Transcription
If you're watching this video, you're probably a user of ChatGPT, but today I'm going to explain to you why I think that may not be a wise choice. Now, I'm not saying ChatGPT is a bad model. In fact, I still sometimes use it to this day, but there is a much better model on the market that in my opinion presents much better value for money.
Now, this gap a couple of months ago was really close. Admittedly, it was neck and neck, but over the past few weeks especially, the new features are just too big to ignore and the functionality is just too great to ignore. So, today I want to do the video to officially say I believe if you are currently using ChatGPT and you can only pick one model, I would be switching over to Anthropic's Claude.
In today's video, I'm going to break down the reasons why I would be doing so, the reasons why I've personally been using Claude as my go-to model over the past few months, and I'm going to give you a live demonstration of some of the stuff that Claude can do that GPT can't do, because honestly, some of it is mind-blowing. So, if you do enjoy content like this, make sure to subscribe to the channel. And without further ado, let's get into the video.
So, what is the problem with GPT? So GPT was the front runner and when it officially launched for an entire year, it was the model that I used the most. It has an extremely sticky user interface and as you talk to it more, it learns more about you. So it becomes harder and harder to migrate away, which I think is one of the reasons why a lot of people are stuck using GPT even though frankly it just isn't the best model in the market anymore. And that's because their rate of shipping has been extremely slow compared to other models. Now it's not a bad model inherently. It's just that Claude has become so much better and wins across a variety of fronts.
All right, so let's go one by one through the areas why I think Claude is better and then I'm going to give you at the end of the video an assessment and if you do choose to switch over, I'll show you exactly how to do so so you can actually transfer your information with you. And stating the obvious here cuz I don't think they do sponsorships, but no, this video is not paid by Anthropic. I'm a genuine user and subscriber of both, but I'm trying to make sure you spend your money in the right place. And I do believe Claude is that place right now for the majority of people. And I actually think if you're only using ChatGPT, you're not just being cost inefficient, you're also costing yourself advancement because you're not getting access to these features.
So firstly, where Claude is really good is the writing quality. The outputs on Claude are a lot more human in my opinion, making it the better fit for most writing tasks. And I personally do a lot of my content on Claude. I find it to be the better model for creating content, for writing scripts, for writing articles and in terms of its even strategic thought process behind the way it lays out writing I find is better than GPT where GPT is actually arguably better strategy. So anything from a business executive CEO perspective I will still often gravitate towards GPT it is pretty much neck and neck but on the writing front is a lot better.
The second thing that I think is so much better on Claude, and in fact, it's pretty much non-existent on GPT, is artifacts. Now, what are artifacts? Artifacts are these pop-ups that open side by side with your chat that allow you to code in real time. So, obviously, there's Claude code, which we all know is one of the best, if not the best coding model in the world. GPT's Codex 5.3 is very good by the way, but Claude still on a value for money efficiency basis is the best. But even without having to use Claude code, just on the application itself or the web app itself, the fact you can get live artifacts like this is just a spreadsheet I was creating before for another video whilst you are typing is absolutely insane. I even did another dashboard before with um a sample financial dashboard that you can open up and you can actually edit a dashboard that you create by just typing in the text box on the left. So I just coded this. I asked it to create a dashboard for a sample $100,000 portfolio. It created a net worth graph asset allocation. Of course, you could even plug in your real data to this and I vibe coded this in literally 30 seconds. And as I make iterations in the chat section, it's coding on the right hand side. You can use this to create websites, funnels, pretty much any visual representation on the spot. That is something that GPT really struggles with. So for anything visual, Claude is just much better. Even this presentation that I'm putting together to collect my thoughts here, this is also created on Claude because it is the best model for anything visual and that is just an element with GPT that you're completely missing out on. It goes deeper than that as you'll see later in the video, but artifacts are just a huge win for Claude.
The next thing that I believe is a huge win for Claude is projects. So, if you go into Claude and you click on projects, you can see that you can create individual folders which act as specific context files for something that you're doing. For example, I've created projects for different content businesses that I have. So, AI Edge, my lifestyle content, my crypto content or my strategic business executive assistant chats which I use to get strategy and ideate. So, I'll give you an example. If you click into one of these projects, I haven't added all the files to this one, but you can create a series of sub-chats underneath it and you can put in specific instructions. So you can write in the description of the project and instructions so it can behave in a certain way when you're chatting. Now this basically means that when you are chatting it's constantly pulling information based on the instructions and you can also put files in. Some of my chats I've actually put up to 36 documents in there 36 pages so it can automatically scan that data to make sure the corresponding outputs are in line with my data. So, for example, what you could do is you could create, you know, a one-pager MVP with all the details about you, your life, your business objectives, put it into the folder, and it will actually help the answers be more tailored for you. If you have a financial analyst project, you could have all of your financial situation information, the way you view the market. Obviously, you can update the numbers, the general context of how you want to invest over a 10-year period, you would put in there. So, then every time you prompt it, oh, what do you think about Apple? What do you think about this? What do you think about that? it's in the context of your investing, which might be a 10 to 20 year time horizon, and then you don't need to keep reminding it.
Now, yes, GPT does allow you to add files to a chat. I've just personally found that the way Claude has laid out its projects and the way that its memory function operates means that it forgets far less. And this is the important part. Actually, in the past week, Claude announced a new update which has enhanced memory even further. So now, memory syncs across all of your chats. you know on GPT where you might message one chat and then the other one might not remember. Well, that is something of the past with Claude because it actually remembers across all chats and that is a massive thing. Now, yes, there are ways to engineer this on GPT by creating JSON profiles and you can keep giving it contextual data and you can achieve a similar effect. I'm not saying it's impossible and you can create custom GPTs. Of course, I want to be unbiased here, but I just think Claude does it a little bit better and I think it's a little bit more intuitive. Hence, I think it's the better option if you actually want it to remember what you're doing.
The next one, and this is a massive one if you are a builder, is the coding ability. Now, I will say this, Claude code is actually in some ways inferior to GPT's Codex 5.3 in terms of really heavy-hitting code. So, you know, for really in-depth code, Codex 5.3, which is OpenAI's model, often has a higher success rate. But the sheer cost efficiency that you get from Opus 4.6 with Claude Code is insane. And I still feel like for day-to-day code, it is the best model to code on. And then you can actually use this in conjunction with 5.3 to review your work. But Claude Code is amazing. I'll do future videos on it. Pretty much all of the vibe coded applications you are seeing are being coded on Claude Code. And as you saw before, you kind of have a light version of Claude Code in your chats as well that represent themselves through the artifact. So you can actually code even mini working applications already in the chats. But if you want to go a step further, you can actually use either the terminal or Claude code within the desktop app itself, which is going to give you even better results. So GPT is good, but Claude is just a lot better in terms of a cost efficiency point of view and time efficiency as well because you can iterate much faster. They even announced a new update where you can actually code from your phone accessing your computer's terminal. So this was one of the big value propositions of OpenClaude. But now in your phone, you can access the terminal. So you can vibe code on the go. You could be in the car, you could be on the beach, you could be sitting in your living room and your computer can be upstairs and you could be vibe coding. That's something you can't do on GPT. And I really feel like Anthropic has taken this consumer-first approach. Like OpenAI is really slow to ship. And I get it. They're trying to be a bit more commercial. They're trying to appeal to more people. They're not shipping as fast. They're not trying to reinvent the wheel. But Claude is like, "We are just going to constantly ship stuff in your face and we're just going to make the value proposition so tantalizing that you're going to have to switch over." And that's why guys like me are making this video now because they're just adding so much stuff.
Something else you can do is you can create a custom writing style. And I feel like Claude does a better job at listening to that writing style. And you can actually create Claude's skills to sound like you. So if you give it a bunch of data, give it a bunch of writing context, you can actually engineer a skill. So whenever you're writing, you can uh forward slash the skill and then it will take the information from that skill in order to give you an output. And skills aren't just for writing. It can be for anything. Recently what Anthropic did is they opened up a bunch of skills that were typically for enterprise use cases and they opened it up to anybody to use either in their work or their small business or their daily life. So now you have a huge skill library you can choose from which is pretty insane and I'll get into it in a second. But co-work is one of the biggest game changers Claude has shipped. In fact, let's get into it now. What is co-work? Co-work is a feature, and this is something GPT doesn't have, and I don't know why. I know Sam Altman hired Peter Steinberger, who created OpenClaude, but Claude is just shipping a lot faster on this stuff. It's a feature called co-work, and it allows you to organize, edit, adjust, or work within a local folder on your computer. So you give it access to a folder or folders and it can do all sorts of stuff like it can find insights within files or it can organize screenshots or it can optimize your week based on your calendar within that folder or you can even go a bit more advanced and use Claude code to plug into APIs which can then automatically scrape data on your behalf and give you the output that you want in the way that you want it. Sky's is the limit, but the simple way I'll describe it is co-work is an agent that can do anything for you within any folder on your computer. So unlike GPT, it can actually edit your local desktop and do stuff on your behalf, whether it be research reports, organizing files, pretty much anything. I use co-work for example to collect data to create a personal operating system dashboard. And I'm doing a video on this in the future, so make sure you subscribe and turn the notification bell on so you see that video as soon as it drops. But that is my personal OS with my finances, my daily brief every morning. It's scanning market research and reporting and all of that is possible due to co-work.
Essentially, what I like about Claude, and I'll talk about the pricing in a second, but it's not even the super advanced features. That's not why I'm making this video. Like Claude Code, not all of you are going to use Claude Code. Not all of you going to use agents and crazy stuff. What I like is the fact that it's all-in-one. It has good memory, good project setup. It makes coding really easy. It makes visuals really easy through the artifacts. It's generally pretty good across the board. It's maybe not quite as good at strategy as GPT 5.2, but it's almost there and it writes much better. So, as a consumer product, especially with the skills that they are now commoditizing and making available to you guys, it just makes a lot more financial sense when you stack them up against each other.
So, how do they stack up against each other when it comes to price? And then if you want to switch over, I'll show you how to do that. Well, ChatGPT Plus is $20 a month. Claude Pro is $20 a month. Both of them have their high tiers at a couple hundred a month and I think they're fairly similar. I think the only real difference is and this is where GPT might win a bit is you will run into slightly more rate limits on Claude especially when using Opus 4.6. Now you can get around this by using Sonnet 4.6 which is almost as smart, not quite as smart but it is still smarter than a lot of the lower tier GPT models but 5.2 is I think a bit more efficient and lower cost than Opus 4.6. But I think in day-to-day use, you probably won't notice the difference unless you know you're doing heavy coding or heavy usage where cost may potentially become a concern. Of course, there are even new tools now like Plexity Computer, which I recommend by the way if you can afford it. It is a little bit expensive right now, but that can combine both models, which is crazy. So, you can actually do research or strategy on GPT and then you can do writing on Claude. So, you can combine both together in one interface without having to like manually, you know, copy and paste across which can get quite annoying over time.
So let's assume you did want to switch over and this applies to any LLM by the way. How do you actually get the data over? Because this is the concern that a lot of people have. It's basically I have all this data. I have all this memory on GPT but it's not on Claude or other LLM. So what you can do is you can go to GPT. You can click on your profile. Then you want to click on settings. Then you want to click on data controls. Then you want to click export data. And then it will send you an email with the data. Once you get the email, you can save it as a zip file, which you can then import into Claude after, of course, you open an account. So, you can just do this by signing up, creating an account. I actually do recommend before you cancel your GPT subscription or do the full export that you do play around with Claude a little bit even on the free tier. And if you do have the luxury of having both, I still feel like that is the best combo because GPT is still a good strategist and it actually has better voice prompting. That's something else it does well. I mean, it's very similar, but on the desktop app, Claude doesn't even have voice prompting. You have to use dictation, which is an annoying workaround, whereas GPT has better inbuilt voice prompting. That is actually a big plus because when you're doing longer strategy sessions, it enables you to get a lot more information out. So, I still find myself in long strategic discussions using GPT, when I actually want to create it into writing or a script or whether I want to, you know, create it into a dashboard or something visual, then I just flick it over to Claude. So, I do use them hand in hand. But anyway, if you can run both simultaneously, even with Claude on the free tier, you can get a feel for what it does well and then you can always upgrade later.
The other thing that will help you set up Claude is, as I said before, actually creating projects and putting in memory or specific details which will help it understand you a bit better. You can even voice prompt GPT for this specific information. So, let's say you have a strategy chat in GPT. You can just be honest and explain, hey, I'm moving to another LLM. I need all the necessary context from this chat. Everything you know about me, who I am as a person, this specific chat and then you can use that to create a one-pager which then you copy and paste into Claude. So you can also do it manually as well for specific project setups on the Claude interface.
So those are some of the reasons why I recommend if you're using GPT and you have been for a while, you should consider making the switch over to Claude. They are shipping at a rapid pace and I'm going to keep you up to date on the channel with how I'm using it, how I'm leveraging it. I also do some content on GPT as well cuz it's something I genuinely do leverage, just nowhere near as much as I'm using Claude at the moment. So stay tuned for that. Make sure to subscribe and I will see you in the next video. Peace out.