WEBVTT 00:00:00.000 --> 00:00:03.280 This week, Anthropic released Opus 4.8, 00:00:03.280 --> 00:00:23.225 which they say is the most advanced AI model in the world. However, others are saying we've entered the iPhone era of AI models where you can't even tell the difference between each model upgrade. We're gonna discuss this today. We're also gonna talk about codex. This week, OpenAI released some insane updates to their super app, codex, 00:00:23.465 --> 00:00:26.585 and some of the updates they didn't even publicly announce. 00:00:27.080 --> 00:00:43.425 You're watching AI native where we cover the most important news and updates on the best AI agent platforms and models. My name is Riley Brown. Let's not waste any more time. Let's dive in. So here we are. This was the Thursday announcement by Anthropic 00:00:43.505 --> 00:00:46.065 introducing Claude Opus 4.8. 00:00:46.225 --> 00:00:51.585 It builds on Opus 4.7 with a sharper judgment, more honesty about its own progress, 00:00:51.745 --> 00:00:56.390 and the ability to work independently for longer than its predecessors. 00:00:56.550 --> 00:01:05.670 And here is the model card. So on the model card, Opus is apparently better at coding. This agentic coding SWE bench pro. 00:01:05.990 --> 00:01:15.305 It's not as good as GPT 5.5 at terminal coding, but it's better than all of the other models including Opus four point seven and five point five 00:01:15.465 --> 00:01:16.825 at reasoning, 00:01:16.905 --> 00:01:21.225 controlling your computer, doing knowledge work like doc sheets and presentations, 00:01:21.465 --> 00:01:28.340 and other finance tasks. And guys, it was genuinely my plan to make a full video on Opus 4.8, 00:01:28.340 --> 00:01:32.580 but I spent three hours comparing the difference between Opus 4.8 00:01:32.580 --> 00:01:34.020 and Opus 4.7, 00:01:34.020 --> 00:01:40.145 their previous model that they released. And guess what? I literally couldn't tell the difference between the two models. 00:01:40.625 --> 00:01:43.825 And I'm not the only one who thinks this. Uh, Greg Eisenberg, 00:01:43.825 --> 00:01:52.520 friend of the show, he said, I didn't cover Claude Opus 4.8 on my pod because I don't think it's meaningfully better than GPT 5.5. 00:01:52.600 --> 00:01:55.960 And I'll add that it's not meaningfully better than 4.7 either. 00:01:56.280 --> 00:02:01.080 And he goes, we are entering the era where model releases start to feel like iPhone releases. 00:02:01.505 --> 00:02:08.065 Remember when every new iPhone had a genuine leap? Now it's a slightly better camera and you can't really tell the difference. 00:02:08.305 --> 00:02:10.945 That's where models are heading. 4.6 00:02:10.945 --> 00:02:13.185 to 4.7 to 4.8. 00:02:13.265 --> 00:02:25.990 Each one is slightly different, but you can't really tell which one is best. In fact, I'll tell you from personal experience, I'm still running AI agents in iMessage, running an AI agent very similar to OpenClaw, and I'm using Opus 4.6. 00:02:26.230 --> 00:02:44.900 I think it is the best for general agent work at least based on how I use it. And I literally can't tell the difference between these three models. And it wasn't just Greg. Here's Matt Wolf agreeing with him. He said, so much this. I spent over one minute talking about OPUS 4.8 in my recent news breakdown and there really wasn't much to say honestly. 00:02:45.140 --> 00:02:55.460 And when there's a big update, Matt will spend five, sometimes ten minutes talking about a huge update and he only talked about it for one minute. And now we're gonna compare GPT 5.5 00:02:55.460 --> 00:02:57.115 to Opus 4.8. 00:02:57.115 --> 00:02:58.475 And DeepSwee, 00:02:58.475 --> 00:03:05.755 which is a company that measures frontier coding agents on original long horizon software engineering tasks, 00:03:06.155 --> 00:03:15.410 they posted some data that was really interesting. And so DeepSwee looks at three things. Right? They look at cost, time, and output tokens, and then additionally, 00:03:15.570 --> 00:03:17.970 they plot it against their score. 00:03:18.210 --> 00:03:23.250 So you can see here, these are the GPT models right here, and here's the OPIS models 00:03:23.330 --> 00:03:29.885 right here. And so the higher up you are on this chart, the better your score. OpenAI got a better score. 00:03:30.125 --> 00:03:33.565 And you notice here that the cost goes this way. 00:03:33.805 --> 00:03:40.680 So the further this direction you are, the more expensive your model is. So GPT 5.5 00:03:40.680 --> 00:03:41.560 medium 00:03:41.560 --> 00:03:47.800 high and extra high are scoring higher for less cost than Anthropix 00:03:47.800 --> 00:03:49.160 Opus 4.8. 00:03:49.160 --> 00:03:54.855 The OpenAI is getting a better score for a lower cost. Here we can see they're getting a better score 00:03:55.095 --> 00:04:04.775 at a lower amount of tokens per task which is better and we're also seeing that the average cost per task is just lower. 00:04:04.775 --> 00:04:08.455 Right? If you see that this model is clearly the most efficient, 00:04:09.090 --> 00:04:13.410 it takes less time and it gets a higher score. 00:04:13.410 --> 00:04:29.275 And also as of late, I've also noticed a lot of people talking about trust and depth of tasks. This guy said I can trust GPT 5.5 with things I would never trust Opus 4.8 to handle. Yeah. Opus 4.8 feels good and can be quite addictive to use especially when vibing, 00:04:29.435 --> 00:04:36.090 but that's mostly surface level. I'll also add that the Opus models in general are better at design. They're better at presentations. 00:04:36.090 --> 00:04:51.595 You're gonna get a better slide deck, a better landing page. It looks more appealing. They put a lot of effort into claw design. However, when you wanna do really long agentic tasks, if you wanna do deep coding work or have it control your computer and even control your text messaging directly 00:04:51.755 --> 00:04:53.035 from the app, 00:04:53.195 --> 00:04:55.995 I highly recommend using GPT 5.5. 00:04:55.995 --> 00:04:59.275 And so now I divide these large labs announcements 00:04:59.275 --> 00:05:11.940 into two categories. Right? There's model updates and then there's super app updates. And like nine months ago, I was way more excited for model updates because every single model update felt like a big step change 00:05:12.340 --> 00:05:25.325 and everything was done in the terminal. So there wasn't really that much innovation happening at the app level or you know the app where you use these AI agent tools. And if you've been watching my content for the last four months, I've been obsessed 00:05:25.405 --> 00:05:31.725 with the super app and the super app like Claude desktop or Codex are these apps where you can 00:05:32.045 --> 00:05:38.650 very easily talk to AI agents. Right? You can speak to AI agents where you have your tasks on the left panel. 00:05:38.890 --> 00:05:58.415 You have your agent and then whatever your agent is working on. And there's so much innovation that needs to be done to make this a very seamless process so that you can interact with agents for all of your work. And this is exactly what OpenAI did this week. They announced a bunch of different things for their codex application, 00:05:58.975 --> 00:06:09.060 new updates to their platform. And so the first update that they announced is there is now Windows computer use. So if you go to the Codex app on Windows, 00:06:09.460 --> 00:06:12.180 you can now officially type at, 00:06:12.660 --> 00:06:15.300 uh, computer use and you can have 00:06:15.555 --> 00:06:16.915 GPT 5.5 00:06:16.915 --> 00:06:22.115 inside Codex. It can control your computer fully. You can say control Canva 00:06:22.115 --> 00:06:23.155 to do 00:06:23.235 --> 00:06:24.035 task 00:06:24.195 --> 00:06:52.665 and you can do this on Windows now. Another one for those Windows users out there, they now have Windows Codex remote. Inside Codex, if you go down to this phone icon right here, this will give you a QR code. If you have ChatGPT downloaded on your phone, you can now type prompts directly through ChatGPT and it will control codecs which can control your computer. If you have an iPhone and a Windows computer, you can connect ChatGPT. 00:06:52.665 --> 00:06:57.930 Right? This is just the ChatGPT app and I'm going to the codec section and now I can press chat 00:06:58.170 --> 00:07:10.170 and now I can message codecs and I can even use computer use inside the iPhone app and I can say please, uh, check my, uh, desktop 00:07:10.170 --> 00:07:20.585 and tell me what's there. And you can see here, right, it's showing up right here. You can do this on Mac and now you can even do this on Windows and these are perfectly synced. 00:07:20.905 --> 00:07:31.060 You can see here this is the same exact chat thread and it shows up on the desktop app and the phone. You can literally control codecs from your phone, Mac or Windows. 00:07:31.380 --> 00:07:37.540 And since codecs can control your computer, you can basically control your computer through ChatGPT, 00:07:37.700 --> 00:07:50.855 which is a really, really underrated and cool feature. Feature. Okay. The second set of updates that OpenAI released for Codex, this is the one that I'm gonna use the most and I think it is just the most useful. 00:07:51.015 --> 00:07:54.535 And so when you're inside Codex when you're inside Codex, 00:07:54.855 --> 00:07:56.375 you can open up a browser. 00:07:56.800 --> 00:07:59.680 Now as of two days ago, 00:08:00.160 --> 00:08:03.520 these stay signed in. So I can go to twitter.com 00:08:03.600 --> 00:08:09.040 and you notice I'm already signed into my profile. I don't know why these tweets aren't loading. There we go. 00:08:09.520 --> 00:08:18.885 This is my Twitter feed and I'm automatically signed in. Can also say something like please get my, uh, latest 00:08:19.365 --> 00:08:20.965 video agent 00:08:21.125 --> 00:08:22.085 native 00:08:22.085 --> 00:08:23.925 to link on 00:08:24.405 --> 00:08:24.805 Notion. 00:08:25.670 --> 00:08:26.790 Summarize 00:08:26.950 --> 00:08:30.390 it, and give me a link here. 00:08:30.630 --> 00:08:33.030 So since iCodecs 00:08:33.030 --> 00:08:44.315 is set up to connect to Notion through the Notion plugin, it can find the exact video I'm talking about. It's gonna give me a link to that video then I can just open it directly inside the Codex browser. 00:08:44.475 --> 00:08:54.770 This is becoming a full browser and take a look at that. So it responded. It thought for two minutes it found the Notion document that I'm working on which is for this video. 00:08:55.250 --> 00:09:02.930 And all I need to do is right click on this and click open in browser and take a look at this. We are automatically 00:09:02.930 --> 00:09:07.785 signed in to Notion. Close the sidebar and here's the app open 00:09:07.945 --> 00:09:08.825 inside 00:09:09.065 --> 00:09:20.920 Codex and I'm signed into Notion so any document that it creates inside Notion for me I can just open it up. So now I'm using Codex. I can ask Codex to change anything inside Notion 00:09:21.160 --> 00:09:31.960 and it will edit the page and I can see it live. I can add things to it just like I'm using Notion except I don't need to leave the AI powered super app which is Codex. 00:09:32.305 --> 00:09:33.425 Now I 00:09:33.425 --> 00:09:41.745 do anticipate that Claude code or the Claude desktop app will have this feature. It just feels like they're really far behind and they're not prioritizing it. 00:09:41.985 --> 00:09:56.150 This right here is something that I've been using every single hour for the past seventy two hours since they released this feature. Now that you stay signed in, it's really useful because before you actually had to sign in every time you opened up a web browser. 00:09:56.630 --> 00:10:13.955 And another thing that I realized, you can open up many browser tabs. You can't hit plus and open a browser tab, but if you're in your browser and you press command open, look at this. It's opening all of these as new browser tabs. So I can go from the main tab to this tab 00:10:14.355 --> 00:10:15.635 to this tab 00:10:15.900 --> 00:10:21.180 to this tab. And so we're starting to see this become a full browser that you can use 00:10:21.500 --> 00:10:23.580 next to your AI agent. 00:10:23.820 --> 00:10:38.655 So that is two, which is browser tabs stay signed in when you're using the browser inside Codex. And we also have multiple browser tabs per task and we're starting to see it become as if you had Google Chrome inside Codex. 00:10:38.735 --> 00:10:42.575 And so this third one is a lot of people's favorites. 00:10:42.815 --> 00:10:44.760 So now when you use use Codex, 00:10:44.920 --> 00:10:51.080 agents can spin up other agents. On top of this agents, you can ask Codex 00:10:51.240 --> 00:10:52.040 about 00:10:52.280 --> 00:10:53.800 any chat 00:10:53.960 --> 00:10:55.720 you have open. 00:10:55.800 --> 00:10:58.555 Let me show you how this works. So if we go to Codex, 00:10:58.555 --> 00:11:01.435 I can now type something like this directly inside Codex. 00:11:01.515 --> 00:11:14.270 So and I call this a super prompt. I want you to spin up new chat sessions inside Codex. So like right now I'm about to fire off a chat session and this chat session will actually create six more chat sessions. 00:11:14.350 --> 00:11:16.590 So check this out. So I'm gonna run this. 00:11:17.310 --> 00:11:26.830 And so now you can see here it says all set this up as six separate codex threads with concrete task prompts. So it's basically going to write prompts 00:11:27.155 --> 00:11:32.595 in new chat sessions and then they'll show up right here. Okay. So it's activating 00:11:32.595 --> 00:11:47.140 some memory. It's it's basically trying to figure out how it wants to prompt the agent and so it says I'm creating six background threads now each with narrow brief and completion criteria. And here it goes. It's created one, two, three, 00:11:47.860 --> 00:11:48.660 four, 00:11:49.620 --> 00:11:50.420 five, 00:11:50.740 --> 00:11:54.820 and six. We're gonna see AI rename them. Watch this. So triage, 00:11:55.825 --> 00:11:56.625 boom, 00:11:56.625 --> 00:11:58.145 boom, and boom. 00:11:58.305 --> 00:12:01.185 So AI created these new chats 00:12:01.265 --> 00:12:13.340 and you can see here the AI basically prompt this. It's sent by Codex from another thread. That's how you know Codex prompted it which is really cool. So you can ask Codex to create 00:12:13.580 --> 00:12:23.180 new threads. So you can start up 10 threads directly inside Codex and here they are all going to work. And so that's really cool and I haven't even fully discovered 00:12:23.180 --> 00:12:47.890 all of the use cases that I wanna use for this. Maybe I might do a full video on that specific feature about using one master agent to spin up sub agents and then you can create an automation which checks in on how those other agent chats went. I think there's a lot of exploration to do there, but that's out of the scope for this video. I do wanna cover some a little other updates that they announced which is there's now better search. So if we go to codex, 00:12:48.130 --> 00:12:49.890 if we go to codex, 00:12:50.295 --> 00:12:53.175 and now if you press command g, 00:12:53.255 --> 00:12:55.335 I believe, I can now search 00:12:55.575 --> 00:13:00.375 way better. Right? You can press command g and I can search for a key term like OpenAI 00:13:00.375 --> 00:13:02.375 and everywhere OpenAI 00:13:02.375 --> 00:13:03.415 is mentioned, 00:13:03.495 --> 00:13:13.290 I can now search not just through the titles but through all of the chats in general. Right? So it's much easier to search through all the chats. Let's see where I mentioned, 00:13:13.530 --> 00:13:33.540 uh, command g, where I mentioned Chorus. These are all of the scripts or all of the chat sessions where I mentioned Chorus. It makes it a lot easier to search through all of the agent chats that I create. Another small thing that they announced was this new GitHub activity page. So, again, if we go to codecs and you go to settings, 00:13:33.860 --> 00:13:34.740 profile, 00:13:35.140 --> 00:14:10.190 here we can see all of the days where I use Codex. I basically started using Codex forty three days ago. I've been using it every day since forty three day streak. My longest task with three hours and seven minutes, and I've used 4,000,000,000 tokens. Pretty fun new update to the app. Okay. So now I wanna move to another trend that I've noticed. A lot of people have been DMing me about their vibe coding platform that they use, whether it's Lovable, Replit, Bolt, etcetera. Many people are moving from these dedicated vibe coding platforms to Codex or Cloud Code because, you know, I think we're about one or two months away from these platforms being full vibe coding platforms. 00:14:10.190 --> 00:14:38.330 And many people who use Replit say that, like, it's just significantly easier to just vibe code an app, get it on the Internet, and use it for internal use or sell it as a SaaS. Many people love these vibe coding platforms because it makes everything easy. Because after all, Codex just generates the code and then it lets you see your app in the browser. Whereas something like Replit generates the code, it makes viewing the app visible while you're building. Right? Just like the in app browser inside codex. It also sets up authentication. 00:14:38.330 --> 00:14:43.610 It sets up database and it also does one other thing which is like it has like some security 00:14:43.610 --> 00:15:03.320 things but mostly that's just an AI prompt and then it also hosts the app on the internet. Well, what people are realizing now is that all of these are just like a single prompt inside codex. Right? On codex, can run a prompt like this. You can say please build an internal tool for my company to track whatever it is that you wanna track. For this example, I'm just using video stats. 00:15:03.560 --> 00:15:36.380 And you could say make this web app. Use Neon Postgres which is a database service for database. As long as you have an account on Neon and you set up the plugin, this just works one shot. And then you can say use Google for sign in, um, and for off. And then you could say use Vercel for hosting. Right? And this, uh, puts the app on the Internet. And then you could say use AI gateway for AI features. So this is another Vercel app where all you need is to sign in to Vercel, get one single API key and once you set that up you can use any AI model. You can also use something called Genmedia 00:15:36.460 --> 00:15:40.220 which is all of the image and video models and this is by FAL. 00:15:40.565 --> 00:15:46.005 And so I've already set this up and made this skill so I can build any app with any AI feature 00:15:46.245 --> 00:15:58.440 or AI video model directly inside the app and then I can just say like make sure to run many security checks. GPT 5.5 extra high is incredible for checking for vulnerabilities. 00:15:58.520 --> 00:16:03.640 So you can just fire off this whole entire prompt and this basically solves 00:16:04.120 --> 00:16:08.040 for the entire value prop of tools like Replit 00:16:08.255 --> 00:16:09.295 and Lovable. 00:16:09.535 --> 00:16:14.255 And soon, I believe there's going to be someone who builds a fully 00:16:14.335 --> 00:16:15.615 AI native 00:16:15.775 --> 00:16:17.935 AI native version 00:16:17.935 --> 00:16:22.990 of Replit and Lovable. And this is a product that our team and I, we considered building this tool, 00:16:23.310 --> 00:16:35.215 um, but we just kind of we fell out of love with building static apps. Agents are just way more fun to work with. But someone could very easily build an AI native Replit and Lovable which acts as a plugin. 00:16:35.375 --> 00:16:36.815 And so you could create 00:16:37.135 --> 00:16:48.070 a skill which handles all of this stuff right here for the user and build it directly inside Codec. So that's one of my big predictions for the rest of 2026. 00:16:48.230 --> 00:16:54.870 Someone's going to build a replet and lovable that makes it as easy it is to use lovable but inside Codex. 00:16:54.950 --> 00:16:57.590 Because with replet and lovable, you use 00:16:57.910 --> 00:17:00.070 their tokens and you use 00:17:00.575 --> 00:17:01.455 their 00:17:01.455 --> 00:17:09.535 agent. And so the replet agent is actually worse than just using codecs out of the box and it's more expensive because OpenAI 00:17:09.535 --> 00:17:10.895 heavily subsidizes 00:17:10.895 --> 00:17:13.100 users to use GPT 5.5 00:17:13.100 --> 00:17:20.700 directly in the app. And so someone could build an AI native version of Replent and Lovable where it's just BYOT 00:17:20.700 --> 00:17:23.500 and BYOA, 00:17:23.500 --> 00:17:28.535 which is bring your own tokens and bring your own agent. You So can imagine a world 00:17:28.775 --> 00:17:34.135 where I go to Codex and I could say build an app and use, 00:17:34.775 --> 00:18:07.485 uh, at use at, uh, Lava Plit. And this is my fictional app that someone could build where it just handles all of that except it acts as a plug in and you use it directly inside Codex and maybe it only cost $10 a month because this company that get that creates it doesn't have to build an agent. They don't have to pay for tokens so it's a bigger margin and they just host the user's web app somewhere and maybe that could cost a little bit more money. But I genuinely believe that many people who love to vibe code are just gonna end up switching over to Codex and Claude 00:18:07.725 --> 00:18:10.845 desktop app over time as they become full 00:18:11.130 --> 00:18:11.770 platforms 00:18:11.930 --> 00:18:25.610 and vibe coding will just be a skill that any AI agent can do. To conclude today's video, I wanna talk about just my biggest obsession for the past two months and it has to do with something called an agent mini app and it stems 00:18:25.930 --> 00:18:31.795 kind of from the in app browser inside Codex and eventually all agent platforms. 00:18:32.195 --> 00:18:36.115 Okay. So in my previous video, I covered a topic 00:18:36.275 --> 00:18:42.515 called an agent native app and I used the example of Dan Shipper who created this app called Proof. 00:18:42.970 --> 00:19:10.005 And Proof is this document editor that's open source that he made to be an agent native app or an app that you use with your agent. So you could say, hi agent. I wanna create a document. And the agent can create the document and then you can edit the document yourself. You can have the agent edit the document, and he basically, he made the connection between the document and the agent incredibly easy. It's very seamless to create a document with this agentic 00:19:10.005 --> 00:19:10.565 application. 00:19:10.900 --> 00:19:33.435 And I've been fascinated by this because we're gonna have agents that will have browsers connected and so many people are gonna make a ton of money building apps that are just agent native. They're not meant to be you for you to go to the app and type a document on their platform. It's made for you to ask your agent to create a document and it just uses this technology and renders it right here. So this is really really cool 00:19:33.675 --> 00:19:45.650 and really interesting and it's possible right now to create and use these agent native apps. In fact, Google Docs now because your agent can fully control Google Docs, it can fully control Notion. 00:19:45.730 --> 00:19:53.170 This is an example of an AI native app. Right? It is an app that's meant to be used by humans but they added like an agent native 00:19:53.250 --> 00:19:55.570 feature. Right? This is just an AI 00:19:55.765 --> 00:19:57.525 agent native feature 00:19:57.605 --> 00:20:02.245 of like an app that's meant to be used by going to the platform. So this is all possible, 00:20:02.405 --> 00:20:09.925 but there's one thing that's not possible. So on Codex, they have these things called the plugins. But within the plugins, 00:20:10.085 --> 00:20:15.610 right, you can actually sign in to all of your apps. And so I have like 30 different plugins like Gmail, 00:20:16.090 --> 00:20:17.370 like Slack, 00:20:17.690 --> 00:20:19.690 like, uh, TypeFully, 00:20:19.930 --> 00:20:24.815 which, uh, allows me to schedule Twitter posts for the future which I use a lot for our company account. 00:20:25.215 --> 00:20:27.855 Um, and you know, the list goes on. GitHub, 00:20:28.335 --> 00:20:29.535 uh, Vercel, 00:20:30.015 --> 00:20:30.815 etcetera. 00:20:30.895 --> 00:20:37.950 All of these different tools. What is not possible right now inside Codecs that I wish was possible, 00:20:38.270 --> 00:20:39.150 you cannot 00:20:39.390 --> 00:20:40.190 create 00:20:40.270 --> 00:20:45.470 an AI native app that connects to these specific integrations. 00:20:45.550 --> 00:20:48.670 Right? When I go to plugins and I sign into my Gmail, 00:20:49.095 --> 00:20:50.455 I'm authenticating. 00:20:50.455 --> 00:20:52.055 Right? I'm authenticating 00:20:52.135 --> 00:20:53.015 to my 00:20:53.255 --> 00:20:58.375 email. What I can't do inside Codex is use this authentication 00:20:58.855 --> 00:21:01.575 to create an app that connects to Gmail. 00:21:01.655 --> 00:21:13.180 Let me explain what I mean by that. So if you think of the way we were describing vibe coding earlier where you have your different agent task, you're chatting with your agent and you can get it to create basically any app you want 00:21:13.580 --> 00:21:16.220 and I'm able to add Neon's, 00:21:16.220 --> 00:21:21.315 uh, database to it by at mentioning Neon. Right? This is just a database provider 00:21:21.395 --> 00:21:25.555 and then it can create an app that has a built in database created by Neon. 00:21:26.035 --> 00:21:29.475 But what if what if your agent 00:21:29.555 --> 00:21:30.675 could generate 00:21:30.675 --> 00:21:36.630 apps here on the side which I call a mini app which could actually 00:21:36.870 --> 00:21:39.430 integrate with all of your plugins. 00:21:39.510 --> 00:21:41.110 And so you could generate 00:21:41.270 --> 00:21:51.885 a email mini app or you wouldn't even need to consciously generate an email mini app. Your agent would generate it for you. So imagine you're using Codex and you say something like, 00:21:52.205 --> 00:21:52.925 I need 00:21:53.565 --> 00:21:54.365 to do 00:21:54.685 --> 00:21:56.605 my email help. 00:21:56.845 --> 00:22:01.805 And the agent one thing the agent could do is just send you a bunch of drafts 00:22:01.290 --> 00:22:12.730 to all of your emails. Right? It can go through and look through your email. It could come up with drafts to send, but it's really hard to like give you that information in a way where you could easily edit those drafts. 00:22:12.810 --> 00:22:18.565 What if it created a mini app and the mini app was like a Tinder 00:22:19.125 --> 00:22:19.685 for, 00:22:19.925 --> 00:22:21.125 uh, email? 00:22:21.125 --> 00:22:33.230 And so it had like it had like a nice input message which is like the person who sent you the message and then it had just like your response. So like it put your response below it and then you could either, 00:22:33.630 --> 00:22:56.045 um, archive, right, if you don't actually wanna send it send the email or you can just send it as is and since the agent has context over all of your different tools, it'll be really good at understanding your goals and everything. It'll actually be able to draft a really good email or there would be like an edit button. Let's say you just wanna edit like a few parts of it. You could very quickly edit it and within the app, 00:22:56.285 --> 00:22:57.965 you could just press send. 00:22:58.285 --> 00:23:21.705 So imagine it created an app that you could easily press send. And as you use these apps, right, as you use these mini apps, you would actually learn. Right? Because every time you press archive, this data would be stored somewhere. I'm not sure how this would technically work but this would be stored somewhere and over time the agent would actually not make suggestions for the types of emails that you would normally archive 00:23:21.945 --> 00:23:39.770 and it would learn from every single message that you send. It would learn from all the edits that you make so that every time it suggests an email, it's one that you will very likely send at a very high confidence. So these can be thought of as just like generative UIs that connect with your integrations because right now 00:23:40.090 --> 00:23:52.715 you could ask it do this but then you'd have to go back to your agent and say send the first one, don't send the second one, send the third one, make an edit to the fourth one, please say this. What if the agent could just send you the best possible interface 00:23:52.920 --> 00:24:03.240 that connect with the tools that allowed you to just make the final 10% edits and send it directly in this little mini app? And users would actually be able to create their own interfaces. 00:24:03.240 --> 00:24:09.105 Right? And you could create your own mini apps and maybe even share them with your team because every person's unique, 00:24:09.345 --> 00:24:15.105 every company's unique, and maybe you want to create your own little mini apps that are integrated 00:24:15.185 --> 00:24:25.930 with all of the things that you've already signed in with. Why would I want to use someone else's external platform if my AI agent can generate a UI for me right when I need it. 00:24:26.330 --> 00:24:28.250 And I think this is next, 00:24:28.410 --> 00:24:37.610 you know, and this is just something that like we've been playing around with and my company in New York, we are I moved my company to New York and we're actually trying to figure this out through iMessage. 00:24:38.055 --> 00:24:41.895 I'm not gonna go into detail because I'm gonna be doing like a big announcement soon, 00:24:42.215 --> 00:24:45.415 but you can actually already use our product. It's chorus.com, 00:24:45.735 --> 00:24:51.975 uh, and you can create an AI agent and add like, uh, an agent like Claude Code or Codex directly inside iMessage. 00:24:51.975 --> 00:25:01.710 And we're trying to figure out how the agent can send you a little link which turns into a mini app. And these mini apps will kind of act as like the operating system for the agent. 00:25:02.030 --> 00:25:21.440 I genuinely believe that all of the major platforms are gonna kind of circle around this idea, and this what's gonna bring out Jarvis. Right? How can the AI agent give you the best possible interface for any given task that you can use and and the app actually connects to the integration? You can actually send an email. You can actually post the social media post. 00:25:21.600 --> 00:25:29.120 You can actually send the Slack message. Right? It can suggest things for you and you can properly edit them directly in the interface 00:25:29.200 --> 00:25:35.600 and I think Codex has a perfect browser for this. The problem is if you try to do this, you actually can't 00:25:35.975 --> 00:25:56.590 connect your plugins to the apps that you create. It's just not possible with the way that they built codecs. Anyway, that's it for the update today. Yes. So I'm here in my Airbnb in New York City. We just moved our company from SF to New York. It's great energy out here, but unfortunately, I don't have a studio. So we're gonna rebuild our office, 00:25:56.990 --> 00:25:58.350 rebuild our studio, 00:25:58.510 --> 00:26:01.790 and I'm going to be 10 x ing my content effort. 00:26:02.535 --> 00:26:19.870 My main goal is just to educate people so that you become agent native, uh, which is the new name of this series. I think people need to become agent native or agents will just start to use you. You could think of social media. Right? If you look at the social media trend over the last ten years, right, there's content creators, 00:26:20.190 --> 00:26:30.590 right, who kind of take advantage of social media. And then there's just like the content consumers who kind of get taken advantage of by the algorithm. It addicts you to the platform. It sells you ads. 00:26:31.275 --> 00:26:34.955 And so like there's kind of this like you're either a producer or a consumer. 00:26:35.275 --> 00:26:41.035 I would much rather be on the producer side of this AI revolution. I think it's really important to learn the different concepts. 00:26:41.275 --> 00:26:43.115 Um, you should learn the surfaces 00:26:43.275 --> 00:26:50.080 that these AI agents will exist on, which is why I started this series. So every week, I cover the most important agent news, 00:26:50.320 --> 00:26:56.880 and I'm I'm loving it right now. And I'll continue to do it every single week. So thank you guys for watching. I'll see you here for the next video.