¶1I've been using Chai GPT for years now and it gets so much done for me. And in this video, I want to share all the best practices that I've learned from using it for hundreds of hours to get the most out of it and to make it as easy as possible. The first thing that comes to mind that I don't think most people take advantage of is browser use. Codeex can literally browse the internet on your behalf, saving you so much time. Watch this.
¶2From one prompt, I'm going to have Codeex go do research on different cameras that I should purchase for my studio that are comparable to the camera I'm using right now, put it in a spreadsheet, and get all the information that I need to make a decision on which camera I should buy. Keep in mind, that's only one use case for browser use. There are so many. I've used it to get refunds, negotiate with customer service, go through my email and archive emails that I don't need to look at, and so much more. 9 minutes later, it put together this gorgeous Excel spreadsheet with the different options, prices.
¶3We have the delta in price versus the camera I already have, the ratings, and a bunch of information that will help me make the right decision about the camera. Very easily done. All done for me. And Codex can not only control your browser, it can also control your computer if you wanted to. And you can have it organize your files.
¶4You can look for bloat and delete it, make your computer speed up. It has helped me so many times clear out my memory, just as an example. All right, the next thing I want to show you is a brand new feature that Chat GPT just launched, which is voice mode natively in Chat GPT codecs, where you can actually have it control the different agents that are working for you. So check this out. You just click new chat and then there's this button right here and then you can talk to it.
¶5And so I can say, okay, spin up a new thread and write a poem that's 100 words long. >> Okay, I'll spin that up. I'll create a fresh thread for that and keep it available for you to talk to. The poem task has the exact length requirement and it's generating now. >> Look, it spun up this new thread.
¶6You can see that right here. sent by codeex from another chat. Write an original poem that is exactly 100 words long. And there's the poem. And I can just continue talking with voice mode indefinitely.
¶7I can run all of my agents, all of my threads, all through voice. It's actually quite incredible. And you need to be using it. Here's another thing you might not know. You can publish basically anything to the public web using something called chatpt sites.
¶8And as an example, let me publish this poem. Publish the poem using chat pt sites that I just wrote. >> Got it. I'll check the publish path. Now >> whether you need to publish a spreadsheet or a PowerPoint or your website, a portfolio, anything, it's very easy using chatgbt sites which is built in to chat GBT.
¶9So, here you can see it is saying create site. This is using the site's feature. And here is the actual poem designed, which I didn't ask it to do, but great. And this is what the website's actually going to look like. So, here we go.
¶10It's finishing publishing. By default, it'll publish it as a private website, but all you have to do is say make it public, and it'll be public. Here's the URL. You can send it to anybody. So, chat GPT sites, very useful.
¶11Put something on the web, share it with anybody, share it with another agent, by the way. You can just copy that URL and give it to another agent. Then that agent has access to whatever you just published. It's super useful. Next, a couple quality of life things I want to show you.
¶12First, pinning. So, if you don't know about pinning, you can pin threads, chats on the top of your left sidebar. So, right here is the pinned category. You can see it right there. All of these are the ones that I've pinned.
¶13And if you want to pin something, it's actually quite easy. You just rightclick on it and do pin chat. There's also this nice little pin chat icon that you can click right here as well. What I use that for is to just keep track of the most important conversations that I'm working on. So, as you can see, I have a bunch of different projects that I'm working on all at the same time.
¶14Sometimes things finish and they're kind of way down in that list and if it's something I want to remember and keep a close eye on, I just pin it. So, next I want to talk about one of the things that intimidates most new users to chat GPT and AI in general, which is model selection. But I'm going to break it down very simply for you. But first, GenSpark is sponsoring this video. Let me tell you about them.
¶15This video is brought to you by GenSpark and their new Second Brain Note device. This is for people like me who finish a day of meetings and think, I don't really remember what I talked about today. Second Brain Note is a credit card thin AI recording device. It snaps to your phone with MagSafe and you just tap it once to capture a conversation. I decided to use it in my morning team routine.
¶16Most recorders hand you a transcript or a summary and just stop. But second brain actually gets more useful to you as you use it. Over time, it builds up a personalized memory based on your meetings, your notes, your ideas, decisions. So I can ask things like what was this morning meeting about and second brain will immediately retrieve the context and then the GenSpark super agent can turn it into a list of important action items that I need to get done. Before that it meant time spent after every meeting organizing my notes, summarizing them, figuring out what action items I actually need to take.
¶17Check out this limited first release of the Gen Spark second brain note and get 10% off through the link down below for my viewers only. One of the most important concepts I want to try to convey today is which model do you choose? There are many different flavors of each of the models. There are different thinking efforts. And choosing the right model for the right task will not only decrease the amount of time it spends on each task.
¶18It'll increase the quality, it'll maximize your usage. So you have this fixed quota that they give you each week. and you don't want to just blow through the quota using kind of the best biggest model on a simple task. So right here in the chat you click right there and first you select the model you have 5.6 and three flavors of that Luna Terra and Soul. You still have 5.5, 5.4, 5.4 mini.
¶19I'm mainly using these three. Luna and Soul are my go-to. I'm either using Luna to do things really quickly and save money or I'm using Soul for my most complex tasks. Now, within Soul, once you've selected that, you also select your thinking effort. For me, I'm usually using medium, high, or extra high.
¶20I rarely use light, and I rarely use the max setting for soul. And the way that I think about whether to use soul, or Luna is obviously based on the complexity of the task. for the really hard task I'm using the bigger model soul. So let's say I'm building a website I will generally use soul to actually build the website and then the fine-tuning of the website like changing fonts or colors or moving text around moving images around that can all be done with Luna because it's capable enough for those simple changes. All right, the next thing I want to show you is scheduled tasks.
¶21On the left side, there's a little tab that says scheduled. And this is where all of your scheduled tasks go. Scheduled tasks are exactly what it sounds like. They are tasks that run on a schedule. So, I have three that I use every day.
¶22I have this daily stale file cleanup suggestions. And if I click into it, you can actually see the full prompt. I have daily Astro production log review. Astro is a project I'm working on. And basically what I'm asking it to do is look at the production logs for any errors and tell me about them.
¶23And then daily forward future production health. Just checking forwardfuture.com for any issues. It just goes uses browser control makes sure everything works. Then within it you can say okay run in a new chat or an existing chat. So a new chat means it's going to spin up a new thread on the left side every single time this runs.
¶24An existing chat means it will use the same thread every single time. I don't really see a huge difference between these. It's more just personal preference. Then we have project. You can put this inside a project inside these folders over here or you can just leave it outside and either one it and either one works.
¶25Again, maybe you want to create like a recurring task project that all of them get grouped into. That's actually not a bad idea. Uh then you select the model. And this is what I really like to do. Luna is so inexpensive and fast that you can set up a ton of these scheduled tasks to run daily.
¶26Anything you can think of that you do on a recurring basis, you should set up as a task and use GPT 5.6 Luna Max Reasoning and you're really not paying much for those tokens. It is nearly free to have these scheduled tasks running every single day. So, you should really be scheduling a lot of tasks for yourself. One that seems really useful, which is a suggestion by Chai GPT, is to start each workday with a summary of your calendar, unread email, and priorities. This is kind of an obvious one, but super useful, and you can use Luna for this.
¶27It is more than capable to do that specific task. So, definitely make sure you're using scheduled tasks. The next thing I want to show you are plugins, which is a way to give your chat GPT many more abilities. and specifically with apps that you already use. So you can give it access to control things like Google documents, to use PDFs, to use spreadsheets.
¶28Here I give it access to GitHub and now it knows how to use GitHub, Google Drive, Gmail, Calendar. There's so many of them. And if you come in here, you can search for all the apps that you already use. Aauna and Notion and Linear and Dropbox. Most likely if you're using a piece of software, it will appear as a plugin that you can give access to chat GPT.
¶29And so by doing that, you make it so that Chat GPT doesn't have to guess at how to use one of these products. And so if you want it to check your email for you, give it access to Gmail and it'll just know how to use Gmail. The next thing I want to briefly talk about are skills. skills are incredibly useful if you do the same kind of task over and over again, but it's not quite scheduled. It's just something, some kind of workflow, some kind of process that you do that you have to usually type out this long prompt, tell Chad GPT about it.
¶30You can be making skills from it. And not only that, you can not only create your own skills, but you can actually find a ton of skills that people have created and published for basically every use case. And to invoke a skill, all you do is type slash and then the list of skills appears right here. And you could just start typing if you want. But these are all the different skills that I have given to my chat GPT.
¶31And it's awesome. So if you find yourself doing something again and again, just make a skill from it. And here's an example. Let's say, I know this is kind of silly, but I need to actually do research on a new camera once a week. So, I can just say, "Create me a skill from this camera research thread." I hit enter and it'll create me a skill.
¶32It'll save it. And then going forward, I just type slash and then whatever it's named. All right. Next, I want to show you something incredibly useful. Slashgo.
¶33Slashgo makes CHP run for a very long time. But not just for the sake of running a long time. It is for specifically achieving some goal. And you're basically telling Chad GBT, continue to run until you hit this goal. And a goal can be one of two things.
¶34It can be something verifiable. Continue until the speed of my website is 50% faster than it is today. Or it can be LLM as a judge and you can say continue until the speed of my website is as fast as you think possible. And so you can see the difference. One is very verifiable.
¶35it is something that is a concrete number and the other you're leaving up to the LLM to decide but either way you can do that. So as soon as I hit /goal you can see this little goal right here and then I will describe the goal and define a measurable outcome and once I do that this can run for a long time. I've had Chad GPT agents running for days using goal. So be careful using it. But it can be very powerful if you really do have a concrete goal in mind.
¶36And keep in mind you can give it limits other than the goal. You can say, "Okay, don't run for more than 3 hours." Just as an example, goals are super useful. They're a very advanced feature, but can be extremely powerful. And by the way, if you need help figuring out the right goal to create or inspiration for the best goals that might be relevant to you, I'll drop a link down below to that and also the loop library which lists a bunch of awesome loops. You can call them loops or goals that you can use right now.
¶37All right, the next thing to keep an eye on is your quota usage. Once you start using your Chat GPT subscription a lot, you need to actually know how much of your quota you're using. And so there's a weekly quota. And you can actually find it by clicking on your name, clicking your usage percentage. You can see I've already run out of quota.
¶38And it will say right here how much quota you have left, as well as the reset date. So it's August 7th. I'm recording this on August 4th. I have three more days before the reset. and I've already used all of my quota.
¶39Now, in settings, if you go to usage and billing, you can see more information. You can see your plan. You can see your current balance. Uh, and then if you have resets, they're called banked resets. You can use them right here.
¶40And what a banked reset is is the Chad GPT team will often, it's been a lot lately, give you resets of your quota. So, you can try to use all of your quota and if you still have a few days left, like I do, you can click reset and your quota will go back up to 100. Keep in mind, they do expire. So, make sure to check right here where it says usage limit resets if you have any. And it'll also say the expiration date.
¶41So, keep an eye on that. All right. The next thing I want to show you, which is really a power user feature, is the fact that a single thread can see all of your other threads in chat GPT. And this is actually super useful. It might not seem so at first glance, but it is extremely valuable.
¶42I have tons of threads and sometimes I have threads that conflict with each other. Sometimes I forget where a thread is or what I did, what was the conclusion. All of that you can check within a thread. And you can also have a thread delegate to other threads. You can have it create other threads.
¶43Just think of chatpt as one giant singular system. The threads are really just for you to help better organize your thoughts. But a single agent can access any of these threads. You can ask it questions about any of these threads to summarize it. You can have it spin up new threads and delegate work out to it and wait for it to finish and then continue in the thread you're on.
¶44There are so many different things you could do. I could say, you know, which of my recent threads was about a poem. So, I forget where it was. I can't find it for some reason. And it'll find it for me.
¶45So, I asked which of my recent threads was about a poem, and it found the two threads. I said, "Link me and it actually opened up that specific thread." And now I can say, "Okay, in that thread, write a 50word poem." And you can see how threads can control other threads. So, I'm sending that request to the 100word poem thread now. And there we go. Write an original poem that is exactly 50 words long.
¶46And you can see again sent by codeex from another chat. So it can be super useful. So I definitely recommend checking that out. All right. Next I want to show you the multiple ways you can actually use chat GPT.
¶47Some of this is for coding use cases but you can use this generally as well. So first the obvious one is local. You're just using it on your computer and you open up the chat GPT app and there it is. Everything that your agent is doing is local on your computer. Next is cloud.
¶48If you're coding, you can put your coding project in a cloud environment so that codeex is running there and will actually run the code, execute the code, edit the code all within that cloud environment. You're basically putting it on somebody else's server. This is nice because you can run many more agents in parallel whereas that might slow down your actual home computer quite a bit. And last, the thing that I use all the time is called connections. Connections allows you to remotely connect to your computer that has Chat GPT running from any other device.
¶49So I am frequently taking my phone. I'm out somewhere and I want to code something. I will open Chat GPT on my phone and it will have a direct connection to my desktop computer. So all of that code that I've been working on is then available anywhere in the world on any device that I'm using. So the way to set that up is you want to go into settings down here under connections.
¶50You want to say control this Mac allow connections. Yes. So you're going to click this add button. It's going to give you this QR code. You simply take out your phone, point it at it, and it'll open up Chad GPT and ask you to accept the connection.
¶51And then from then on out in the Chad GPT app on your phone, you can connect to your desktop computer. So then on your phone, there's this little remote tab. You just click that. And then at the top, you can see I'm now connected to my Mac Studio and all of the threads that I have on my Mac Studio are now available to me from my phone. I can continue coding from anywhere.
¶52So, those are a bunch of ways to get the most out of Chad GPT, the most out of Codeex. Hopefully, you found value. And I know I talked about loops. I actually made an entire video specifically about loops. Check that out right here.