Loading summary
Transcript17 lines
- [00:00]
A
If the world were like a Sleep Number mattress, everything would adapt for your comfort. Your coffee would never get cold, your flight would always be on time, and your mattress would always respond to you and your unique sleep needs. Sleep Number mattresses are designed to respond to your body and shift with your changing life, helping you fall asleep faster and stay asleep longer. To experience a whole new world of comfort, visit a Sleep Number store or go to sleepnumber.com sleep number to a good life Sleep when you're a maintenance engineer in a beverage manufacturing plant, you keep production lines moving and quality on track because there is no room for slowdowns. With Grainger's vast selection of high quality motors, sensors, belts and hard to find parts, you can get what you need fast and all in one place so nothing gets in the way of getting the job done. Call 1-800-GRAINGER clickranger.com or just stop by Granger for the ones who get it
- [00:59]
B
done well, we have big news on the podcast today. Jaden is no longer using Claude exclusively. He's moved back to using Codex and OpenAI, which is big news because Claude has been the reigning king or queen you could say for forever with for at least the past six to nine months I should say, in the development space. So Jaden has Now switched to OpenAI. So we're going to talk a little bit about today. Claude versus OpenAI, anthropic versus OpenAI, maybe why you would want to use one over the other. What are advantages and disadvantages of each? So Jaden, why don't you take it away, tell us what's been going on with you and OpenAI.
- [01:43]
C
So you know that I talk on the school community that Jamie and I do about Claude a lot and vibe coding everything with Claude. At this point I feel like the two products have become similar enough and I actually think OpenAI is a little bit better. I'll show you some examples of why I think OpenAI is a little bit better. But if you want to get some of the insider scoops of like the literal tools I'm building, how much money I'm making from them, how much money I'm spending on promotion, how I'm growing and scaling my business with AI tools, we'd love to have you in the AI Hustle school community for this week or for this month only. It is free to join. We'll probably make it paid again in the Future. It was 20 to $50 a month before that, depending on the promo that we were running. But it's free to join this Month only if you've ever wanted to, would love to have you. You get access to 5 videos for free, a weekly live stream every single week for free, and you get access to a community of over 500 members that are in there. So go join the school community. You can still get the full library if you want for 19amonth, but you get a ton of freebies and it's linked in the description. Okay, so what I'm loving about chat GPT right now is. And about OpenAI right now is I think the quality of the output that they're giving you is better. I got sick of my Fable 5 credit, so I have two Claude Max subscriptions right now, spending $400 a month for both of them. And the reason why I do that instead of. I know a lot of people are like, well, why do you have two accounts? Why don't you just have one account and buy, like, API credits? Because that's what they try to push you to do, guys. The API credits are, like, they give you like a tenth of the usage that the max plans, like, cost per credit would give you. So If I'm spending $200 a month on the max plan to get that many tokens, which I use them all up, it would cost me like $2,000. So I just stick with. And by the way, $2,000, I think is being generous. I think it's probably like $11,000. But everyone thinks I'm crazy when I say that, and they make it kind of hard to see exactly how much usage you get. But either way, I have two accounts. I fully max them out. I'm using Fable 5 and, like, getting a bunch of stuff done, building a bunch of tools and software and apps and. And Facebook ad campaigns, and, like, running like, a company of 20 people with, like, five. So it's pretty amazing what you can do with it. The reason I've switched back over to Codex is two reasons. The first being that I got. Well, first of all, I have two cloud accounts and both of them are, like, maxed out of my tokens. So since they're maxed out of my tokens, I have no other option but to go to OpenAI. And I think. I don't know if Claude could have done anything better about this. They. They did something that rubbed me and a lot of people the wrong way. And essentially they had. For the last three weeks, they keep saying the deadline for Fable 5 to get removed from Claude is, you know, the 1st of July and then the 7th of July and then the 14th of July and then the 19th of July. So like every week they would tweet out and say, like, look, we extended it for one more week. The problem was in anticipation, every single week of it getting cut off, I would go and use up all of my Fable credits for the entire week or the entire month, however many I could use. I would use them to the deadline. And then when the deadline came, it was like, oh, just kidding, you get it for another week. But now I have no more Fable 5 usage unless I go pay their API. So I got sick of like, I don't know them. I just feel like not treating people like responsible adults because I have these two completely maxed out accounts that I can't even use anymore. So anyways, that is probably the, the one reason that I'm the most salty. I had a post on LinkedIn where I actually responded to Claude on, on Twitter. They said, we're extending Claude Fable 5 access on all plate on all paid plans as well as keeping Claude code weekly rate limits 50 higher through July 19th. And I said, except I already used all my tokens in anticipation of the deadline, so I have no more credits left. Thanks for nothing. Yeah, anyways, a little bit salty, but it is true. And I think a ton of people were saying something very similar. So that is the first reason why I'm moving to OpenAI.
- [05:39]
B
I mean, that's, that's valid. And I. So I am definitely not as much of a power user as you. I'm on the $100 a month plan, but I found myself maxing it out over really simple, simple tasks. I was like, I don't think this should be using this much, this many credits. Granted, I was using Fable, but one of the things I had to do was I had, I had an email list like a CSV with a bunch of data in it, and then I had, I wanted to remove some of the. I want to clean it up essentially and get rid of things that businesses that weren't really who I was trying to reach and kind of segment it a little bit and that like, I would reach my daily usage within like 20 minutes. And it would just keep like rolling and working really slow. And so I was really one disappointed with the speed, but also the amount of credits used. And I think a lot of people are in the same boat. But my question for you is, okay, so you're going to start using codecs. Can you work on the same projects or is it.
- [06:46]
C
So that is the amazing thing. Okay, now let me tell you all the things That I love about Codex. I tried Codex out like a few months ago, wasn't there. I tried OpenAI agents out two years ago or a year ago. Whenever it first came out was so bad, wasn't there. They came out with their new model, which is 6 point or 5.6 Sol. And that is equivalent to Fable 5. There's so many things I love about it. Number one, it doesn't nanny you or babysit you. If I ask it like a health related question, Fable 5's like, sorry, for biological warfare reasons, we can't answer, like if you're hyperthyroid, question, whatever. I don't have any hyperthyroid problems. But like theoretically, right? Like we can't answer that. I actually recently got life insurance and for my life insurance they came and did like a blood draw and like a urine sample and put it off for like a health screen. And they sent me back the whole report and I was like, this is awesome, Kate. You can't ask Fable 5 those kind of questions. But I plugged into ChatGPT and it gave me like a whole breakdown of like all the different things, the things I should be concerned about, look out for. If there was like an issue, it told me, like, things to like work on on that issue. Really amazing. Fable 5 does not. So number one is like the health stuff. Number two, more importantly for me was security stuff. I'm like, hey, can you do a security audit? Because like I'm vibe coding stuff. Security flaws come in all the time. I've had issues with this in the past with different projects. I don't know what I'm doing. And it's like, well, we're sorry, like for security reasons. For like, for national security reasons. In case you just happen to be North Korea trying to hack the US government, you can't do that. So it lowers you to like a dumber model, which isn't that bad. Opus 4.8. But it really, it just bothers me that it like tells me it can't do stuff. It switches to older models and it doesn't let me like do what I want to do with the best model. OpenAI has none of those problems. You turn it to 5.6 Sol extra high and it will do anything you ask it to do, which is amazing. But beyond that, also what's really cool about it is it takes about half as many tokens to get the same amount of work done on OpenAI versus Fable 5. And they also help you optimize. So I'm just, I've opened up My Codex thing. Oh, by the way. So OpenAI did this really cool thing where they used to have, like, Codec or they used to have Codex. That was their Claude code competitor. They had an app for that and I had it. Then they had an app for ChatGPT, which was just chatting with ChatGPT. And then they had an app called the Atlas browser, which was like a browser. Anyways, they smashed all of them together, deleted the other two. There's just the Chat GPT app, but it has what's called ChatGPT work. It's OpenAI's cowork competitor inside of there now. And they also put the Atlas browser in there. So it can, like, browse for things. It can pull up a browser, it can show you what's going on, and it could. It could do everything for you in one app. So that's number one. It's amazing. They consolidated that. I mean, they're just copied Claude. They made everything, like a really nice copy of Claude plus better. They have folders for your projects on the side where Claude is just like a giant. A giant, like, thread that you have to go find old stuff. But it's like folders by my projects. So the cool thing with Codex, though, is I'm like, I want to work on a project. I literally go to my Claude because, like, Claude makes a folder on your computer and then it has your projects inside of there. I just go click on the folder inside of my Claude folder and it, like, if there's a code base in there for a project, it just starts editing that code base directly. And I also just. I changed the folder from being called the Claude folder to just AI Projects because I'm like, sorry, Claude, you're not the one running my computer anymore. And so I can point OpenAI and Claude both at the same folder and they both can edit the same stuff. So I can interchange them back and forth by run out of credits on one, I switch the other, it goes back and forth and the same folder is being updated and they're not.
- [10:14]
B
I mean, that's a great point. If you're. If you're coding on a folder on your computer, why can't both of them access it? And you know, you have to interact with it yourself. If you're putting, like, API keys in and stuff, you have to put them in the envelope or env. Whatever it is. But yeah, that's a great point. What about. If you have. I guess you would just connect it to your GitHub as well. Then it would. And it would have access to all that stuff too.
- [10:40]
C
Exactly. Like Claude, it just has connections to Vercel, GitHub, Supabase. You get MCP for anything and plug it all in. Like I put my AI box MCP in there. So like I literally get everything that I had on Claude is on here. Plus there's nicer folders and plus it doesn't boss you around. And plus it use half as many tokens. And plus it has this cool new feature that just popped up right now that says introducing Fast mode. Based on your work in the last seven tasks, fast could have saved you 1 hour and 13 minutes, increases your plan usage. And so you could just click Enable. And now it's going to go into fast mode. Now you can switch it off of fast mode, I think smarter, faster, you have a little toggle that you can do. But anyways, it's trying to optimize what I'm working on. It knows what I'm working on. Anyways, I just freaking love it. And the last thing I want to say is it has this really cool visual component. So I actually went. I've had to do a whole bunch of audits lately of a bug that I had on a software and if you're watching this, I was like, hey, I have this problem. I explained my problem I was having with my code base and it's like we did an audit and this is what, what's like causing your problem. And it drew a full like, like diagram of how my code base all links together where the bug is and what was causing the issues. I've never seen a visual of what my code base looks like because I'm just a dumb non developer. I just yell at the Void with Claude and tell it to do stuff. I don't know where my code base is. I don't know what line of code has the issue. I don't know why. I think I just like, hey, it's broken, it's broken, it's broken. I could actually see where the problem is now and address it directly. And if I wanted to learn more about getting into the code, it has this really cool thing as you scroll down where it's like critical findings. See line 837 and if I click that, it literally can open up my file, which isn't connected right now, but it can literally open up my file and show me exactly where in my code base the bug is.
- [12:37]
B
That's crazy.
- [12:38]
C
Really, really cool.
- [12:39]
B
That's amazing. Also guys, just news. This is not sponsored by.
- [12:44]
C
This is highly sponsored. Sam Altman himself has paid me $1 million to relentlessly shill his new app.
- [12:52]
B
But if you were trying to sell me on it on the switch, I might be convinced because that's all very compelling.
- [12:59]
C
I have been the biggest Claude Stan for like six months now, but officially they got to pass the torch on. And here's the thing with these models. It's like. Or these companies, OpenAI really cooked because they saw that Anthropic was eating their lunch for a long time, like six months, and they kept coming, making updates and coming out with stuff and putting stuff together, and it just finally reached a tipping point where they had to build so many better features, consolidate all of their applications, and come up with a brand new model that beat it. And when all of those things were right, then I'm like, okay, the software is better, the tools are better, the model is better, the speed is better, the efficiency is better. Okay, I'll go over here. And inevitably Anthropic is going to do the same thing in like, I don't know, maybe six months. I don't know how long it'll take Anthropic to catch up because I would say opening eye is like 30% ahead of them. Like, my experience here is 30% better, which is for me, that's pretty good, right?
- [13:52]
B
That's really good. Yeah.
- [13:54]
C
How. What will it take for Anthropic to catch up on the 30% and then surpass by 30% for me to have to switch over again? Right. Because, like, I'm not going to go for 5 or 10% better. Like, it needs to be a significantly better product. We'll have to see what they do, but might be six months before I'm back on Anthropic. I like having the competition and feeling like I can go back and forth and everything's interchangeable and they're like, fiercely competing. But yeah, right now it feels like OpenAI has. Has taken the cake.
- [14:20]
B
Nice. Cool. Hey, thanks for sharing. And if you guys want to hear more from Jaden and some of the projects he's working on and also some of the ones I'm working on, join our school community and you can get access to those. Find out how we're actually using these tools to make money online and grow our businesses. Thanks for listening and we'll see you next time.
- [14:40]
A
Grainger knows when you're a procurement manager for an office park, you're not managing one building, you're managing all of them. And to stay ahead, you need to see through walls and around corners, lights about to fail. Filters ready to clog H Vac on its last leg. If you wait until something breaks, you're already behind. Count on Grainger for quality products, easy reordering, and 247 support. Call 1-800-Grainger. Click grainger.com or just stop by Grainger for the ones who get it done. You're listening to a podcast right now. Driving, working out, walking the dog. If you're into podcasts, chances are you have something to say, too. With RSS.com, starting your own is free and easy. Upload an episode, and we distribute it to Apple Podcasts, Spotify, Amazon Music, and hundreds more. Track your listeners, see where they're from, and start earning from ads like this. Even with just 10 listeners a month, if you've been thinking about starting a podcast, this is your sign. Start free at rss. Com.