AI Strategies for Startups with Mike Grandinetti

First of all, it's always great to present to such a global audience, and obviously we have a lot of people that are with us today, so not surprising, it's an incredibly exciting topic. I do wanna talk about artificial intelligence specifically for startups. And let me just spend a little time talking about my background so you know where I'm coming from. I, I have been a Founders Institute mentor. Um, I started my career in Silicon Valley as an engineer. After my MBA at Yale, I worked at McKinsey for several years out of New York City, and then I helped co-found 8 enterprise technology startups. 2 went public on the NASDAQ, 7 were acquired by companies like IBM and Oracle and Symantec. And over the last 20-plus years, I've been teaching in addition to being very active in the startup world at MIT, at Harvard, where I teach very actively today, at Brown, at Cornell, at Carnegie Mellon, at Babson College. And so, you know, bring a lot of expertise to the world of startups, having done 8 over 25 years. And today I do a lot of consulting to large global companies on innovation. So you'll recognize a lot of these logos. So I spend a lot of my time consulting in the tech industry, and I also spend a lot of time consulting in more traditional industries— automotive companies like Ferrari and Stellantis, medical tech companies like Medtronic and Quest, financial services companies like Capital One and AXA. So this is kind of what I do. When I'm not giving talks for the Founders Institute. And so we're all here as entrepreneurs because we have a dream. We want to create the next great tech company. And if we go back, and you're all global citizens, many of you probably have heard of the great Niccolò Machiavelli who wrote this 500 years ago. And it's probably more true today than it ever is, right? There is nothing as difficult or as dangerous to do than to undertake to change the order of things. And of course, that's exactly what founders do. That's exactly what startups are designed to do. To change the way that we work, to change the way that we live, to change the way that we interact, right? We're changemakers. But getting people to change is hard. It does not come naturally. And so we might come into our startup with a huge vision, like we really want to put a massive dent in the universe. We want to leave a significant legacy, but we need to start small and scale incrementally over time, right? One step at a time. And so let's get into the content now. Okay, let's talk about how transformative AI is and how much more it will be in the years ahead. Let's talk about the fact that AI already has reached a very interesting point of commoditization as an underlying element because of the models that are becoming increasingly similar and interchangeable. Very importantly for all of you as founders, Let's talk about how you can build a defensible moat around your startup if you're using AI, because it's not straightforward. And then let's talk about— now let's look inside the company. Let's not think about the product now, but let's look at how you might use AI as a force multiplier inside of your company for everything from coding to marketing and sales and customer support. Okay, so I don't think I have to convince anybody that we are living in the age of AI, right? It is here and it is now. Okay, and so I'll ask you to complete this poll, um, if you would, um, just to let us know whether you're actually using AI inside of your operations today. Are you using it to code? Are you using it for sales? Are you using it and your customer support ops? Are you using it in multiple parts, multiple functions of your startup? And so, Amelia, I don't know if you're able to share the results, but, um, I'm going to keep moving forward. Right, and so this is a fascinating chart. Some of you, I hope many of you, know the Gartner Group. The Gartner Group is without question the most respected IT analyst firm in the country and in the world, literally. And they generate billions of dollars every single year by consulting to C-suite executives about their use of next-generation technology. And on the right hand, you see, beginning with the, the very first days of digital disruption with the advent of the World Wide Web and the Mosaic web browser from Netscape and early companies like Amazon and Netflix, the single biggest digital disruption to ever happen until recently was the introduction of the World Wide Web and the democratization of access to the world's information. And it is considered to be a massive disruption with literally 10 to the 5th secondary disruptions. Now, what's fascinating is that Gartner is listing generative AI already just a year and a half into its introduction at being almost at stage 4, right? There's a lot of hype around a lot of technologies, and Gartner is not a firm prone to hype. They're very pragmatic in their approach. And when Gartner lists generative AI at nearly the reinvention stage after year 1, when the World Wide Web has been around for 30 years, tells me everything I need to know about how disruptive this technology is and will continue to be. Okay, some of you may have been blown away by the OpenAI text-to-video app known as Sora. Okay, and what you may have seen is the American movie mogul Tyler Perry, who has one of the most sophisticated of all of the Hollywood studios in the United States, and who's made Marvel movies, and he's done a lot of, um, a lot of superhero movies, including Black Panther. When he saw the demonstration of Sora, he instantly decided to put a nearly $1 billion expansion of his studio on hold. That's just how profoundly disruptive this technology is. And so Of course, distribution of funding is never equal, right? Nearly $4 billion went into 240 deals in the US. Another roughly $2 billion went into 120 deals in Europe. Um, another $1.2 billion went into about 170 deals in Asia. And we see far smaller amounts in Africa and in Latin America, right? Measured not in billions, but in millions. So obviously it's not an equal distribution of funding. And for those of you who pay attention, just this morning the Sovereign Wealth Fund of Abu Dhabi announced that they're going to get far more aggressive, and rather than invest in other vehicles, they're going to invest directly. So I, I suspect we're going to see a lot more, uh, investment coming out of MENA very soon. Okay, now Tufts University recently did a study that was looking at the technical readiness of AI nations. The acronym is TRAIN. And they looked at 4 different variables including the amount of data available, the compute intensity, the amount of IP. And the United States was significantly ranked higher than any other country by a very significant measure. Okay? And so there's no question that a lot of countries are going to compete heavily in AI. You'll notice that the UAE and Saudi Arabia are not in here. Singapore is not in here, and that's not because they're not making massive investments— they are— but because they just didn't have enough data to allow themselves to be ranked given the small size of their respective nations. Okay, and so now I want to shift to A very critical insight for all of you building startups using AI technology. Of course, water is a commodity. And you know, you can get water from your tap, you can get water from a fountain, and you can find hundreds and hundreds of varieties of water. But it's just water. And if you think about what Perrier did, right, they branded water for the first time. They made water part of a lifestyle. What with AI, it is so easy today to have even fairly unsophisticated engineers get access to an API and within the course of an afternoon create an AI app. Okay, so it's— it's— this is something that has never been possible before. And as a result of that, we're seeing a lot of commoditized startups today. There's very little differentiation across the vast majority of startups that are using the same foundation models for their startups. And so what are the challenges, right? And I spend a lot of time on panels with venture capitalists. I'm on the board of a number of startups that are raising money. I invite VCs into my classes to talk. And what we hear consistently is that they have looked at thousands and thousands of business plans, listened to thousands and thousands of pitches where GenAI is sort of at the center of the startup. And most of them are almost identical under the hood, right? Because it's becoming impossible to distinguish yourself now When it comes to access to data and models, right, everyone's grabbing the same models, whether it be from OpenAI or whether it be from Google or Facebook or Mistral or Cohere or Anthropic. Okay, so the challenge is, is that of course plays directly into the hands of the big incumbents. And the only way that a startup today is going to be able to have some kind of unique competitive edge, create some kind of moat is it's gonna have to have some insights, uh, because it's not typically going to be able to compete on access to data. And so just a couple of data points. If you look at the winter 2024 cohort of Y Combinator, 70% or 175 startups are building models on top of generative AI. There are two real themes coming out of this current winter cohort that just had their demo day. A number of them are attacking vulnerable components, incumbents like Salesforce and ZoomInfo. You're seeing a lot of sort of revenue ops related types of startups. A number of them are attacking customer support-centric use cases from companies like Zendesk. And you're seeing an increasing focus on niche verticals, which makes a lot of sense because these will not be nearly as competitively intense as some of the more common mainstream ones, right? And IMD, the, uh, the business school out of Switzerland, um, you know, has done a very good job of categorizing again hundreds of different apps just for productivity, all built on common models. All using generative AI, right? And it's really hard to distinguish. And so here's just one more example, right? If anybody out there is creating an education technology-based AI company, Reach Capital, one of the most active early-stage edtech VC funds, has identified literally about 280 companies in the edtech space, 116 alone in personalized learning, another $100 million in career and skill development, which is where a lot of the funding has come from, right? So you're seeing about $400 million, close to half a billion dollars coming, you know, into this one space. So there's a massive amount of competitive intensity. Of course, as founders, you need to think very clearly about how you're going to distinguish yourselves. And so if we learn from history, some of you may have followed the history of the sequencing of the genome. And in the early phases of that technological revolution, it costs many millions of dollars to do genome sequencing. But eventually that technology became commodified. And you can get a $100 genome sequence. Well, it turns out in the world of generative AI, we got there on day one. Right? We don't need for this technology to become inexpensive and accessible. It already is, right? When, when OpenAI put out 3.5 and it's out there for free, or you can get free versions of Claude, you can get a free version of Gemini Plus, you can get a free version of Copilot. It's already commoditized. So one question I really encourage all of you to ask yourselves is if you took AI out of your pitch deck, have you still really built a fundamentally strong, scalable, some— sometime in the future profitable business? Are you able to think about delivering value outside of the AI solution and build on top of it? Okay, and just to put a point on it, many of you probably heard of the company known as Inflection. You couldn't ask for a better pedigree. The gentleman on the left in the glasses is none other than Reid Hoffman, one of the early team members of PayPal, the founder of LinkedIn, a multi-billionaire, formerly a venture capitalist with one of the most prestigious firms in the Valley, Greylock. And on the right, Mustafa Suleiman, who was one of the early guys at Google DeepMind, without question the most sophisticated AI research firm on the planet. And these guys raised literally billions of dollars at a sky-high valuation. And in the last month, realized that they just were not getting any kind of traction. And they basically decided to sell the assets of the company to Microsoft. And Suleiman, the gentleman with the beard on the right, is now running consumer AI for Microsoft, right? And so when a company with this pedigree and this level of capital funding decides to throw in the towel, it's a real signal about how unique we need to be. And just to put a point on it, some of you probably had heard of Stability AI, which was a very hot company when it came out about a year ago. They had a, just a billion-dollar Series A. Okay, and they were competing with Midjourney and DALL-E, right? This was a text-to-image company, and they've been having a lot of challenges. They've really struggled, even though they were using an AI-based model, to, to get any kind of significant revenue traction. And the CEO and a lot of the leadership team quit, and this company is in serious jeopardy. So all of this is to make the point that we need to be very focused about how we're going to create differentiation. Because let's face it, if a startup can't create differentiation, there's no reason why a customer won't go buy from an incumbent. It's a much safer choice. Okay, so what is a business moat or a competitive advantage, and how do you identify them? Well, let's look at some opportunities. Okay, McKinsey came out in January and said that generative AI is the single highest priority for global CEOs this year, 2024. Okay. Secondly, right, I'm a judge in the MIT CIO Leadership Awards, and every year I get to look at plans that CIOs are submitting to be able to be named the CIO of the Year. And every single plan I looked at, it was very clear generative AI is a very critical part of the agenda, of the roadmap, of the budget, of all of these CIOs. So it's not as if there's not opportunity. IBM launched its own half a billion dollar enterprise AI venture fund. IBM has never done anything like this. They've never been an aggressive venture capital investor. Okay, so we're seeing massive opportunity available to us if we could just figure out how to create value. And so what are the implications for startups, right? You're going to need to have a very clear, very validated vision and perspective on how you're going to win. And I'm looking at the green emoji on the right here, okay? You're going to need to give your prospects a very big why. Why work with me? What is it that I'm bringing? What insight do I have that others don't? Okay. And so it's gonna require you to think very holistically about not just the product, but where is distribution going to be an advantage? How do you create unique branding? How do you create unique messaging and marketing and go-to-market and pricing and customer service? And the beauty is generative AI can help you with all of those different things. And so I, I gotta believe that all of you know Lean Startup methodology and you understand the value of the Lean Value Canvas and constantly iterating on getting a good problem-solution fit. This is foundational for any founder. I hope that all of you know the Value Pyramid that was created by Bain, one of the world's great strategy consultancies and understand that there is a B2C as well as a B2B value pyramid. And there are multiple levels. Think of this as Maslow's hierarchy, but only for, you know, creating value in the startup space. And of course, the higher in the pyramid that you are at delivering value, the more likely that you are to be able to really get the attention of a customer and to have some kind of sustainable differentiation. And I also hope, because this is critically important today, is that you are well-versed in the world of human-centric design, or what some people call design thinking. This is a process that has really allowed companies to get to the essence of what customers care about. The really, the very first commercial practitioner of this was none other than the late, great Steve Jobs. And it is the mindset of design thinking that allowed Apple to kind of create the very unique and long sustainable differentiation that they've had in the market over a period of 40 years, right? So these are just critical methodologies that if you're not familiar with them, you need to become. If you are, but you're not really expert in using them, you really need to do so. Okay. And so what is the winning formula today? For a company building on top of a data model. Obviously, the data plus the model is a selection criteria that you're going to make. Some of you may choose to go with an open-source model from Facebook or from Mistral or now from Groq, the company known as XAI from Musk, or you may decide to go with a proprietary model from OpenAI or Anthropic or Google. And that's a very small part now because those are more or less the same, right? But of course, the user experience. How are you creating a world-class experience for your users? Think about the new search engine called Proximity. I don't know how many of you are using it, but to me, Proximity is a brilliant new approach to search because it just gives me the answers I want. I don't have to scroll through line after line of advertisement. And snippets and all kinds of other related junk that just gets in my way. And it provides all of the sources to all of the, uh, the reference documents that I need, right? It's a brilliant user experience, and I've more or less stopped using Google Search as a result of it. Okay, one of the cool things about the company known as Proximity is, um, you know, they're working concert with this fascinating new AI device called the R1 Rabbit that some of you may have seen. That was really the most compelling announcement at the 2024 Consumer Electronics Show and is going to be shipping starting next week around the world, right? So these distribution agreements are critical. And so how do you create a very clear perception of value to your customers? Thinking about those value pyramids I just showed you. And this is really where the founding team needs to have a diverse set of perspectives. Of course, you need the hacker, the techie. You need the hustler, the commercial person who's out in the market. And you need what is that third H, which is sometimes called the hipster or the designer, right? You need to bring all of those perspectives to your strategic product direction. Okay, now clearly most of you are not going to be working down at the data or the infrastructure or the foundation model, right? This requires billions or tens of billions of dollars of capital, and your opportunities lay at the app level. Or of course, we're moving into the era of autonomous agents, and there's huge opportunity for startups to start to create some really compelling end-to-end agent type of experiences for your users. Okay, where do you look for opportunity to distinguish yourself? You can look for it starting at the top within the business model. You know, can you provide information-based service extensions to existing applications? Can you bring AI to existing multi-sided platform businesses? Right? Can you use AI to create much better experiences like proximity has done? Can you provide access to customer intelligence with real insight? Create more emotion. I'm working with a startup right now that's creating more of an emotional experience around the use of social media. Right? There's, there's an endless opportunity for each of you to create some unique value, but it's gonna require a lot of heavy thinking. Okay, and so now I'm going to pivot to the second half of my presentation, which is how do we now bring in AI into our startup to optimize our operations. Okay, and so let me just start off with a really interesting prompt. Many of you probably have heard of the great Wharton professor Ethan Mollick. He's been doing a lot of publishing around AI. This was a prompt that he gave to his students in a Wharton entrepreneurship class, right? And he basically is asking his students to come up with understanding potential problems and corresponding solutions that would be for people who make less than $50,000 a year. Are in a certain age demographic, and then coming up for interesting ideas for the company. And here's what happened, okay? The MBAs were competing directly against ChatGPT. The MBAs were asked to come up with 200 ideas. ChatGPT was asked to come up with 200 ideas. So in total, 400 ideas. And what they then did is they went out and they used customer intent to purchase surveys, and they were able to take those 400 ideas and narrow them down to the top 10%, the 40 best. And guess what happened? 35 of the top 40 ideas actually came from that very simple ChatGPT prompt. Only 5 came from students. And so there's a couple of very important implications for all of you in the audience today. The first is that because of the massive amount of data within ChatGPT, it's gonna have context and perspective that nobody, no human could ever possibly have. Okay? But more importantly, you can take advantage of this same exact kind of prompting within your own model about how do you come up with the right idea, the right type of approach within your startup to identify some unique capability that you're, you're looking for. Okay, he also found that the next step was, now that we know who our, you know, what our sort of our target market is, how can we begin to identify specific personas that we can begin to market to in a very precise way, right? And of course, for those of you who are in marketing or advertising or design, personas are a foundational element, okay? And this became such a powerful insight that set this— this has now been used just in the last few weeks several thousand times by many different companies. And these are, you know, MBA students that are not even working in, you know, real startups yet, and they're already beginning to create this level of unique value. Okay, so let me give you a more sophisticated example. In the Boston-Cambridge area, there is a startup, a VC fund— it's really a venture fund— it's called Flagship Pioneering. And Flagship is a fascinating company, right? They were behind the Moderna vaccine as one example. And they draw an incredible amount of talent from MIT and Harvard. And so what they, what they did is they said, you know, as a venture studio, they are always looking for what is the next life sciences medical startup that we want to build in our portfolio to create another billion-dollar company. And in the past, that meant very tedious brute force work, pounding their way through thousands and thousands of academic journal articles and scientific literature. And now what they've done is they've brought some value to a large language model, and they're now able to automatically sort through literally hundreds of thousands of pieces of scientific literature. And be able to identify in almost real time where they believe there are opportunities to build the next great life sciences venture. Okay. And it even, you know, just like proximity, it provides you with the scientific sources so that there's no, there's no question about it hallucinated. You can go right back to the source. So this is a very powerful model for anybody that's thinking about, you know, if we're, if we're starting from scratch, Where do we look? So all of these examples together, I think, should give you a lot of confidence that you can really have a much more strategic approach than you ever might have, given the resource constraints that startups have. Okay, of course, raising money is a very important part of the journey, and here is a startup, you know, fundraising market map. Pitch decks, financial models, investors. CRM tools, right? And now you're seeing an increasing number of them being AI-powered. And here's another example from, you know, from what's happening in the real world. A group of students were asked to give pitches in the same Wharton classroom at the end of a semester, and the people that was listening to the pitch— one was the professor of entrepreneurship, the second was a real-world VC, and the third was OpenAI. And people were communicating, right, using multimodality. They were speaking into what was called OpenWhisper, and OpenAI was listening in real time and, you know, and doing its research and providing incredibly valuable feedback. The VC and the professor both said this feedback is extremely informative and realistic. So imagine as you're, you know, as you're preparing your pitch to get into an accelerator, to go out and raise a seed round or an angel round, imagine having the opportunity to just, you know, speak your pitch directly into a large language model and get feedback. How much more prepared you're gonna be. When you go out and do that. Okay, so let's talk about some use cases, some other use cases now, now that we're getting beyond sort of the foundation of your startup. Okay, McKinsey put together this chart, and what you're seeing is that there are some universal cases— marketing and sales, customer operations, and software engineering— that work pretty much across every major industry. These are really the ones that, you know, we're seeing a tremendous amount of early success with. Okay, and so let's talk about software engineering, which is considered one of the, I think, most impactful use cases. And what's interesting, and some of you probably have already been using GitHub Copilot or other, you know, auto code generation tools, but the results are phenomenal, right? So right now, about 1.5 million developers are using GitHub Copilot every single day. Okay. And what is the— and, you know, and it's not just junior coders. You may not know the name Andrés Carpathy, but he is a living legend in the Valley. He got his PhD from Stanford. He was a senior researcher at Google DeepMind. He ran robotics at Tesla, and he was a senior researcher at OpenAI. And he said that he is writing 80% of his code now on GitHub Copilot with 80% accuracy. So think of Andrej Karpathy as sort of the LeBron James or the Caitlin Clark of coding. And if it's good enough for this guy, it should be good enough for anybody else on the call. Okay, and there's obviously a lot of other variations that you can get. It's one of many, but of course a very well-known one because it's built on top of OpenAI. And so what are we seeing here? Well, this is really where it gets fascinating. Generative AI is dramatically reducing the amount of time it takes to complete tasks. I don't know if I've ever met a coder that likes doing documentation. Well, the good news is OpenAI's GitHub Copilot allows you to reduce the amount of time it takes to document your code by up to 50%. It reduces the amount of time it takes you to generate clean code by 35% to 45%. Code refactoring, of course, is a much more sophisticated activity. This is when you are really going back and you were you know, making your code a lot more efficient. Um, and it takes a lot more of an architectural mindset and a lot more of a sophisticated approach to do that. And yet you can reduce the amount of time it takes to refactor your code by 20 to 30%. Now here's where it gets really interesting. One of the things that using GitHub Copilot has done is it has dramatically improved improved the mindset, the attitude of coders. So you see this, this point of demarcation between negative and positive. So the question then was asked of coders, do you feel happy without generative AI or with generative AI? And of course, the number that felt happy using generative AI, right, GitHub Copilot equivalent, significantly more, right? A very small number are unhappy. How about your ability to focus on satisfying and meaningful work? Again, a much lower number of negative people when they're using the AI-based tool. I am in a flow state. I am at one with the keyboard. A very small number of people are not in a positive flow state, right? So these numbers are extraordinary. We dream of these kinds of value-producing outcomes when we create new products. And this is not something that is new. This has been around now for several years, okay? And so I would encourage all of you, if you've not done so— I know in some of the courses I've taught, um, I had mentioned these tools. Some of the students in the room, you know, executives who were running engineering and development teams, weren't aware of them. And within weeks, they started seeing unbelievable upside. And I got a lot of notes back saying, oh my God, thank you so much for, you know, forcing us to think about this for the first time. Okay, now it's also very valuable in a sales and marketing and customer ops context. So let's look at sales, right? We have moved from the early part of my career where we were talking about sort of being champions of information using a variety of sales methodologies like value-based selling. Um, and we're now very much in the AI-assisted sales era, and we're never going to return to the past. And so this is a, an example of AI across the entire end-to-end sales and marketing workflow. And what we see from awareness to engagement to consideration through decision right, retention, loyalty, advocacy, is using a variety of different AI-enabled sales and marketing tools. We're seeing dramatic improvements ranging anywhere from 20% to 40% increases in productivity. Okay, and so there's a, there's an endless number of these applications that are now available But a lot of this work can be used with some of the more common models that are out there. Okay, so if you're not using generative AI-based sales and marketing capabilities, then you're not getting the value out of the sales and marketing team that you should be. Here's one example. This is a new startup that was born at the Founders Institute. The name is Ava, the world's first AI employee, and she's essentially an AI salesperson, right? So you can onboard Ava in about 10 minutes. She's able to get access to 270 million email addresses. She's able to ghostwrite and send thousands of these sales emails at a time. And when you get a response, able to book those meetings directly into your calendar. Now, I started using a very early version of Ava about 7 years ago in one of my startups. And I will tell you, it was pretty effective 7 years ago. So this is something that I think is just such a force multiplier for startups to, you know, and I'm not advocating that you use any one, but I just think they've done a really good job of messaging what the value is of this tool. Okay, Klarna. Some of you may have seen Klarna. Klarna just came out recently. Klarna, of course, many of you may know, was a buy now pay later fintech app. And they're live in 22 markets around the world. They built a chatbot on top of OpenAI for customer support. In the first month, they reported that they literally had set the— that Klarna performed the work of 700 full-time equivalent customer service agents, okay, out of their 3,000. It engaged in 2.3 million customer contacts. In one month. So it handled two-thirds of their customer service interactions, and they got a higher level of accuracy. They saw a 25% reduction in repeat inquiries. They reduced resolution time from 11 to 2 minutes. Wow, that is mind-bending. Okay, and then marketing. Okay, we're seeing more and more This was a study done by BCG across the world. And marketing executives are very, very aggressively using AI, right? Marketing has become a very technically sophisticated function, very data-driven function. So they have been really early adopters in the use of AI. And the number of use cases that are very high impact from lead ID to marketing optimization, to personalized outreach, right? All of these have very significant impact already today. Okay, and I'm going to skip over a couple of these things. Um, and what I want to show you now, some of you may know, uh, Rosie. Rosie is an avatar that was created by a South Korean advertising agency, um, using a combination of computer vision, 3D modeling, and generative AI. And Rosie literally, um, has generated already this year— we're only in April— $2 million in fees because she's appearing in advertisements on behalf of the agency. Okay, and, and they've been able to generate relationships with Calvin Klein, Tiffany, and a number of other very well-known brands. As a result, Rosie has really caught on in South Korea alone. There are 150 of these avatars and the list is growing every day. Right? So for those of you who are focused more in the B2C space, we're starting to see this dramatic increase in virtual social media influencers in Korea. South Korea is leading the pack. Okay. We're seeing the same in the use of augmenting creativity. Um, recent study was done, 1,000+ content creators were, were, um, queried. Bloggers, podcasters, short-form video producers. Two-thirds are using, uh, tools for enhancing, augmenting their creativity. Significant increases in their creativity, their productivity, the number of followers that they're able to generate. Okay. And so it is a force multiplier. I hope I've convinced you of that. And so I wanna wrap up and I wanna just encourage you, right? I'll be teaching a course on artificial intelligence at Babson College in the United States in Boston. Babson has been named the number 1 school in entrepreneurship 31 years in a row, and it was just ranked as the number 10 best college in America by the Wall Street Journal. So we literally just scratched the surface today, but I'll be offering a 3-week virtual course starting in June, offered on Wednesdays. And for those of you who are interested, the QR code is here for you to register. So I'm ready to answer some questions. And I appreciate the opportunity to speak with you today. So, Amelia, um, thanks, Mike.

Thanks. That was really comprehensive and inspiring. Inspiring, and I learned a lot, and you have a lot of great reactions from everyone. So feel free to connect with Mike. You can find him on LinkedIn or scan the QR code. Everyone will receive the recording within 48 hours. And after the Q&A session, just to remind you, we're gonna jump into the networking lounge. Where we can also, you can meet us and maybe Mike, if he has more time.

Yeah, yeah. I'll spend a few minutes in the lounge. Absolutely. Yep.

So we have a lot of questions. So I'm going to go with the ones that were voted the highest. And let's start with this one. I'm going to put it on stage. An ASI, super intelligent AI, unleashed on internet in the coming years. Is that really? And how prepared are we? A question from Cedric.

OK, so I mean, I know that this is a question that everyone loves to ask. And so let me answer it this way, right? If you think about the 3 people today that are considered the fathers of modern AI, these 3 men all won a Turing Award, Turing Award, a few years ago. And the Turing Award is essentially a Nobel Prize in computer science. And if you ask all 3 of those men how close we are to artificial general intelligence, you will get dramatically different answers. And so I don't think anybody could look you in the eye and truthfully give you an answer. Is that day coming? It's coming. Is it 5 years away, as some people tell us? Is it 30 years away, as some people tell us? Nobody really knows. But what I will say is, you know, AI is advancing at a pace that we have never seen technology advance before. Some people are saying it's becoming 1,000 times more powerful every year. Okay? And so, you know, the, the question we have to ask ourselves is, where does AI governance best happen? Does it best happen at a global level? Is that realistic? Right? We look at how ineffective the United Nations has been. We look at how challenging it has been just to get countries within the same region to agree on anything, right? So this is a very, very challenging area. I just gave a TED Talk 2 weeks ago on, on AI at Babson, and here's what I said to everyone, right? None of us individually are going to be able to manage the downside. We have to leave that to experts. We need to focus on the upside, right? We need to make the world a better place. We need to help grow our startups. We should be focusing on how AI can make the world a better place and let policymakers and regulators focus on managing the downside. I think there's nothing else that can be done there.

Thank you, that's a great motivation. And running into the next question: What if you build a tech startup without any AI today? Does that mean you are bound to fail? Does it mean the same proposition with AI will kill your startup overnight? And how do traction and customer service impact on this scenario?

It's a very interesting question, Manoj. Thank you for this. Right. So, so listen, no one says you have to have AI in your company. Okay. That, I mean, I know that everybody loves to chase the new shiny thing. And, you know, and if you think about my, my Y Combinator example, if 71% are building AI startups, that means 29% are not. And they're still getting into Y Combinator. Right? So it's not about do I have AI or do I not? It's if I choose to use AI, am I using it in a way that is truly unique? Because a lot of people say, oh, you know, my differentiation is I have AI. No, that's not, that's not possible. So by no means should you feel like you are compelled to use AI. And there's nothing that says that if you do use AI, it's going to kill your startup. But it's gonna require a lot of really deep thought, a lot of human-centric design and market discovery to use it in a way that actually adds value, right? Now, right now, of course, the funding environment is like a tale of two cities, right? Those people that do not have AI in their business plans are challenged more than people with AI. But, you know, that will level itself out fairly soon. At the end of the day, VCs care more about one thing: are you creating real value? Are your customers buying what you're selling? And are you being able to show that you can scale your business and drive towards some path toward profitability? That will never change. So think most about how am I creating customer value, and the rest will take care of itself. Great question, Manoj.

Okay, um, next The question is also connected. I think Icaro is referring to the beginning of your presentation.

Yes.

Small start should also be applied to our target public. First, provide a solution that's focused in a public, then traction, and then expand.

Yeah, yeah, Icaro, that's exactly right. So when I say start small, right, I mean I want you to have a very big vision, right? I'm— every founder needs to believe that they're going to create something meaningful and impactful. But you can't do that, right? Rome was not built in a day. You need to identify a very clear set of early adopter customers that are perfectly targeted for your value proposition, that can't imagine living in a world where your solution doesn't exist, where they use it, they use it aggressively, they get great value out of it. They're willing to be advocates for you in the marketplace, both with new prospects and with investors. And when you build that core foundation, that solid foundation, then you expand from there, right? You don't try to do it overnight, right? Even Airbnb or Uber, two companies that went global, right? They both started, right? Uber started in one zip code in San Francisco. Airbnb literally started at the South by Southwest conference in a in a single city in Austin, Texas, right? And it was the ability to sort of prove out the model there that gave them the confidence to continue to incrementally expand until they wound up going global. That's just best practices.

Thanks for this answer. Next one is on marketing. So how can brand marketers marketers effectively reach and work with emerging companies to create these points of differentiation? Where are these connections happening? First of all, thank you.

Very interesting company, Emily. A question, and so I assume, Emily, if— are we able to hear Emily? Can I ask her a question? Will she be able to respond to me or not?

She can respond in the chat.

All right, so Emily, do you work for an ad agency or a big consumer brand, or help me understand, I want to try to answer your question as best as I can. What— when you say you're a brand marketer, what, what exactly does that mean?

Let's see if she's still here. Yeah, I'm not sure if she, she's with us now.

All right, well anyway, I'm going to try to answer it quickly since she may not be here. But, but, you know, for those of you who care, right, if you're working for a big brand, right, a lot of big brands have created their own, um, their own partnerships with emerging startups, right? I mean, startups love working with big brands because it expands their reach, it gives them a high level of endorsement value. Um, but, you know, if you're a brand person, you need to figure out where in your city— and New York City is a very active city for startups You need to go out and network, right? You need to go to these startup events, whether they're lunchtime events or evening events or conferences. And, you know, it should be very straightforward, right? These events are broadly marketed and advertised. And let me tell you, the founders would love to work with big known brands. So I think it's a very straightforward thing to do. If you want to do it, it's totally possible. Now, I keep seeing Your LinkedIn is not working. Is that being directed at me? I don't know. If you can't just— if you log on to LinkedIn and you put in the name Mike Grandinetti, I assure you I will come up. Okay, but this is my correct LinkedIn address. If people are struggling with my LinkedIn, I'm not sure who— I've seen a few times your LinkedIn is not working. Okay, great. So someone said it worked. Great. Okay, so Amelia, are you muted now? So I see your lips moving, I don't hear you. Can you hear me? Yeah, you're muted, Amelia.

¡Ah!

Amelia? So I don't know if we're still live or not, guys. I know we've lost Emilia, um, so I just— I don't know whether we will continue here or not. Okay, so if you can hear me, I'm going to answer that last question. I mean, there's no doubt that OpenAI, having raised $13 billion from Microsoft, having a distribution relationship with Microsoft, now having a significant number of enterprise customers, right, about to launch Sora, which will be a revolutionary video-to-image generator, um, the use of their, their DALL-E voice of their image to, uh, their text-to-image tool, right? They are going to be a winner, um, and they are so far out front in terms of their capability that a lot of companies that did not differentiate themselves are not going to make it. And it's exactly the point of the first half of my presentation. If you're not distinguishing yourself from big giants like OpenAI, then you just can't possibly create enough value to continue to exist in the market, which is why Inflection AI closed its doors even though it had billions of dollars in the bank. So, Emilia, I know we've had some tech issues here. I know we're well beyond the time. So how do you want to proceed?

We're right on time, so thank you so much for joining us today. It was an amazing presentation. We're gonna cut the technical issues out and we're gonna make a great video that everybody can find on YouTube, and it's— when we also going to send it to you. So join us at the lounge and connect with Mike, and also don't forget to attend our next event. Mike, do you have something else to say?

No, just I'll make an appearance in the lab and you'll automatically bring me into the lounge then. That's something you'll be able to do, okay. And if you found this interesting, and it looks like a lot of you did, then please take a look at the course. I think you'll have a chance to have a lot of experiential learning. Um, and in the Babson course, and a lot of it will be very much applied to the startup world, which is where we all live. All right, and Amelia, thank you for doing a great job of moderating. I appreciate it.