EP 136: Why so many AI startups will die

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life.

So many AI startups are going to die. They're going to fail, they're going to run on money, they're going to get squashed, they're going to become obsolete. Jarring maybe, but this is a fact. This is the future. And I'll tell you why on today's episode of Everyday AI. Thanks for joining. My name is Jordan Wilson and Everyday AI is for you. It is your daily live stream, podcast and free daily newsletter helping everyday people like you and me keep up with what's going on in the world of generative AI, how we can use it to grow our companies, grow our business just become generally more awesome. So FYI, this show is debuting live, but technically pre recorded. I think it was. Too many people accused me of being an AI robot for working too much, so I had to take a day or two off to prove to you all that I'm actually human. But don't worry, live stream audience, if you're joining me, thank you. I'll still be in the comments, chatting with you, taking your questions, responding to your comments. So let me know what you think. Are AI startups going to die? And also if you're joining us on the podcast, thank you as always. Please check out the show notes. We have so many more resources in there for you. Related episodes, how to subscribe to the free daily newsletter, which you should be doing and more. And hey, just as a reminder, go to your everyday AI dot com. You didn't miss out on today's daily news. Don't worry, we're still going to be putting out the newsletter so you can keep up with what's going on. But today I'm coming with some hot takes. Coming with some hot takes. I hope, I hope you all don't mind because I, I've noticed something over the past couple of years and I wanted to talk about it because it's, it's actually becoming easier and easier to create a startup, which is not a bad thing. But there are two sides to that coin because what's happened is startups that are trying to play in the field of generative AI are growing faster than ever before and it seems like innovation has slowly started to die off because it's become a, a boring game of follow the leader. Everyone's trying to just take cues from chat, GPT from mid journey, from Runway, from Google, Bard from Microsoft and Guess what? You can't always compete with the tech giants. That's one of the reasons why so many AI startups will die. And it's already happening and it's going to continue to happen and it's actually going to ramp up. As we wrap up 2023 and head into 2024, expect to see a lot of this in the news. If you follow startups, even in our daily newsletter, I'm sure there's going to be stories that these AI startups that did not really carve out a niche, that did not build a sustainable, purposeful product, they're going to fall by the wayside just like gnats that are being swatted out of the air. All right, but this isn't just a hot take. I got some facts. So let's, let's, let's talk a little bit about why I am so confident that so many AI startups are going to die. But what do you think? Let me know if you're joining us live. Am I off base here? And here's the thing, I, I feel bad because, hey, a lot of even people in our everyday AI community are creating AI startups. So this isn't a shot at anyone in particular or any company in particular, but I think it's also, it's a sign of reality. But I hope it's also some guidance, right, because we all have to see the writing on the wall. All right, so here's what we're going to go over in today's episode of Everyday AI. We're going to talk a little bit about the influx of AI startups and funding. We're going to dive into a little bit, the lack of USP for startups or unique selling proposition. And then last but not least, we're going to talk about how feature and product updates from big companies are affecting this AI startup landscape. All right, so first, even, even before we get into this very structured overview and if you are joining on the podcast, I'm going to be sharing some things on my screen. But don't worry, I'm going to try my best to go ahead and describe everything to you, but always check out those show notes. Come, come join. Come join. We have a growing, growing community of AI enthusiasts who join us live every day. We have special live events too going on. More of those coming up. This, this, this month for the AI inner circle. But if you're a podcast listener, make sure to come join us. Come join us, meet other AI enthusiast people building great products and hopefully maybe better products. Now, after Today's show. So let's dive into it. So this, this is important, this is important. So I have a graph on my, on my screen here that shows kind of the average generative AI pre money valuation for different startups. All right, we don't need to get into the technicalities of this, but this is according to PitchBook data and it's a graph put together by Fortune, Fortune magazine. Okay? So according to PitchBook data, the median pre money valuation for generative AI companies has risen to $90 million in 2023, up from $42.5 million in 2022 and about $10 million in 2020. What's that mean? Okay, it means a lot of things. Hype. Too much money in the space and not good enough startups. Yeah, I'm going there today. I'm sorry, it's, it's, I'm feeling spicy. I'm feeling like old man, old man Wilson on the front porch shaking his fist at these young startup founders saying, why? Okay, well I'll tell you why. Because the money, right, this isn't just geared towards startup founders maybe who aren't creating a unique or future proof enough product. It's also at the VC industry in private equity. And you'll see here on this chart, right, the fact. Let me just go over the last three and a half years, okay? And when we talk about pre money valuation, that's essentially like, hey, startup that hasn't risen any money, let's look at your company top to bottom and we're gonna slap a value on it, right? And there's obviously startups valued at, you know, a billion dollars, that unicorn status and you know, there's startups video that valued in the generative AI space at many hundreds of millions of dollars. But this is median, right? So this also takes into account all those startups that don't really have any valuation, okay, 10x. 10x, that is in about three short years how much the median valuation of generative AI companies has gone up. I chuckle at the end because it's hilariously telling of what's wrong with this industry, right? There's, there's no other, there's no other space. You know, when you talk about VC funding, so funding from venture capitalists or private equity, there's no other industries that explode 10x over the course of under three years. That's unheard of. So what that means is the floor is built on paper thin material, all right? It's gonna drop, it's gonna drop quickly. Why this, why this giant, this giant Ramp. Well, I'll tell you, and you can probably guess the exploding popularity of open AI, specifically Chat GPT. Right? So there's the biggest, the biggest jump on this chart obviously comes from 2020 to 2021. So guess what happened during that time. That's when the GPT technology was made available to third party companies, right? So what that means is OpenAI's large language model that they had been working on, I believe since 2015 in various iterations. But when it got to GPT3 and it was released to the public via their, you had a host of a ton of companies in late 2020 and early 2021 using the GPT technology to create products, right? And this is again, keep in mind, this is before chat GPT. So as that started to happen, this whole generative AI kind of startup scene started to explode, right? And then you see the biggest jump. The biggest jump, yes, there was the jump from 2020, which was about 10 million dollar average or, sorry, median valuation. So 2020 to 2021, 2022, about the same, right? Around 40 million. Okay, big jump, big jump from 10 to 40. But then it skyrockets. It skyrockets to about 90 million right after that. Obviously, that is because the release of Chat GPT. And as much as I love Chat GPT, and again, I say this all the time, I think Chat GPT with plugins is the most untapped business tool out there, bar none. If you aren't using ChatGPT with plugins, if you aren't exploring plugin packs, which is stacking different plugins, y', all, I've done it live on this show before with different plugin packs that we've built for ourselves and others literally can accomplish what used to take 30, 40, 50 hours with the press of one button. It's extremely powerful. However, investors, startup founders, family offices, private equity, they all saw, they all saw this happen in late 2022. And ChatGPT, literally, it didn't just change the business landscape, it changed the world. And it continues to, right? And now we have large language models popping up everywhere. You know, the big consulting firms are creating their own financial institutions are creating their own. Now we live in this world of generative AI, where I think, and I'm fairly confident it to be true, that generative AI will be exponentially more, more powerful and more impactful than the Internet itself. Check the show notes. I've actually done a whole episode on that backed by data, backed by facts, you know, from, from consulting companies themselves that have done these large Scale studies. So why this huge explosion? Money, money, money, money, right? Everyone saw the popularity of chat GPT and they said, oh, let's build the next thing. And then the generative AI craze started. And like I opened the show with, it's actually ironic that how easy it became to create a simple startup by, as an example, tapping into Open OpenAI's API, right? So you can essentially use their technology, you pay them for it, but you can use their technology, add a couple bells and whistles, throw on an interface, and people call it a product. And that is probably, probably the biggest problem and the biggest reason why so many AI startups will die. Let's keep going. So let's talk a little bit more about this influx of AI startups and this influx of funding. All right? So this year, and this year's not over 2023, so it's going to go up. There's been more than. According to Exploding Topics study, they did more than 10,000 AI startups that were launched globally this year so far. That's too many, y'. All. Don't get me wrong, don't get me wrong. I love these gen AI products. I love them, right? Like I said, we started, Our team started using these late 2020, early 2021, all these different, essentially wrappers, right? Where small startups, SaaS companies tap into a bigger company's API. You know, most of them, many of them, you know, if they're creating GPT tools, are using OpenAI's, some companies are using others, you know, I know, you know, anthropic cloud has, has done this as well. But so many, so many of these AI startups are essentially just using someone else's technology, putting an interface on it, maybe a feature or two here or there, and maybe tailoring it to one certain group, one certain niche, one certain vertical. I have news for you. Startup founders and VCs and private equity people investing in these startups, y' all are going to lose it all, okay? And I'm going to tell you why. But if you did not build an actual company, if you just built a product based around a feature, your days are numbered. And again, not saying that to be mean, I'm saying it hopefully to motivate some of you listening. Or if you, hey, if you have a couple startup founder friends in your life, make sure to send this to them, right? If you're listening on the podcast, you click like two buttons and you can text this. If you're listening on live on LinkedIn, YouTube, whatever, send someone the link, tag them in the comments and let them know you need to be building an actual product that solves problems, not using someone else's technology and adding some new features or a better interface. Right. It's important. All right, let's look at this next stat on the screen here. So this is from Yahoo Finance, but the generative AI market potential is protected, predicted to reach almost a hundred billion dollars by 2026. So everyone sees the money. Everyone sees the money, but they forget the method. All right, Last but not least, AI startups saw a 50 increase in funding in 2023. We already kind of showed that on the first chart. And this is despite the overall decline in tech startup funding. Again, there's another side to this story as well. Everyone's investing in generative AI, which has caused two things to happen. Number one, great tech startups. That said, you know what? We don't need generative AI in our product. They're not getting funded when they should. Okay? Because a lot of these VC firms are just chasing, are just chasing anything with generative AI in it. So that's problem number one. Problem number two, at least when it comes to kind of this angle of startups is you have startups just trying to force force generative AI into their product, their service, their offerings when it doesn't need to be there. All right? We don't need LLMs and gen AI and AI art in every single product we use. We don't. Is it cool? Sure. Do you need it? No. Right. So I think that this is problematic and we're going to see a slow demise over the next couple of quarters here of, you know, these, these VC companies, private equity, family offices, whoever, aren't going to be getting their usual returns that they're used to getting from tech companies, startups, SaaS, companies, whatever, and I'm going to tell you why here in a couple of minutes and it's going to change. We are going to have a whiplash effect of AI gen AI startups popping up every minute. You know, these, these VCs and everyone else throwing more money at them faster than a major league bag, like a major league baseball fastball. That's what's happening. That's what's happening. People are throwing money around. It's like, oh, you're, you're X Google X meta and you're coming together and you're creating a large language model. Take all of our money, we don't need to know anything else. It's, it's literally there are companies out there that have done pretty much that without even a working product, right? It's just, hey, I'm X, this company, X, this company, X, this company. A group of three of us coming together who have worked at big companies. I'm not going to name names, but this literally happened. And they're saying we don't have a product yet, but we have a great idea. We're going to create a large language model gen AI. Buzzword, buzzword, buzzword. Raising tens of millions, hundreds of millions of dollars on a promise, not a product. All right, let's keep going. Why so many startups? So many AI startups will die. Lack of usp, Lack of unique selling proposition. Okay? If you are not uniquely better than the rest, you will be put to the test and you will probably fail the test. Okay, I didn't mean to rhyme there. It just happens, you know, you put a mic in front of me, I might rhyme. But let's look at this high failure rate. So this is from Startup Stash, but showed that nine of ten AI startups will fail within the first five years. All right? Which means many of them will fail way before that. But all these startups that popped up 20, 20, 2021, you're already starting to see them drop because there's too many, there's too many options, too many competitors, right? I've mentioned this already and I'll mention it again, even just if you look at GPT, the GPT technology made available through OpenAI's API, so you can tap into OpenAI's large language model GPT4. You know, I literally, people do this where they do a YouTube video and they build something, live in a couple of hours, right? And you can have a working AI startup in a weekend that is actually good, right? Because of now how easy it is to code and to develop with these AI tools that help you do this, right? But so many of them are going to fail because there's too many of them now. I've tried, I've lost track. I've used more than 70 GPT wrappers. So these different, essentially GPT writing tools, you know, they have different interfaces, different features. I've lost track. It's probably a hundred, but I, I've extensively tried so, so many over the last three years. And you've already seen mass layoffs at some of these companies that were able to raise tens of millions of dollars early on because the competition got stiff. And the biggest reason is coming up. But I'll tell you about that, all right? Another reason for failure, lack of Data, the inability to compete with big tech and not addressing a real world problem. That's the biggest thing. Are you building a product, service or software with a unique selling proposition around a real world problem? Or are you just tapping into buzzwords and putting your own logo on it in your own interface? That's not addressing a real world problem. And then last but probably most important, the overarching problem is that many startups are building on top of existing AI models without adding significant value. All right, and then what happens, which is our, our last point here. But then what happens when you're building on top of those models and those companies then release updates and features themselves? And I think we're going to talk specifically about two big companies because that's what we're going to see. That's what we're going to see. Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on gaming Gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can partner with us too, just like some of the biggest companies in the world do. Go to your everydayai.com partner to get in contact with our team or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on gen. Let's look at one example. So I have, I have shared on my screen now a screenshot that was shared on Twitter. So I had, I had an episode about this earlier this week. So make sure you go check that out on kind of the future of chat GPT. But hopefully this is an example here. So again this was shared on Twitter from the user Brian McAnulti. Hopefully I got your name right, Brian. But when we talk about how even just chat GPT is changing. So there is this new mode that was announced and presumably it will be rolling out to most users here soon, essentially called the All Tools model or mode. And that's when actually let Me just, let me just walk through what, you know, Brian shared in this screenshot. So presumably he got early access, took some screenshots. I know there's been dozens of people that have taken screenshots, shared videos of this new All Tools mode previously. And again, go Listen to my ChatGPT episode I just did about the future of ChatGPT. But previously you had to work in all these different modes to take advantage of the different features and functionality of OpenAI's ChatGPT. Okay, Dall E, the AI image generation was in its own mode. Kind of the vision GPT4V, which a lot of people called that was in the default mode. So if you uploaded a photo or something like that, you would have to go in that mode. And then you had Advanced Data Analysis, formally called code Interpreter, which is essentially advanced computation, uploading, spreadsheets, data visualization, things like that. And then you also had Browse with Bing, which is the ability to query the Internet. Not a big fan of that mode. I'd much rather use plugins, but have an episode on that. Go dive in. So those were different modes, right? So GPT Vision was in default. Then you had Dolly. Okay. Then you had Advanced Data Analysis. Excuse me. And then you had Browse with Bing. That's four different modes. Now those four modes are brought into one. Okay, why does that matter in the context of today's conversation? So like in Brian's example here, he uploaded an image, it looks like of some sushi. And this is all in one mode now because it used to be different modes and he said different Chat GPT to presumably the All Tools mode. Can you generate an image similar to this one but with more. So then the Dolly within this mode created a very nice looking photo by the way of a more robust plate of sushi. Okay, and then Brian in his, according to his screenshot says, thanks. Now, can you list each of the food items in the generated image and place them in a row of a new CSV with the following columns. Food name, calories, protein, fat, carbs. Right? And then Chat GPT does it. So this is an example of the future of Chat GBT and multimodality. Right? So when you can input not just text, but you can input images and voice. Right? This example doesn't show voice, but you know, it's live right now on the ChatGPT app where you can talk to ChatGPT with your voice. So why does this matter and what does it have to do with startups dying and failing? Well, even what we just showed Here, this little example of uploading an image, uploading a PDF, you know, kind of this multimodality. There are already hundreds of startups that over the past 18 months, I'd say, that have made a certain step in this process that I just described their core value. Right. There's countless different startups that all, all they really did is tapped into OpenAI's API and you can upload a PDF. Guess what? Most of those are going to die. Many of them, I'm sure. Or they're going to have to develop an actual usp. Right. And not a feature. Right. Are we getting it now? So like just a couple of these updates from chat GPT from one company made dozens or maybe hundreds of startups. I'm not going to say useless, but not very useful anymore, you know, Are you still going to pay a couple dollars a month for that program that allowed you to upload PDFs and talk with them? I don't think so. I don't, I, I, I, I really don't think so. Why, why would you? Right. Unless they offer some sort of additional functionality. You know, I'm, I'm pretty sure I even have one or two of them that I paid for myself. Don't need those anymore. Unless they solve a real world problem, not just offer a feature or an interface. All right, so let's go into this a little bit more. So this is why so many startups are going to die, because of feature and product updates from big companies. So we just talked about ChatGPT updates as an example, the introduction of PDF analysis, automated tool shifting, which is huge in the all tools mode and combining those different abilities into just that one mode that right there devastated so many gen AI startups. And I'm sure they're already hopefully working, hopefully working on building new features, new functionality if that was their sole usp. Right. And then the big one also very timely because already rolling out this week, Microsoft 365 copilot. Right. I'm going to try to check myself and I'm putting myself on the clock. I'm going to talk about this for one minute. Microsoft 365 copilot is going to change the way the world works. I've done multiple episodes just on this and I don't think people fully realize, yes, there will be a learning curve, yes, it's going to be a slow rollout. It's starting this week with, with enterprise companies who opt in to paying the, I believe $30 per user per month. But essentially it's this if you haven't heard a single thing or if you want to know, okay, why will this Microsoft 365 Copilot kill startups? Well, it's bringing generative AI functionality across the board, across mediums, to the desktop, which has not yet happened. All right, all like all of generative AI right now is so disjointed. It's so disjointed, right? So many even myself, right? I'm using Chat, GPT, I'm using Cloud, I'm using Mid Journey, I'm using Runway, I'm using Pika, I'm using so many Canva, I'm using so many different generative AI tools and they're all over the place. Microsoft 365 Copilot brings generative AI to your desktop, to all of those Office 365 programs, your email, your Excel docs, your word docs, your PowerPoint presentations, your teams meetings. Right, your teams meetings. Again, there's probably at least, at least 100 AI startups or more that are just, you know, hey, we record your meetings and transcribe them for you using AI, which essentially a lot of them are just using Whisper, you know, from OpenAI, kind of tapping into that API. So guess what's going to happen once this, this, this rollout in the learning curve for Microsoft 365 and once it becomes more commercially available to individual users, guess what's going to happen then we're going to see another wave of AI startups, generative AI startups fall. Okay, I didn't want this, y'. All. And let me know, what do you think if you're joining us live. What do you think I want to know? Leave, leave me a comment, leave me a question. Am I off base here? You know, one thing that I really want to emphasize is I don't say these things lightly, right? And if you're listening on the podcast, you, you know, all of you that, that tune in live, you can see my age, you can see the gray hair there, right? I've been getting paid to write for 20 years. I've been designing websites for 25 +. I've working, I've worked with dozens of, of companies, big, small startups, helping them develop digital strategies, teaching them generative AI, you know, so I'm not a, you know, a 20 year old crypto bro trying to push some generative AI or say this or say that. I make these shows as, as you hopefully have heard. Have you? As you've seen, as I share my screen, they are well researched okay, I'm not just throwing out my, my opinions and my feelings. I'm. But I don't want this to be a pessimistic, pessimistic episode. But I want this too. If you're listening to this, number one, if you found value, please share it with someone. Share this podcast, tag them in the comments, shoot the link over to someone that you know that works at a startup or in venture capital, private equity, family office, whatever. Because this industry, this crossover between generative AI startups in the funding, it needs to be re examined. We may not see it yet, but it's broken. We'll see it eventually, right? I think as Chat GPT Enterprise gets rolled out and people get acclimated as Microsoft 365 Copilot gets rolled out, people find the value in that and y', all, we haven't even talked about it. People are pumping hundred million dollars and more into anthropic cloud. Okay, Google, Google. I'm sure once they release their updated large language model Gemini into Google Bard, their chat and once they get it working right. I'm not, I'm not throwing shade at Google. It's just, I don't know, Google Bard's not very helpful for me right now. It doesn't work very well. But think these big companies, they're going to make acquisitions, they're going to be investing in their systems and then you're going to have hundreds or thousands of startups left scrambling. We're still early, right? Consider this the, the warning, the call to action VC peeps be smart about what you invest in, who you invest in. Don't just invest in, you know, a bunch of startups that are just rappers, that are just slapping an interface and a small feature inside OpenAI's chat GPT. You also need to learn how to speak the language. I've seen so many terrible investments that into companies that have already failed. And I called it at the time, I said this is a wrapper. Someone's putting a sticker on this thing, calling it a startup and raising tens of millions of dollars. You need to do more due diligence on the IP. Do they have their own IP? What happens startup founder when OpenAI releases this feature that you're building a product on? What happens startup founder when Google acquires your biggest competitor and they create a conglomerate that is unstoppable? Okay. And then hey, VC company, private equity. I might ruffle some feathers here. Why do you continue to invest tens of millions of dollars into features, not companies we need to collectively figure this out because I love Generative AI. I want these companies to succeed again. I don't. I don't want your startup to fail. AI startup founder. I don't want your investment to go to waste. Venture capital firm but if we aren't smart and look into the future, that is what is going to happen. We are focused on the past. We're only looking at money. When we're building these things, we're saying, oh look what Open AI did. 100 million users in three weeks or whatever it was. Oh, look at the valuation of Mid journey. We need 50 more of these. No we don't. No, we don't build unique products and companies, not features. All right, Old man Jordan's getting off his front porch and reminding you. Please go to your everydayai.com Sign up for the free daily newsletter. We're going to recap today's episode and you didn't miss out on the AI news. Don't worry about it. We got it coming in the newsletter here in like two hours. I'll be in the comments too. Let me know are my hot takes off. Do you agree? Are you going to come back? I hope you come back. Thanks for tuning in and I hope to see you back on another episode of Everyday AI. Thanks y'. All.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit your everydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.