other Great Podcasts:
Join us monthly for Startup Investor Drinks
Register for Startup and Entrepreneur Events here
TorontoStarts Entrepreneur and Startup Community
Starting A Business in Ontario: A Resources Guide For Entrepreneurs, Startups, and Scaleups
Automated Transcript of Waverly AI on The Startup Talk Podcast
Announcer
Direct from the six world renowned Canada’s largest city, with Canada’s biggest thinkers, visionaries and hustlers. This is startup talk featuring the founders, funders, innovators, and community leaders who’ve led Canada’s startup ecosystem right here in Toronto. You’ll hear the challenges, the failures, the successes, Toronto StartUp podcast gives you the full story direct from the entrepreneurs and influencers who’ve made a difference. Now, the host of startup talk, the founder of TorontoStarts this startup coach.
The Startup Coach
Welcome back to startup talk. I’m your host and startup coach, founder of Toronto starts one of the largest startup communities in Canada. And with me today is Philippe Beaudoin, CEO and co founder of Waverly AI, the innovative deep tech artificial intelligence company that’s changing the way we discover and consume content. Welcome, Philippe.
Philippe Beaudoin
Hi and thanks for having me.
The Startup Coach
It’s great to have you AI is a big rage, everyone’s talking about content creation. And I thought maybe we should talk about content consumption. So this is a great time to chat with you. Can you tell us the inspiration behind Waverly? Yeah,
Philippe Beaudoin
I mean, the idea of Yeah, and we’ve really dates back to when I was at Google working on some of these products. I was working on Chrome, where we have some content discovery tool there. And I just felt like the path we were on, had some problems like it led to our current content discovery platforms, Twitter, Facebook, tik, Tok, Instagram, and I felt like they were the fact we were on made these platforms really good at making me angry or addicted or is throwing me to two places. I didn’t want to go. And I was like, Okay, is there a different way to build things like that? And the question stayed with me for a long time a couple of years ago, thanks to the advances what we’re seeing today with chat GBT. So all these advances in natural language understanding, we had this, this idea for a new kind of technology to power these things that would basically keep the frictionless experience that user love on these platforms. So everybody loves the, you know, the feed of a tick tock or an Instagram. But while keeping this could give the user some control over what they get.
The Startup Coach
Yeah, I find more and more, I have less control of what I’m seeing. I generally like the experience in the format of consuming this way, the scroll and the graphic and the highlighting the text and whatever, that works really well. But you know, there’s a lot of content in there. I don’t care about
Philippe Beaudoin
Yeah, exactly. It’s almost as if you had to relinquish any desire to shape your environment, in order to have the frictionless experience that you love. And we think we can actually combine both.
The Startup Coach
And the funny thing is, most of these other platforms, they, they don’t show you, the people you’re following and the stuff they’re doing, you have to kind of go out of your way to see that. It’s like, what’s the point?
Philippe Beaudoin
That’s right. Well, this is where they all started, right? With the social graph, the idea that the best way to understand you is to know who your friends were, give you a lot of that, eventually, they figured out that well, it only captures a part of your interest, that social graph, there’s a lot of things you like that are not necessarily from your friends. And this is this is tick tock showed the way for that tick tock basically said, we don’t even care who your friends are. We’re gonna give you a very quote unquote good feed in the sense that it’s a very addictive here, because it’s easy to find addictive videos, especially short videos, but everybody follows suit. Everybody was like, Oh, we’re stuck with this social graph. And now Instagram moved away from that, and everybody moves away. But the truth is, we want a bit of boat. Sometimes I want my friends. Sometimes I want my interest. But the fact that these are built with zero control mechanisms, makes it impossible for you to drive it to where you want it to go today, or
The Startup Coach
how does Waverly’s empathic AI differ from other AI use and news and content industry?
Philippe Beaudoin
The current AI’s are the either the ones that are using the platforms, you were mentioning the tick tock, the recommender, recommendation, media platforms, they are trained on your past behavior. They look at what you clicked on, the people you’ve interacted with, which videos that you watch until the end these kinds of things. And they build a profile for you from all of these interactions. But we think that this is not very empathic in the sense that it imagines that everything you do on the platform is a perfect picture of who you are as a person. And we all know that we do things that we end up regretting. Right so for us, an empathic AI is one that recognizes these two sides of people their desire to have a good frictionless experience, but also the fact that some Sometimes they might regret the time they spent. So if you build 100% of your profile and interactions, you’re missing the second part out. So empathy is let’s have a dialogue with the user, let’s let’s try to understand them at the right moment in time without increasing the friction. But just so that we have a fuller picture, a better understanding of who they are. And nobody’s doing that nobody’s building these kinds of systems at the moment apart from us.
The Startup Coach
We’re talking about the personal user experience. And what we want to see, can you walk us through the process of how Waverly personalizes the user experience,
Philippe Beaudoin
it’s a simple three step process, when you install the weaver lead today, first thing you do is you enter a couple of keywords. So you can say I don’t know if you’re a professional interested in flight simulators development, you can write flight simulator developments, or if you’re trying to populate your blog about the secret life of cats, you say, you know, scientific articles about cat, whatever you it’s like a search box, you can type whatever you want, and there, but instead of getting a search result page, you know, a blue links kind of thing, what we really gives you is a longer piece of text that you can read to and that tries to capture the questions you might be interested in over the long time. So if you’re interested in scientific facts about Jack, maybe it will say Oh, interviews with people who study cats, or maybe like there will be a couple of questions in there. And you can select the ones you like, you say, Okay, I’m interested that question that that question that question. And what the technology if we really does is it takes everything you’ve done, and it curates a feed for you, it looks at 10s of 1000s of articles published on the internet every day, and it ranks them for you so that you have the frictionless feed experience that you love, right. And we really, but this is all based on these few keywords you type. And then the the little, you know, interactions, you had to refine, refine it. But it’s a totally new way of tuning a feed to your needs this and yeah, it’s very easy to do. And as you go, you know, we are what we are developing right now. And we’ve are the our abilities to shape that feed over time. If you feel you’re getting too much of x not enough a fly, you will have more and more ways going forward. And we’ve really to nudge the feed in the direction you want it to go.
The Startup Coach
Talking about the feed, can you tell us about the sources we really uses to gather the content and how it ensures their credibility.
Philippe Beaudoin
So I come from Google, I believe in the open internet and indexing it. So that’s what we built, we actually built an index system that is able to ingest almost anything that’s open out there on the internet. So at the moment, we ingest large newspapers, small blogs, newsletters, all the kinds of stuff you can find on Google. And we our API reads all of that the content itself understands what it talks about, and very subtle details, and then decides whether or not it’s aligned with what the user asked. But yeah, it’s it’s the entire Internet. It’s not like we do source curation at the beginning. But the source variation comes from what the users Express, if you say, I’m not interested in marketing content, well, the marketing, the more marketing feed will not. Blogs will not appear on your feed. Yeah, it’s basically we, we give that power to the user, we make it easy for users to deploy it to use it at the source. Our goal is not to filter, we filter a little at the moment, because we don’t have the technology to index absolutely everything. It’s really giving that control to the users.
The Startup Coach
With the rise of deep fakes, AI voice and chatGTP. It’s really easy to create tons of content and the content wars really heating up. How does that impact content consumption and the use of tools, like Waverly AI when users are trying to get information that they really want and filter out the spam?
Philippe Beaudoin
That’s right. There’s one finding and we really that is working really well is that people care about originality, if we find the same content 10,000 times, it’s typically gets down ranked, right? It’s typically not something people care for. And what we find is with automatically generated content tends to be very, very repetitive, very unimaginative. And it’s not like we’re looking to match the text, we’ve really eyes using the same kind of technology than what exists in this platform to understand the subtle nature of the text. And if it’s talking always about the same kind of stuff, it’s just trying to fit like a new angle in the same marketing pitch. It gets down ranked naturally, because it’s not original enough. So I don’t know how much of that will play in the long term. I think we will have to rely on trust networks. Who do we trust? Which sources do we trust, all of that will make a comeback, it’s going to be important to fight against this flood of automatically generated content. But I think there are other mechanisms to try to do it and originality is one of them. If you if you use GPT to create a totally original piece, maybe it’s interesting, right? It’s not necessarily because you use the chat GPT as a writing assistant, or I mean it’s similar at that point, it becomes similar to a spell checker, right. But if the content is totally original, and you work hard to make it original, but you use GPT, to help you and find maybe it deserves a space on my, on my feed, but it’s super, if it’s super repetitive, very, you know, no value add, then I don’t need that.
The Startup Coach
Yeah, there’s a difference between using these tools to help the creative process and then using these tools create spam that floods the market, I use it for different reasons. For example, if I’m good, this is, for me is a good use case, if I’ve got to create 100 different tweets about one event, because I got to share 10 separate tweets with each partner that’s unique to them. You know, I don’t want to write that 100 separate tweets, I can feed that information to the chatGTP. And they could do that for me. And that’s a perfect example. Another thing I like to do is outlines because even though I don’t use outlines, it comes up with it gives me three or four points I didn’t think of and I’ll add to mine, and so it helps with the blank page thing. These things are great for getting started or giving you ideas. But if you use them for total creation process, you end up with not great content.
Philippe Beaudoin
And I think you know, we already have mechanisms to weed out that kind of not great content, there will be more of it. So we will have to rely on it more. But it’s not something new. So you know, since the beginning of the internet, the constant thing that we’ve seen is more and more and more and more content and a big fraction of it is uninteresting.
The Startup Coach
How does Waverly recommendation algorithm work, and what sets it apart from others in the marketplace.
Philippe Beaudoin
So it’s in the big family of what you would call content based recommendation systems. So we actually read the content. In our case, right now we’re working with with written content. So we read the text, and we understand what it talks about, but in a lot of details, and then build our own what we call intent and understanding technology. So from the piece of text that’s generated from the user, we can extract from the the language itself, we can extract the intent, try to understand, do they care more about a topic about the source do they care about the thing being popular, being recent, being original, all of these characteristics we try to extract from the text. And it’s from this mix, you know, it’s a, it’s a big mishmash of these characteristics that we try to rank all the other text. So this sets us apart, nobody is trying to do that kind of recommender system. recommender systems are traditionally as always seeing built on trying to match your user behavior to the ranking. But in our case, it’s very much like trying to understand the intent from a piece of text. And this, this is what sets us apart.
The Startup Coach
Are there any standout features you can share by Waverly mobile reader,
Philippe Beaudoin
if you pick it up today, you get that that great experience of being able to create your own very personal lives feed, if you’re a curator for for your for a blog, or for a newsletter or something like that, it’s a super useful tool for you, just in minutes, you will get a feed right there, if you are more casual readers are like hundreds of these feeds have been made public by their curator so you can just follow one. And what’s interesting about them is they are they tend to be around niche topics like Lego for adults, or you can have some around is nicly are making a comeback. So we’d like people who have these these specific interests, but you can just follow one that already exists and you become kind of part of that community. So it’s interesting for readers to if you like to contextualize things you find on the internet for for your community or for your friends or your co workers. We’ve read the as a super feature for that if you read an article, you get into the reader view right away, which makes it super clean. And you can highlight whatever you find interesting in the article. And from there, we just let you share a summary of what you’ve what you’ve read together with the link so that the person you’re sharing it to understands why it’s interesting in your shared context. So that’s a really interesting feature. If you care about sharing healthy lead, if you want to help people understand why something is interesting, we’ve already makes it really, really easy to do that on your mobile and you basically just the time it takes to read the article is all you need in order to generate a nice to read view of it.
The Startup Coach
That sounds like an excellent feature. I’m gonna have to try that out. How does Waverly aim to combat the issue of doom scrolling?
Philippe Beaudoin
So where does Doom strolling come from right it comes from having an AI that doesn’t care if you regret the time you spend on it, right? It just doesn’t care. So basically, if you spend time the Adi assumes that you’re happy, which we know isn’t true, which we know that it’s exactly what Doom scrolling is, it’s fine new span. That doesn’t make you happy. The idea with Weaver is essentially when we were talking about embedded PKI was that it’s basically if you want limits, we want to make it easy for you to set them We want to make it easy for you to say, You know what a good session for me last 30 minutes less than an hour, right? Or it involves interacting with my friends but not with random strangers to trigger me. We’re not there yet. But this is where we’re going to kind of control we are spare heading in Waverly is the kind of control you need in order to do that, and basically craft your entire experience to the point where you can say, I’m going to make sure this environment doesn’t lead me to doom scrolling, I’m going to make sure this environment doesn’t lead me to flame wars, right. And so that’s where we want to go. That’s why we think it’s such a revolutionary idea.
The Startup Coach
Seems that Facebook and many other platforms are built around driving your emotions, they’re not happy unless you’re either angry or really happy when you’re in the middle. They don’t know what to do.
Philippe Beaudoin
Yeah, that’s right. I mean, that’s what creates engagement, right? That’s what they realize is like, hey, if I’m put you put a post that makes you angry, the probability that you comment is higher than if I put a post that just informs you but doesn’t trigger any emotion. So they they learn to, you know, to promote flame, flame war triggering tweets, right. But the thing is, it’s an unhealthy loot, because that’s what people learn to create. When you see only this kind of content, and it’s your time to tweet, he will tweet inflammatory stuff, right. And we’ve evolved, I think these algorithms have contributed to evolve the environment we’re in where people start flame or ours, and people consume flame wars or whatever, right, these type of triggering content, but by saying everybody isn’t controlled, you can go back to it, you can if you realize your experience is not what you want it to be, you can shape it, you can change it, keep the friction lasting, everybody loves it, I we’re not taking that away. It’s important, but you shave it.
The Startup Coach
Can you discuss any challenges you’ve had with Waverly and how you’ve overcome them?
Philippe Beaudoin
The technical challenge was big to cost quite a bit of time to develop that kind of technology, the things that make a feed interesting to someone, it’s very subtle, you cannot do it just from a couple of search terms. So yeah, just just figuring this out, the first version of we’ve really, you had to write everything yourself. So it was you had to write a pretty long paragraph, or we’re ready to start understanding what your intention was kind of knew that ChatGPT was coming. So in the latest version, we just released in January, we actually integrate with GPT three, which is the engine behind ChatGPT. And so now we support this, this fluid, like talked about there in a year when you can where you can just type three words and get what you want. So yeah, overcoming that challenge, where the first version of we’ve really had too much friction in order to get it started. That was pretty challenging. But it’s super cool that, yeah, now we basically have just this new technology to build on.
The Startup Coach
Definitely a chat technology is much better than filling out a paragraph or filling out all these forms and stuff. It is friction. And, you know, I teach growth, hacking or whatever. And you got to bring that friction down to zero because desire minus friction equals conversion, right? So
Philippe Beaudoin
that’s exactly,
The Startup Coach
how’s the response been from the users?
Philippe Beaudoin
Oh, I got two users love it, what we have out on the i iPhone store today, I would call an open beta, right, it’s just like part part of our vision getting a we just wanted to put it out there early, but people already love it, they create feed that matches their their super niche interest. And it’s the first time that they have a healthy space with only that in there. And that’s the kind of feedback we get. It’s not working for everything yet, you know, for a bunch of different reasons, limitations, that technology limitations and what we’re indexing. But when it works, the people just just love it, and they come back and they have this feed, and we see them like coming back again. And again. And again, what we get asked more and more often is I want to bring my community in like it feels lonely. But there’s, there’s a bunch of us who cares about flight simulator development, there’s a bunch of us who want to become better at building 3d games, or whatever, right? And I’d like to bring my friends in, or I’d like to start my own thought leadership space here. So we’re going to be building more and more of that going forward. So you can you can do that. So if you’re a curator, you can you can use it to curate content for your communities outside if we’ve already, which is our primary target. But you can also maybe gain new followers, and we really find new people who care for your for your things. Yeah, at the moment, I would say when you use it, it feels very fresh. And that’s the kind of feedback we get from our users.
The Startup Coach
Can you tell us about the role of artificial intelligence in the future of news and content industry?
Philippe Beaudoin
Yeah, you talked a bit about it when you said it’s going to help generate content, and we’ll probably see it used as such a little bit. I think the first the first use is going to be as a way to see your imagination, like even the mid journey and things like that they’re really good at letting you explore a bunch of variation in idea. I think we’ll see the same for written content which ChatGPT And then you start a more complex process to produce the piece that you want to produce, right? The, the article or the piece, the piece of visual art or whatever. But yeah, I think the future is going to be exploring these, this this new world. One thing that doesn’t get talked about a lot, and that we strongly believe in is that our ability to understand natural language means that we will have more control over some apps or some technology that were traditionally really hard to use, because they were limited to experts, there was too many things to control. So yeah, a future of future Photoshop will probably have some mix where you use the mouse a little bit, you use the chat a little bit, and you converge more quickly to want what you want to do. And we’ve really is directly that essentially saying, hey, the future of our recommendation media, social media is going to be this mix of Yes, I have the frictionless experience. And I have these moments where I can nudge this experience. Because the truth is, these platforms even today, they’re the way in which they are configured to be super personalized for you. It’s very complex, it’s a lot of parameters behind the scenes that you never get to see that are powering your Facebook feed that are powering your Twitter feed. And all these power meters, there is no there is no settings page for it, because it would be overwhelming. But with natural language, we think you can finally start configuring them. And we think we will see that in a bunch of other spaces, it’s going to be a super fun time to discover this.
The Startup Coach
I could definitely see a a interface into Photoshop to do what you want to do. My mind automatically goes to excel and saying, you know, just being able to Hey, take this dataset, are these datasets create a pivot table based on these things and show me the graph. And like being able to do that rather than figuring out manually how to do all that stuff? Just telling a computer or something to do it is going to be so much better for us in the future?
Philippe Beaudoin
Yeah, it’s good. Like, I don’t know it. Will Clippy make a comeback? Who knows?
The Startup Coach
I think about this, and I’m a little scared, because I think we’re entering in a new age of the internet. Because what happens to search engines, when AI takes over search engines monetary is based on you clicking around everywhere. And if you tell AI and they give you an answer, you’re not clicking around everywhere. So what does that mean to Google and Bing and whatnot? And then as all this content is out there, you know, do you trust this particular AI that you’re asking? Is it going to feed you the content that you want? And you know, is it going to be I got to use the best AI to filter out the content against the AI that they’re using to break through my filters.
Philippe Beaudoin
Yeah. Is it an arms race the AI against AI? I don’t know. But I think we we’ve entered though i i think it’s an iPhone moment. You remember your touchscreens before the iPhone, you know who nobody would have taught you could like eventually had a keyboard and a touchscreen, right? I remember going in the airplane and trying to change the channel and just banging on it didn’t work. But as soon as the iPhone was introduced, suddenly the imagination of the entire world was captured. And everybody figured, ah, touchscreen work now. Right? And the fact that new experiences that were enabled were not only because of technology made them possible, but because you like people, you and I suddenly became aware that it was possible. I think we have a similar moment now with with ChatGPT and Andy’s chatbots, you know, we will look at the clippies of the past or chatbots of the past and say well are sure nobody wanted to interact with them, they were so bad. But now what we think we might be getting in that moment where people will eventually trust them enough or trust that they will have a good interaction with them to accept them, embrace them. And that’s what will will cause the new the next wave of tech the next wave of so will will this translate into an arms race or content versus filtering? I don’t know. I think I think the arms race has been on since since a long time. But it had a bit less AI in it and more a bit more humans but you know, content farms have existed since forever. And we develop mechanisms to try to counter them. It’s always a mix of technology AI like Spamfilter kind of thing, human moderators who are using these technology to do it first fast. And then social mechanisms like trust networks, who do you trust? Who do you follow? Who do you have vilify? And I think it’s going to be again a mix of all of these things that that help us tackle the next wave with it. In fact, search engine for sure, but I’m with Janna German this one. I think that the current technology is really good at capturing our imagination. But there’s a huge huge gap between like my background is in AI. I did my PhD in AI so I know that stuff a little bit, but there’s a huge gap between the technology that can power ChatGPT and technology that could do something like that, but always grounded in truth always grounded. In sources, it’s just not the nature of the research product that’s powering chatGPT. It’s not good at connecting to a knowledge base. So clearly Bing is trying to do it today. But they’re trying to patch it our way through it. And it’s all full of holes. I think we need more research. And it’s very unclear how much we need and when we will make that step. But yeah, I think that’s why Google was very hesitant to put it out. People are saying, Oh, it’s going to play with their business model. For sure. It will, you know, it’s much harder to monetize your ad bluelinx When you have a chatbot answering the question for you. But it’s not only that, it’s because it’s not the right technology for knowledge seekers, it’s not the right technology for knowledge gathering, it’s good at chatting, it’s good at making up sentences, then sometimes these sentences are true, but there’s no guarantee that they will be.
The Startup Coach
Yeah, it’s funny that you mentioned that I was ranting about this the other day, but I’ll mention it again, using chatGTP. I was asking for summary of something 200 characters or shorter. And the first response was 373 characters, and I said no 200 characters or shorter second response 313. And then I had to explain what a character was. And it came back to 76. And then I said, how many characters is that? And I said, 200. I said, No, let’s do 276 And and finally came back to 200. I’m like, really?
Philippe Beaudoin
Yeah, that thing is not built on. Understanding logical things. You know, it can do mad, but it’s it’s almost by accident, right? I cannot do it all the time. You ask if you ask him that question with with two large numbers. It’s not gonna work. So it’s really weird. It’s really strange. I mean, even as researchers, I would say, we don’t understand why these things work. It’s a bit it’s a bit shameful for researchers to admit that but it’s true. We build these machines we started with, with deep neural networks. And then we evolved and we evolved, and we evolve. And now we have that. And it’s like, Yep, it was inspired by the brain. That’s how we started. And still today, we made a lot of intelligence choice choices along the way, based on what we learn. It’s not it’s not total random exploration. But it’s not like it’s it’s grounded in super strong first principles. It just happens to work. Yeah, it’s a real stretch from where we are to a system that could say, hey, this thing is working is able to apply a logical reasoning it’s able to ground with it says in a factual database of facts or something like that. Yeah, we’re far from that.
The Startup Coach
What are your plans for Waverly for the next few years? Are there exciting projects going up, you’re improving the AI what’s going on?
Philippe Beaudoin
Pretty, pretty exciting time for for us, we’re raising our seed round at the moment. And the goal is really to invest more on the curator aspect of the platform. So we can start start really powering What the what users are asking us for, which is just a great place to find content aligned with, with our interests. So we have like people who seek knowledge, they can’t find a good but data can’t find a good video, but they have a specific interest, we want to be the best platform for them. And so the way we do that is just make sure we are the best place for curators like your reader. I mean, people who I don’t know have a mailing list already published on Twitter or the the user things like HootSuite or HubSpot to try to get a really good feed for their audience we want we want them to come to every because it’s the best place for them to find content is the best place for them to write these little knowledge nuggets for the people they care for. Once we you know, once we build a next best experience for them, and they start coming to the platform, this is where we start realizing our dreams. So this is where we’re going for the next eight to 12 months. So if you’re one of these characters, and you are curious about bringing better content to your users, check us out, give us a call. And I think we can be a great starting point for you.
The Startup Coach
I was gonna ask you about your funding situation. But you’re currently raising your seed round. We are if you are in Toronto next week, it’s a startup investor drink. So I’ll send you an invite, and you should come out we have a bunch of investors from different organizations coming out in startups get the pitch and it’s a networking thing. I don’t think you’re in the area though.
Philippe Beaudoin
I’m in Montreal, and I’m traveling around for that seed round the next few years, the next few weeks.
The Startup Coach
Good luck with that. Thank you. So having worked at Google being a serial entrepreneur, founded Waverly, what’s your favorite book for entrepreneurs?
Philippe Beaudoin
Yeah, I love that question. I mean, I’m gonna answer with a book that’s not that’s not for entrepreneur, but I think it’s a super important book. It’s Stand out of our Light. It’s a book about Stand out of our Light is a book about the attention economy. It’s the best book about the attention economy. It’s really whether you’re a leader or whether you’re a knowledge worker, no matter where you are, you are embedded in it today. Everything out there is trying to grab your attention and if you’re a leader, you’re trying to get grab the attention of others. But this just goes so much into the mechanics of the attention economy, why some of them are bad, you know, and how can we work with them or get you know it? If we care about our attention, how can we can we understand it enough so we can subtract ourselves from it. It’s a fascinating book, it’s a short read, it’s really going to help you understand the world we live in. When it comes to knowledge gathering, by convention. The second one, the book that has always been really, really close to me is Thinking Fast and Slow by Daniel Kahneman. For me, it’s an essential one, it really goes deep into these two modes of thinking that we all have, like our ability to make snap decisions, quick decisions, super important. And then our ability to think to sit back thinks reflect on our lives and what’s what’s important on it, and you need both no matter what you’re doing, but especially as an intrapreneurs, you need to understand how you work with these two things. So they are related and assess these two books, but they really capture what’s the most important for me and what I had to work on the most to become the leader I am today, I guess, or the intrapreneur. I am and the problems I’m trying to fix with the world.
The Startup Coach
Do you have any tips for founders?
Philippe Beaudoin
Tips or father’s a? Here’s one, some someone when I left Google, I went to talk to someone who’s now one of my mentors. You had been at Google too. And I was you know, debating do I leave Google, I left Google to fund a company called element AI. And what he what he asked me the question you asked me is make sure you’re running towards something not away from something. And I love that for me it really it to me into some of the most important question you have to ask yourself, because as a founder, it’s a tough journey. It’s a long journey. It’s an I know everybody’s saying it, you don’t know what it means until you try to do it. Even myself. You know, I try. And I forget, and I try again. That’s part what you need to be really, really sure of is that you’re not running away from something but you’re running towards what you’re creating. And don’t lie to yourself. Just like go there, figure it out for yourself and, and then do the do the tough things.
The Startup Coach
Yeah, we didn’t even talk about element AI at all, is there a lesson learned coming out of element AI, your big exit,
Philippe Beaudoin
I mean, the lesson learned is, it’s the best decision I made to leave Google to do this, I learned so much out of it. But in terms of business lesson, I would say element AI was a super interesting but bold play, I wrote about it a little bit, if you’re interested, I can send you the link to a blog post, I wrote about it that had really good responses. And when I talk about it, there is essentially the perils of raising a huge round a huge venture capital round before you have product market fit, which is a, you know, some something some companies do when we did that element. And it’s exciting, you know, because large rounds are kind of self fulfilling prophecies, it’s dangerous, it creates a bunch of internal tension, because you don’t have you’re not directly connected to to a product that generates values. For some some people, that’s very risky. It’s very hard then to focus your energy on building, building good product, especially since the first step is to find the product. So find a product find find the customers for the product. And then you have to find something so big, because you over promised so much that every little thing you find looks too small, very risky. I like we’ve really, we took a radically different approach small precede round, we raised a million dollar small precede round product product product. I mean, we had the technology to build, so build the technology, but right after its product, product product, look for PMF, who loves that work with them? Like it changes the dynamic very much. And the reason I did it like that, which we read is total rebound, total rebound from from element. Yeah,
The Startup Coach
I might want to have you back or even in a personal live event just to talk about your element III journey. And even just that topic in the broad blog posts of warning people about, you know, raising, you know, $100 million before you have product market fit and trying to figure it out. That’s a great session.
Philippe Beaudoin
I’d be happy to do so. But there are a couple of things that I might have to check with my lawyers.
The Startup Coach
We’ll we’ll put it six months a year you’re raising now we’ll wait till you’re done. And then yes, obviously, you can check with your lawyers, but so you can edit out anything that’s gonna make trouble. If people want to go and find out more about you Waverly download the app, where do they go?
Philippe Beaudoin
Mywaverly.com. It’s our landing page, go there with your iPhone. And then you have a link to the webstore where you can the iPhone store where you can get it or just search for we’ve really read with intention on the App Store. And you can just install it from there, give it a spin, try it out. Tell us what you think we’re very much looking forward to that kind of feedback. And if you’re a curator and you’re interested in using it, to power your community, we especially want to talk to you so don’t be shy reach out. I’d be happy to have a chat with you.
The Startup Coach
I really appreciate you time to leave today talking to talk about Waverly AI, and all the wonderful things you’ve been working on our audience really appreciate the insights. It’s been a wonderful conversation.
Philippe Beaudoin
You’re very welcome. It was a pleasure to be on.
Announcer
This has been startup talk Toronto StartUp podcast for more exclusive content, the episode vault and to be part of TorontoStarts community, visit torontostarts.com. Get your name on the newsletter mailing list and check out our upcoming events. For more episodes subscribe now, and please recognize the time and work behind the scenes put into connecting you with the biggest visionaries entrepreneurs and innovators in Toronto by leaving a five star review. Join us for more next episode from Toronto’s most active entrepreneur and startup community on startup Talk.
