Ep 631: AI’s App Store Moment? Are ChatGPT’s Apps The Next Big Thing or Smoke and Mirrors?

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.

Do you remember using a smartphone without apps? So I don't know if you remember this, but there was a period for more than a year after the original iPhone launched where there weren't even apps. So imagine today using an iPhone without apps and maybe how useless that phone would feel. Maybe in one year or two years or five, we might feel the same way about ChatGPT and say, hey, do you remember when ChatGPT didn't have apps and how little work we were actually able to accomplish? Well, we'll see if that's the case or if this is just another kind of failed experiment from OpenAI. Plugins didn't quite take off. GPTs weren't the next big thing. So what about now that OpenAI has recently unveiled their apps? Is this the true app Store moment or is this just some smoke and mirrors and the latest marketing toy from OpenAI? Well, we're going to be digging in to that on today's Hot Take Tuesday edition of Everyday AI. What's going on y'? All? Welcome to Everyday AI. My name is Jordan Wilson. I'm the host and this is your daily daily live stream, podcast and free daily newsletter helping everyday business leaders like you and me make sense of all these whirlwind of AI updates and how we can leverage all this information to grow our companies and our careers. So if you're constantly looking at chat GPT and saying like what can I do? Well, it's episodes like this that are going to help you do that. So it starts here with the live stream podcast, but if you really want to take it to the next level, you got to go to our website at your Everyday AI dot com. That's your everydayai dot com. This is there each and every day. Well, first go sign up for the free daily newsletter, but each and every day we recap the highlights of the newsletter as well as bring you the no nonsense take on all of the latest AI news. But let's get straight into today's episode. OpenAI at their Dev Day about a week ago unveiled a lot of things and actually I think it was apps that maybe got the least love or the least attention. They unveiled drag and drop agents. They unveiled and released new updates in their API so millions of developers could start adding new functionality. But then they released apps and eventually the apps SDK that will allow millions of developers to bring native experiences inside of ChatGPT. So if you want to know a little bit more, we're actually going to be going hands on. So if you are new here much Mondays we do AI News that matters. Tuesday Hot take. I'm going to give you my hot take on this here in a couple minutes. And then Wednesdays we do more of a at work on Wednesdays putting AI to work. So tomorrow we're going to be doing some hands on demos but today we're going to be talking about how Chad GPT apps work and kind of what they are. We're going to cover the marketing behind it, but also cut through the hype. I'm going to dish my final hot take on if this is an actual big future move or if it's just another feature bound to die eventually. RIP ChatGPT plugins and last but not least, I'm going to give you my hot take and explanation on why I ultimately think ChatGPT apps will be a short term flop, but an absolute long term home run if OpenAI does this one thing that they've failed to do. All right. And if you did miss all of the OpenAI Dev Day announcements, make sure you go listen to episode six twenty six. We cover that all in depth. But ChatGPT apps are here. If you have a free, a free plan, a paid plan, they're there. You may not have even noticed. But don't worry, I'm going to show you everything, how to enable them, how they work and really the fundamental shift that this could represent, emphasis on could, you know, depends on how these are ultimately rolled out, how they're adopted and also the the information and education that ChatGPT continues to pour into this. So here's the bullet points if you missed it, if you've took a nice vacation on a yacht in Greece and maybe missed the last week or so, here's what you need to know. So OpenAI launched ChatGPT apps with integrated services out of the gate like Spotify, Zillow, Canva, Figma and others. Their strategy essentially is to transform ChatGPT into a universal operating system for completing digital tasks. So again, in the same way, like you can run a lot of your day to day business operations on your phone unless you have fat fingers like me and can't type right. But that's kind of the hope and the long term goal with OpenAI is one of their initiatives bringing full working essentially web experiences inside of ChatGPT without ever having to leave. It's built on an open protocol, the model context Protocol from Anthropic for interactive user interfaces and enterprise security developers. Huge opportunity out there. 800 million weekly active users. That's reported by OpenAI. Also, Goldman Sachs separately projected that this is a 15 to $25 billion markets. Also, there's a lot of two sides of this when it first came out, right? A lot of media coverage and people on social media talking about how this is now AI's App Store moment, let alone OpenAI's kind of app store moment. And others are, well, no, this is just a startup killer or this is another piece of marketing from OpenAI that may ultimately flop or maybe may not take off, just like plugins and GPTs. So how, how can you get your hands on it? Who can use it? So right now, whether you're on a free plan, paid plan, doesn't matter, but any individual plan will have access to apps. So it's going to be a little different how you access them depending on what type of plan you're on. But you should have a little plus icon and then an option. Once you do, click that plus icon to go to add sources. Okay. And then from there you'll go click Connect more and then you'll scroll to the bottom in your settings and you should see something that says browse apps. Or you can just go directly into your settings in your chat GPT account. Again, for personal accounts only. Right now look on the left hand side, apps and connectors and then browse down to apps. All right, so you do need to one time individually install and at times authorize these if you are connected to one of the seven apps that launched on launch day that requires you to have an account. Also, in the future there's going to be paid apps, free apps service fees, right? Just like in the full, you know, Apple iOS app store. But right now all the apps from launch are free, but you do have to have an account like as an example, canva. And if you are in the European Union, Switzerland, uk etc, support is not out yet. Also if you're on a business plan, enterprise plan education, the rollout has not fully happened yet. That will be for later this year alongside once OpenAI does unveil the developer submission process. So yeah, right now there's no mom is the word on monetization. We still don't even fully have the process for developers to submit an app for approval, but we should be getting word on that soon. So let's talk about a couple examples and how these work now. So one of the biggest differences and one of the newest pieces of utility that come along with apps. And I'm going to go over later kind of the difference between apps and connectors and GPTs, but I want to focus a little bit on apps. So the biggest thing is it essentially brings in a graphical user interface or GUI of these seven websites. And you might say, okay, well why, why wouldn't I just go to Spotify? Or why wouldn't I just go to Canva? We're going to talk about some of the stats and some of the results, but I think internally OpenAI saw a 70% increase in productivity doing things within ChatGPT versus, you know, doing things separately on these other apps. And when you think about it, it actually makes a ton of sense. And a lot of this goes to context, right? There's been all this buzz recently about the terminology context engineering, right? Which, not trying to brag, I've technically been teaching context engineering before anyone was talking about it, but that's another story for another day, right? But essentially what this means is this allows all of the apps within ChatGPT to understand and work with your context within the chat window that you're using. You don't have to redescribe anything. So think, let's say Canva, because it's a great example and probably one of the use cases that I'll tackle on tomorrow's AI at Work Wednesday show. But let's just say that you're using ChatGPT to pre apps, you're using ChatGPT to build a Canva outline, right? Ultimately what you still have to do is you have to go copy and paste all that into canva. You have to find a template that works, right? You have to resize all the fonts, bring in your colors, all those things. Right now, well, you can just, if you're doing a deep research, if you're, you know, bringing in information from connectors, you know, bringing in your own data, working in your context of a chat, you can just then call on the canva app and just be like, hey buddy, turn this into a 10 page deck for me. And within a minute or two it's there. You don't have to copy and paste, you don't have to switch between apps. So that's ultimately what ChatGPT or OpenAI is trying to do. They're actually trying to make this an AI operating system and bring all of your day to day business processes inside of these apps. Pretty big deal, right? And we're going to talk later about, you know, some of these new features that OpenAI has been rolling out that just happen to make a lot more sense now, right? Like a checkout page inside OpenAI. But a couple. Or, sorry, inside of ChatGPT. A couple use case examples, though. So let's say. And some of these are based off of OpenAI's demo, some are. So, you know, let's say you're that inside ChatGPT, you've been discussing moving to Pittsburgh, right? And then something about a dog walk. And then ChatGPT remembers the context and then it'll surface results. If you're using the Zillow app and you could say, and it might suggest something next to a dog park. Right. The example that I just kind of gave, if you're working within ChatGPT, maybe you're working in Canvas mode and you're iterating in body, by the way, Canvas mode, if you haven't used it so good. Right. But maybe you're using canvas mode in iterating with ChatGPT to get kind of that outline just right. Well, now you can just. You don't have to do anything. No copy paste. You just call on the canva app, so you can use the at symbol, bring up canva app once you have it installed and just say, turn this into deck, right? And then you can iterate on it conversationally with natural language. That's another huge thing. Another example, one that I tested out and I really like. You can with a Coursera app also, by the way, I know a lot of people are asking, like, hey, Jordan, like, your Coursera course, it's so old, don't take it. It's like three years old. But if you're watching a Coursera course that's not three years old and outdated, it's great because you can actually pause and just say, explain this concept to me. And it just knows, it knows what you're watching. It understands the context of the video. The other thing that I've been testing out, which is really cool, calling on the Coursera app after you've already been having a conversation inside ChatGPT. So it has this kind of preloaded context before you even start watching a video. So then you don't. If you're in the middle of it, you don't have to pause and type out a big paragraph or dictate a big paragraph, because it has the context of whatever you're working on. But that's great. Just say, saying, explain this concept deeper without ever having to leave the player. So why? What's the point? Well, this is OpenAI's technically third attempt at creating essentially an ecosystem that allows you to connect to third parties. The first one, ChatGPT plugins, they came and they went. By the way, I loved plugins. We, we taught something called plug and packs. I literally don't think anyone at OpenAI ever realize what they had because you could use three plugins at a time. This was technically the first agentic features of ChatGPT. Before that it had a reasoning model because you could share the context between three plugins at once and it would pass the context on. Anyways, plugins came and went, right? And then In I believe November 2023, OpenAI made a big splash with the GPT Store. And at the time a lot of the headlines were saying the same thing. Is this the App Store moment? Right. And OpenAI made a lot of lofty promises, but what happened? Well, OpenAI pretty much ignored GPTs. It went, Geez, a year and a half without any major updates, right. Finally it was about a half year ago that they finally allowed GPTs to use, you know, newer models because previously, you know, you could only use the much older GPT4O and finally before GPT5 came out, they said okay, you can use the O3 models. So this is technically their third iteration and I mean we'll see what ultimately happens with GPTs because they have the GPT Store. And the good thing is. Well, the good thing about GPTs is also the bad thing is it's so easy to make a GPT. You can make a GPT in literally about 20 seconds and it can be okay. But the problem is the GPT store got flooded with garbage because I'm not a mathematician, but 99.9999% of everything in the GPT store is hot garbage. Very little utility because the bar for entry is so little. Conversely, plugins, most plugins were of decent quality because it was difficult to get a plugin approved and you had to have a little bit of technical, some technical chops. So we'll see ultimately where apps lands and what the Future is for GPTs. But the goal is for OpenAI to move beyond and to use apps to move beyond just ChatGPT being an AI chatbot and they're trying to become the operating system in the AI era. So some of the apps available today and reported some of these stats so far based on those app usage from pre launched testing. So Canva reportedly delivers a 400% faster design creation. Zillow processes 2 million queries daily with 500% conversion improvement using the ChatGPT app Coursera serves 300 right now. Not in the app, but Coursera serves 300,000 students weekly with a 60% higher course completion rate than traditional methods. Also Spotify, Figma, Expedia, booking.com complete kind of the launch of the initial seven apps, there's also 11 more on the way. We don't have an exact launch date for the rest of these apps, but we do know that DoorDash, Instacart, Uber, OpenTable, the Fork, Target and Peloton are on their way as is all trails in TripAdvisor. Side note, I'm tired of like all these AI companies when they're demoing something like, you know, oh, watch this agent, you know, buy your tickets for the concert or book your travel, y', all, can we just start doing business use cases when we launch these products? Anyways, we don't know yet how or when third party developers will be able to submit their apps, but OpenAI did release first set of guidelines, essentially what developers have to do in order to submit those. But we don't know how or when they will be approved. We did just hear that they'll be releasing that later this year. So right now, here's how they work, right? If you want to know what's going on under the hood. Well, they're built on the MCP standard which enables developers to define both logic and the interface. The first use, like I talked about, the first time you use one of these apps, you do have to connect and consent to it first, right? Go over the privacy policy, all that good stuff, sign away your firstborn. Not actually, that's just a joke. And also the big difference, apps render an interactive user interface in the chat. So it does feel kind of like you're using, I mean different apps. I've used I think five of them so far. Different apps have a different level of like reflecting the web interface, right? I think Zillow, pretty good Spotify, it's just more of like a, you know, a very slimmed down version of what you might get in Spotify. So if you're like me, you're probably thinking, wait, didn't OpenAI just release these connectors a couple of months ago? And what about GPTs? What's the difference? Because even Canva is a great example. Canva has a connector, Canva has a GPT and Canva has an app. Well, by this time Canva's GPT is pretty much useless now, although previously it was kind of valuable. So Canva's GPT previously you could give it Some information and it would kind of create a design for you, but not in the interface. You had to click to see it and it would open up a new window. It sometimes worked, but mainly didn't. Right. So Canvas GPT specifically, great in theory, but that's not really what GPTs are for. So let me just kind of go over some of the differences. Well, actually, no, let me finish this Canva piece. So Canvas app, great. I think it's one of the best, shiniest use cases of the seven that they announced that are available. And then Canvas connector, conversely, something I use all the time. I love doing deep research across my canva. I know that's weird, but I've done 630 podcasts and each and every one of them has a little Canva deck that I show on the live stream just with some notes and some bullet points. Right. So I use the connector I can ask deep questions across. So 600, some of them are 20 pages. So yeah, thousand plus pages of Canva docs. So the connector can just grab in all this information and it's almost like a database. And then the app, obviously you can make and iterate and see and view designs in the app. And we'll go over that one tomorrow. I think that'll be one of our three use cases. So essentially the apps embed a rich interactive user interface that you can interface with directly in the chat. So you can interact with actual website graphical interfaces and work with your context. Some, I mean, we'll see. But like, I don't know, like the booking ones, you know, the booking.com one, it's like, doesn't really do too much for me. You know, it shows you examples and then you click, oh, like this hotel in Chicago looks great. And then it clicks and it just opens up a new tab. So some have great front to end user experiences, some not so much. But they all do have an option to essentially hand off or continue in the actual website. So connectors, like I said, these just securely bring in your personal data into conversations for personalized answers. So connectors are kind of like a mini miniature version of retrieval, augmented generation and then custom GPTs. These are just kind of personalized versions of ChatGPT that you can modify with text. Right. So that's kind of the difference. But let's talk about the potential. All right. Because with the apps we'll talk to our developers here for a second. OpenAI did also release an apps SDK. So they did release the platform for developers to build on top of. And right now they did say that over 50,000 developers already registered within the first few weeks. So later this year submissions should be open. Right? So if you are a larger organization, here's what you should probably be thinking of. You might want to essentially build an app, maybe of your internal database, maybe of your website. If you have a, a comprehensive website. Let's say that you're an E commerce brand. Let's say that you're a brand that puts out a lot of, a company that puts out a lot of information online. You might want to convert this into an app to bring both for your customers and maybe for your own internal employees. I mean, we'll see some of the use cases, but submissions, we should know more later in this year about how that works as well as how they will be displayed within ChatGPT for app discovery. It does seem it's going to be similar to the GPT Store where there will be categories. I mean, we'll see if organizations, once this does roll out to the business tier, previously called teams, the enterprise tier, all that, if there will be apps specifically for just enterprises. So I would think they will since they have that for GPTs. And I can see a lot of big use cases for that, especially for organizations that have technical departments. But there's especially, especially for companies that use some sort of proprietary piece of software that's specific for a, their, their niche or vertical. I mean, to be able to work with that context inside ChatGPT with natural language is huge. This could be the future, right? This could be the future, right? Think of a CRM, right? And some CRMs are super clunky. I remember the early days, like using pipedrive, like very early. I would have to like read all these long help docs just to figure out like how to make, you know, an update to an opportunity or something. But imagine sometimes, even when you're working with today's apps, all the duct tapes, all the duct tape that you have to put together to make something work the way you want it, you know, customizations, a lot of companies pay extra or have developers build on top of their current software. Well, imagine if it just worked with your natural language inside ChatGPT, those pieces of software that you use. I do assume probably any huge enterprise piece of software, your email marketing tools, your CRMs, they're going to have to go inside ChatGPT apps. I don't think they're going to have a choice because what's going to happen, I think some of the more AI native organizations are going to do that, right? So let me take as an example. If you sign up for our email newsletter, we use Beehive, right? Great. Previously we used mailchimp. I would assume a company like Beehive would probably be early to adopt something like this, and they're growing and they're making everyone else in the industry sweat. I would assume even the legacy players are eventually going to have to bring their apps in. It's the same thing with the like the iOS App Store. There's no big enterprise piece of software out there that doesn't have an iOS app, right? You, you, you can't not. So I do think eventually if OpenAI does this the right way and I'll give my 2 cents hot take on this at the end, everything's going to have to move in there. And I do think ultimately we're going to see anthropic Google and Microsoft follow suit. 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 Genai. 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 everyday AI.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. So here's been some of the good vibes so far, right? Venture Beat, you know, just kind of going over some of the media coverage. VentureBeat called this the new App Store with a vibrant interactive platform. Forbes said that this is the this actually makes ChatGPT the next big browser as a primary interface to the Internet. TechCrunch praised the integrations feeling native to the medium with 800 million users. Obviously inside ChatGPT some of the downside so artificial ignorance argued that Agent Kit so something else that they launched on Dev day is probably a startup killer. And that one's interesting because OpenAI releasing a lot of things that are viewed as startup killers, even though a lot of their developers and some of their big account holders are developers. Anyways, Bloomberg describes some of the early apps as underwhelming and sometimes frustrating and I agree with that. Some problems with responsiveness and functionality issues. Yeah, it's buggy. Some of the apps work well, some not so much. And then Wired in their article did remind readers that OpenAI previously attempted this with GPTs and that did not achieve widespread success two years ago. So here's my final hot take. You're not going to hear hardly anything about apps until you do. This is one of those instances with OpenAI. Maybe it's because they wanted to unveil this during dev day, right? I think just releasing what they released in the API. They released GPT5, Pro Access, Sora 2, a new real time model, and then they released Agent Kit, drag and drop agents. Right. So I don't know why rush. It seems rushed. Why rush ChatGPT apps with only 7 apps now and 11 apps later and developers not even necessarily knowing how to submit their apps or when? Right. It's almost like you only get one shot of the grand opening. So I don't know, maybe there's. I'm sure there's a rationale behind this. I'm sure they used, I don't know, GPT 7 to plan out their go to market. So I don't necessarily understand this go to market strategy because I think, you know, yeah, there's all this initial coverage and people are talking about it and then people are just going to forget about it because I don't know, unless you're a designer or, you know, enjoy taking a couple extra steps when, you know, trying to book something, you know, book a hotel on booking.com or whatever, most people aren't going to use this. So I do think in the short term it's going to be a massive flop. Right, because no one's going to care. That's the reality. Long run, it, I think is going to be a massive hit if one thing happens. And here's why, like I was describing, if this catches on, you can't not release a chat GBT app. Right? Mailchimp is gonna have to, right? Big CRMs. They're gonna have to. Essentially, if you want to compete in the new business landscape, which is AI operating systems, and I've been super bullish on this before anyone was saying the words, I've been saying it, people are gonna have to bring their services to where the Demand is and OpenAI owns Demand 800. According to OpenAI, 800 million weekly users. So that obliterates anyone's, any company's traditional app distribution. No one else. I, I won't say no one else out there, unless you're Google or so, right? Or maybe Salesforce, but no one has 800 million users, okay? This is, I think, the potentially the distribution platform of the future. If they get this one thing right, Open AI with the App Store, they need to be minting millionaires as soon as possible. And here's why. They fumbled the GPT store. GPTs are insanely useful and they still are to this day, right? To almost two years later, OpenAI largely ignored GPTs, right? And kind of continues to ignore GPTs. And I think the power is immense and I really hope they don't get rid of them because GPTs, like I said, for small teams that just can't go out and build on the, you know, apps, SDK, GPTs I think have become essential almost for a lot of companies, day to day business, business operations, right? So I hope OpenAI continues to support and maybe promote and improve GPTs. But one thing they did not do that, they kind of promised, they talked about monetization of GPTs, right? They're like, hey, developers, go build your GPTs and put them on the GPT Store. You know, and you know, for our most used GPTs, we're, you know, we're going to share the, share the wealth, right? Do you know how many stories I heard of people getting money from that? 0. 0. Okay. If you look at the iOS app store in 2024, 5,000 apps earned more than a million dollars. 5,000, okay, I'm not saying OpenAI needs to make 5,000 millionaires in 2026, but they need to be handing out money to developers, all right? Because if you want to incentivize everyone to get on board, you need to put your money where your mouth is. And OpenAI, in my opinion, maybe, I don't know, maybe they paid a couple developers for the GPTs that they put out on the GPT Store. But I remember seeing stories, and we shared about it in our newsletter of some of the most prominent developers that had, you know, GPTs that were, you know, top on the GPT store were like, I haven't seen a penny. So OpenAI, you have to pay people and you have to make them insanely rich. Talk about a great PR strategy. Because let's be honest, you fumbled this rollout. OpenAI should have waited and had at least a concrete Timeline. They should have said, hey, October 6th, right? This is Deb Day. We're announcing it. You know, November 1st, you know, you're going to be able to submit your apps. The, you know, App Store is going to launch November 19th, my birthday, right? So have a plan, right? Here's, here's how monetization is going to work. You know, we're going to make money on, you know, fees and paid apps and these things. And, you know, developers are going to get, you know, 30% right. You don't get a second chance to roll this thing out. So you already messed up once. So OpenAI, if you want this to work, don't mess this up again. You need to make so many people insanely rich, very quick, instant overnight millionaires. Because what that does is it creates this cyclical PR buzz, right? Because now all of a sudden you're going to start seeing these things on the news. You're going to start seeing, right, these stories pick up everywhere and then that brings even more and more users. So you're welcome. OpenAI do that. And this won't fail because let's, let's face it, plugins obviously failed. GPTs for what they currently can do, I think have largely failed because they haven't given them the attention and promotion they deserve. So OpenAI, third time can be the charm with the App Store. This can be your App Store moment if you do it right. So far you haven't done it, but it's not too late. There's my hot take, y'. All. I hope this one was helpful. So if you want, make sure you tune in tomorrow. We're gonna be doing a more practical hands on walkthrough. So on Wednesdays we do AI at work on Wednesdays and we're gonna be showing you, I think, three of the most practical use cases so far. I'm gonna be walking through those with you live. So it is a visual episode. So if you normally listen on Spotify, Apple podcasts, thanks for the support, appreciate that. But in the show notes, we always leave a link back to our website where you can go watch the video version. So make sure you tune in tomorrow for that one. Go to your everydayai.com Sign up for the free daily newsletter. Thank you for tuning in today. Hope to see you back tomorrow for that one and every day for more 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 youreverydayai.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.