EP 416: Breaking - OpenAI’s New o1 and ChatGPT Pro - Revolution or Rip-Off?
2 Jan 2026 · 0 min
Jordan Wilson (Everyday AI) reviews OpenAI's announcement (Dec 6, 2024) of new reasoning-class models in the O1 family (O1 Mini, O1, and O1 Pro) and a new ChatGPT Pro subscription tier. Key points: OpenAI markets O1 as a reasoning model (distinct from GPT-series) and claims a ~50% performance increase; ChatGPT Pro is a $200/month plan that provides unlimited access to O1, O1 Pro, O1 Mini, GPT-4o, and advanced voice mode. Wilson demonstrates O1’s multimodal capability (PNG/JPEG uploads), runs live tests (prompt-based business use cases and an analysis of a 14,000-cell spreadsheet screenshot), and reports substantial quality and reliability gains in benchmarks (host quotes competition-math scores and increased "four-for-four" reliability). He frames the release in business terms (who should consider $200/month), notes uncertainty on API/enterprise rollout, and raises questions about accessibility and tiering.
Questions this episode answers
$200 per month
$20 per month (ChatGPT Plus)
O1 Preview: 50; O1 (new): 78; O1 Pro: 86
PNG and JPEG (images only, PDFs/spreadsheets not directly accepted as files)
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.
Open AI has just released a new model. That's right, the new open AI01 and 01 Pro. But that's not it. There's more. We also have Chat GPT Pro, and it comes at a price, both on the wallet at $200 a month. But also, I think this is really going to impact businesses. And is this separation needed? Is there now an elite club? Right. I think that one of the exciting things over the past two years since Chat GBT was first released was that it democratized access to large language models for everyone. But $20 a month for the world's most capable model I think is reasonable for most people, small business owners, and everyone. But 200amonth for chat gbt pro and for the new OpenAI 01 and 01 pro models. Is this a revolution or is it a ripoff? Well, we're gonna at least start to dig into that. All right, what's going on, y'? All? My name is Jordan Wilson, and welcome to Everyday AI. This is your daily livestream, podcast, and free daily newsletter helping everyday people like you and me learn and leverage generative AI to grow our companies and our careers. If you're tuning in live, you might be like, wait, what's going on? Yeah, double dose today. Yeah, we already had a podcast and live stream today. Our newsletter already shipped out the door. But speaking of ship, it is Open AI's ship. Miss the 1212 days of Open AI. It started today. And, you know, we knew that they were probably going to release something big on day one, and they did not disappoint OpenAI. So, yes, this is our second podcast of the day. So thank you, everyone. Who's. Who's tuning in for the evening edition? Absolutely. Peter and Tara and Nathan and Douglas and Dr. Harvey. Thanks for tuning in. So let's jump straight into it, y'. All. Let's talk about what's new, what this 12 Days of OpenAI even is, and we'll talk a little bit about this new 01. It is a new model. So we had 01 Preview and O1 Mini. Now we have the O1 Full model, as well as an O1 Pro model, as well as the new Chat GPT Pro, which is $200 a month, which go back a year ago. I said that we were going to see a version of this that was going to be multiple hundred dollars a month. And that I would pay for it. And I'm going to. Right. But we're going to take a look at all of this a little bit here live. So let's get to it. And hey, podcast audience appreciate especially if you listen twice today. So thanks, thanks for tuning in, but I'm going to try to do my best to describe anything that we're going to do. We're going to jump in and test out the new Full 01 model as well. So let's talk a little bit about what was announced today. So CEO Sam Altman and some engineers at OpenAI did a live stream. So that's essentially what's going to be happening over the next 11 now weekdays. So at 10am Pacific Time it's you know, for my Chicago people, uh, noon. So uh, you know, we'll be uh, probably sending out our newsletter a little late every day, probably right after noon. So if you want to see what comes out, it should be. Speaking of shipping, it should be shipping out then. So CEO Sam Altman and crew today announced live what was new with OpenAI in their first 12 days of Christmas release. So here's what it is. It is ChatGPT Pro 200amonth unlimited use. That is the big thing, right? So no more, no more messaging caps or anything like that. And let's talk about the models. Let's talk a little bit about the model. So OpenAI claims that the new O1 is the smartest model in the world. We'll wait to see all the third party benchmarks, but what OpenAI has released, it is bonkers. Good. So we'll wait for all the third party benchmarks just to see how big of a gap is it or will it be between this new Zero1 model and everyone else in the world. So we've seen a lot of, you know, of the Chinese startups in the last week, you know, with Quen. Qwq Alibaba, all the big kind of large language model and AI companies in China are, are also releasing reasoning models as well. So that's another important thing to know. We this is not GPT5. I've been on the record saying we're going to get a GPT 4.5 before we get a GPT 5. That is not it. So the O1 models are a different class. They are not a GPT class, they are a reasoning model. What that means, if you haven't heard of it, maybe you're brand new, but there was essentially some reports more than a year ago about something called Q, right? And then that Q Star project, that this was a code named, then it was called Strawberry, and we ultimately learned it was the O1 model in the O1 series. So when OpenAI did release this, it was 01 preview, which was kind of like a lite version, and then 01 mini, which was a smaller, faster, cheaper version. So I should say this, this is all live now, right? So you can go use the complete Zero1 model. You can use the new Zero1 Pro model, which we will be talking about, as well as the new Chat GPT Pro plan, that $200 a month plan, which essentially lifts all the limits off of everything. So it lifts the limits off of the 4O model. It lifts the limits off of, obviously, the 01 models, where normally you would have a certain number of messages a week or on the 01 preview, the 01 Mini had a higher limit per day. So a lot going on here. But let's just talk about the model. So OpenAI says that we are seeing a 50% performance increase. All right, so let's go over. So now just reading straight off of OpenAI's website and, you know, we're going to give it a little, a little poke here at the end, but we're going to be covering this new model in the coming weeks, I'm sure, as we get more time to go super in depth. So here's what we're seeing. OpenAI says today we're adding Chat GPT Pro, a 200 monthly plan that enables scaled access to the best of OpenAI's models and tools. This plan includes unlimited access to our smartest model, OpenAI 01, as well as O1 mini, GPT4O and advanced voice mode. It also includes O1 Pro mode. Okay, so there's not a lot of details on this new O1 Pro. So there's technically now three models of O1. There's 01 Mini, there's 01, which used to be called 01 Preview. So 01 Preview is gone. So it's 01 Mini, 01 and then 01 Pro. But you have to have that 200amonth plan to use the 01 Pro. Although this morning it was popping up for me before the announcement, now it's gone. All right, so what OpenAI says, it says it. So it also includes 01 Pro mode, a version of O1 that uses more computer to think harder and provide even better answers to the hardest problems. So, yeah, not a ton of information yet, aside from benchmarks, I'm sure that we'll be seeing and hearing more from OpenAI on what actually differentiates this new 01 and the 01 Pro. But for now, they're essentially just saying, yo, you get more electricity, more or less, right? You get more brains, you get more electricity and it's better, right? And hey, look at the benchmarks, it's better. And then you have to think, is it actually a different model or is there some throttling going on, right, in the normal kind of like quote unquote, 01. So, you know, not sure. I'm sure we'll get a lot more information, but essentially they're saying we devote more compute to, to O1 Pro to think harder and work better, essentially. Right? So let's look at the difference here. So if you're listening on the podcast, it's, it's a graph, right? We're looking at graphs here and we're looking at Zero1 Preview, right? Which is now not the default Oh1 model, but it was released in September. It was the most powerful model in the world until today. So now that is 01. So if you are on the normal Chat GPT paid plan, that $20 a month ChatGPT plus, or if you're on ChatGPT Teams, you have access to the O1 model. And they're saying the O1 Pro, so that's the one you have to pay 200amonth for. It performs much in challenging machine learning benchmarks across math, science and coding, right? So if you are an analyst, if you're a researcher, if you're a software developer, if you are just a power user, you might want to look at that 200amonth plan, right? Sounds, it sounds crazy to say that out loud, right? Because we've been absolutely spoiled humans in the world. We've been spoiled for the last, you know, year and a half since there was a paid plan of Chat GPT and then Google Bard, which became Google Gemini, and then we had Claude Anthropic, Microsoft Copilot. All these other models, we've been spoiled, right? I still don't think people, number one, people don't care to learn about large language models, how they work, all that, right? But to have the amount of power into now, how much a single person can accomplish for $20 a month, it's, it's nuttier than a squirrel on keto, right? So now you have to think, is this a good thing, right? This new tier, it's almost like the, there's an elite club now. Whereas before, I think generative AI large language models, it, it became this place for everyone, right? And you could almost feel this, this energy and you know, almost inclusiveness of, you know, kind of leveling the playing field. And that's something even for me personally, I loved about generative AI and large language models. Early on I said anyone can do this. Literally anyone can scrap up, you know, almost anyone can scrap up, you know, $20 a month and it pays for itself immediately. Now we have tiers, right? We have the elite club. We have not everyone, not everyone can afford, you know, $200 a month for this new chat GPT Pro. And I, I haven't yet seen because again, y', all, this is just hours in, hours, you know, new or hours old, right? I, I, I haven't seen a lot of details yet on teams accounts, enterprise accounts, what the pricing, if it's even available, what it will be as we figure that out, we'll let you know. But anyways, benchmarks, I mean, O1 Pro is much, much higher, right? So in math, right? So we go from 01 preview, which 12 hours ago was the world's most powerful model, you got a 50 on competition math, which is the AME AIME 2024 benchmark. So we got a 50, right? So now the new 01 model 78 and 01 Pro 86. Yeah. So if you are in any, any, any role related to business intelligence, anything with, with stats, research, analytics, that what I just said right there and then, I mean, we're going to get into coding and science research, right? It is such a large gap between the O1 preview model that we, that we all had to the O1 model now in the O1 Pro. Also, I should mention OpenAI did say the API for this should be released soon. So at least as of right now, this isn't on the API. And so what that means if you're a newbie, don't worry. Everyday AI, it's for everyday people, what that means, right? There's probably, and you don't even know probably so many of the different softwares, apps, enterprise tools that you use, websites, right? A lot of them are powered by this technology under the hood, right? So so many thousands of enterprise companies, software apps, et cetera, use OpenAI's technology. They use their API. So right now, if you really want a taste of this and if you want to, you know, figure out for yourself if this is a revolution or ripoff, you're going to have to log into Chat GPT. This isn't available for free users as of now. I'm sure they're going to give everyone a little taste, right? Give everyone a gateway, little gateway drug here of 01 and 01 Pro. Maybe. But I mean the competition coding, so code forces another metric here. 01 Preview 62:0189. 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, 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. All right. I mean we are talking about 30% plus gains overnight, right? So a lot of these instances where smart business leaders would look and they'd be like, ah, you know, still not there. Well, now you have to think about it, but you also have to think, is it worth that 200amonth price tag? All right, another thing, and again, this is all chat GPT. We got to do our own kind of a very unofficial internal testing. So another thing that they're talking about is improved reliability. So they're saying that they used in their evaluation, which there was a whole model card. Right. So we'll share that in the newsletter, you know, today or if you're listening live tomorrow. So we will share this. But they're also reportedly improving the reliability. So what they said here on their page is to highlight the main strengths of O1 Pro mode. They use a stricter evaluation setting. A model is only considered to solve a question if it gets the answer right. In four, four out of four attempts. Four out of four reliability, not just one. So in kind of even benchmarking their own internal models, you see this huge, huge improvement. So, right, like we talked about that competition math doubled the reliability from 01 preview to O1 Pro. Right? That's, that's the only Thing people don't really understand about large language models, they're generative, right? It's rolling the dice, especially if you're not good at, you know, prompting or if you don't know how models work, you're even rolling the dice even more so. Right. If you know the proper prompt engineering year, you have a little bit more control of those die. But I mean, the reliability scores here are off the charts. All right, so then let's quickly, let's quickly look at the pricing, all right? And then we're going to run, we're going to run a little test here. Don't worry. So $200 a month, is that worth it? Is it worth it for you? Is it worth it for me? Is it worth it for my company? Is it worth it for your department? Is this where we should be looking now? Right? Is this what we're going to see in the future, y'? All? I'm literally on record I said last year that eventually there would be more premium versions of large language models that would cost thousands of dollars a month. That's a good thing about having a daily podcast. You know, got receipts, receipts for days and days, right? And I still think that's coming. Right? So, you know, a lot of people are, are talking about this on, online, on the Twitter machine, on Reddit, right? And they're saying, oh, OpenAI is just doing this because they're reportedly losing $5 billion a month and they need to get profitable. Well, yes, right, and, and anyone that follows the space has known all along that especially OpenAI was burning through cash. Right? If you go try to use Claude Anthropic at the same rate, you can't. It is so limited. Right? So OpenAI has been very generous and I think that's one of the reasons why they are literally crushing everyone else, right? Because for $20 a month, you get access to the world's most powerful models with the best features. And it's. Even though there's limits, it's very hard to hit those limits. So now with this pro plan, it is unlimited access again. That's 200 per month for one user. So you do get everything else, but you get unlimited access to GPT4O and 01. Right. So I haven't personally jumped on the $200 a month Pro plan. I will just been, been a little bit busy today. But, you know, we will be upgrading to that plan. Our team here at Everyday AI. We will. But one of the things that I think is great is unlimited voice mode. All right. Because the Advanced voice mode from OpenAI is so sneaky good. It's so sneaky good, right? And if you haven't used it before, I encourage you to, you do have to have at least the $20 a month plus plan to use that. But that's something that not a lot of people talked about. I'm like, yo, like, the rates are different. I, I, I think it's been about, I think you get like, originally it was 24 or 2020 to 25 minutes. When it first came out, I think they expanded it to 60 minutes, but that's a huge upgrade there as well. All right, so let's, let's jump in. We're gonna do a super, super unofficial, unofficial test here, all right? And we're gonna see, we're just gonna figure out if this is worth it or not. So I'm jumping into chat GPT01 again, not GPT01. I'm gonna run this little prompt and we're gonna, we're gonna look. So is, is this any good? I don't know. Let's, let's find out, right? So I'm saying, please give me exact and specific examples of what OpenAI's O1 model can do for someone's business. Be short and succinct, yet very detailed. Do not waste words, yada, yada, yada. And I'm saying, you know, give me five examples and in doing so, tell me how a subject matter expert would tackle these same tasks, projects, or challenges, and then how the O1 model could do it. All right? And I'm saying be short, be direct. Let's take a look, right? So let's take a look because everyone's, everyone's always talking about use cases, all right? And so when we think about it, the, the O1 model, it is a reasoning model, okay? So to simplify that, in theory, you're not getting something that wasn't possible before. With Chat GPT's 4.0, it would just require to get the same outputs some of the smartest and most talented humans in the world, right? Because, you know, if you think of like kind of quote unquote, prompt engineering, it's really steering the world's intelligence at the right place with the right skill set in the right way, right? So technically, nothing is, is different in terms of what it can achieve, but by having a model that can use logic and reasoning, right? It's, it's literally if you get the smartest human in the world, give them GPT4O, they can probably get similar results it would just maybe take them many hours and probably only a handful of people in the world could even perform at the same level. So the O1 model, it does that all for you. It uses this kind of chain of thought reasoning under the hood. So I like to tell people like this, if you're in the kind of the, the default version of chat GPT plus GPT4O, right? You, it's, it's kind of like you can get to the destination, but you have to help Chat GPT build the car, you have to tell it which road to take. You have to examine, right? Oh, is this the right route? Right. So you still have to be very involved. It's like you're, you're building the car, you're, you're paving the roads, right? But again, capabilities five years ago we never thought we would have with that baseline model. But with the reasoning model and now with this new updated version of it, you tell it the destination and then this reasoning or kind of chain of thought processing does it all for you, right? It does a lot of those, you know, best practice prompt engineering for you under the hood. So if you haven't used the O1 model before, and this is a little bit different user interface with this, this newest version. So when I asked this, right, So I said, hey, tell me, you know, five specific things that OpenAI would be great or, sorry, the O1 model would be great at, and then tell me how a normal, very smart human would tackle the same things. So it said, thought about O1 model applications for seven seconds. So I can click that. Okay. And then on my screen here I can go see kind of how the model handled this. So it said highlighting. So the, the bullets here says highlighting capabilities, clarifying business impact, highlighting precise tasks, crafting marketing campaigns and transforming tasks. Right? So you can literally see how it's thinking, how it's piecing it together and it's literally going through step by step. Right? So let's, let's, let's quickly see what are the business values and are these things that your company could use? All right, so let me, let me see, let me see something that hopefully. Here we go. I'm not going to read all five, but I'm going to read just number two here. So it says producing accurate sales forecast reports from raw data. So it says a subject matter expert has to manually extract historical sales figures from spreadsheets, identify seasonal patterns, apply statistical models and write a thorough analysis report. A process that could take days, right? A1 model, all right, it rapidly processes Upload historical sales data, identifies trends, runs predictive modeling, and outputs a polished data backed forecast report with charts in minutes or maybe barely a minute. All right, so I'm not going to go through. I'll at least read all five examples, but not the details. So number one, it said generating highly specific marketing copy for niche audiences. Number two, producing accurate sales forecast reports from raw data. Number three, conducting detailed code reviews for complex software. Number four, summarizing and analyzing customer support tickets at scale. And then number five, checking marketing materials against regulatory guidelines. All right, so I'm going to do another, another quick little test here. And hey, live stream audience, let me know if you have any questions. I, I've played around with the model, but let me know if you have any questions. And thanks. Thanks everyone for tuning in twice in one day. Love to see it. All right, so I'm going to do one other thing, y', all, because there's another feature that I haven't even Talked about in 01 that was not previous before. So not only did we get the full 01 model, if you pay for the $200, you get the 01 Pro, but there is one other nice little feature that's available that was not previously available. So let's actually take a look here. All right, so you can upload photos, all right? So you unfortunately can't upload PDFs, you can't upload spreadsheets, but you can upload PNGs and JPEGs, right? So I can do probably something, you know, kind of fun and cutesy and you know, we do some standard tests that we do here on the everyday AI show pretty often. But I'm going to do something else, right? Because I want this, I want to be able to upload because I think one of the best things, one of the best things that you can use a large language model for right now is analyzing data, right? Because the availability of data has made people that are not data analysts all of a sudden. You got to make use of it, right? Just think of like 15 years ago who put together sales decks and presentations, right? It was probably a graphic designer. Guess who puts them together now? Literally everyone. And I think being able to tell stories with data is everyone's job now because data is available everywhere and we can use it and leverage it and make sense of it in seconds. All right, so I'm going to try a little trick, y'. All. I haven't done this. So if anyone is. If you're looking at my screen on the live screen, this is Chaos, right? So we, we, we do this on our YouTube channel a little, a little bit. But I just export a ton of data here out of our YouTube channel. It's, it's video titles, views, subscribers gained on each, video impressions, click through rates, shares, right? All these just thousands of data points. So right now I can copy and paste this in, which I'm actually going to try. But here's a little tip that most people aren't going to tell you. All right, Zoom out a ton. Actually, I'm gonna, I'm gonna put this to the test. All right, So I just took a screenshot, all right. And then I gotta find where my chat GBT went. All right, so now I have a screenshot with a bunch of data, right? You can barely read it. I zoomed all the way out like it would take a human, like a friggin telescope and a lot of time just to be able to see this data. I want to see, okay, how are its computer vision capabilities? This, if it does this, this is an extremely difficult task. All right, so here's what I'm going to say. Luckily I typed this out so you didn't have to watch me type this live. So I'm going to go ahead and click enter and I'm saying based on this breakdown, please give me 20 in depth insights that can help me grow the YouTube channel. Be pins, be pinpoint specific drawing on between the line insights and advanced tactics. For each insight, give me actual stats or examples from this spreadsheet. Also give me step by step directions on how to implement all 20 of these insights. Make sure they are specific to the content I've uploaded and do not suggest anything general. Only focus on the results, trends and statistics. Statistical insights you can glean from the data in this screenshot. All right, so it thought for a whopping 18 seconds. So what I will say is sometimes when you click it and you want to see what's going on under the hood, and I encourage you to do that, it's going to make you better at AI, it's going to make you better at your job. So sometimes it says no details available, sometimes it says, oh, there we go. It says occasionally the model generates the best response without anything to summarize. Right? So it fought for 18 seconds before replying and then it says, below are 20 pinpoint insights drawn directly from the data patterns visible on the spreadsheet. And here we go. Right? So even though you can't upload spreadsheets, which I think people are going to love, once you get that capability. Because before today, when working with the O1 model, you couldn't upload anything. So today you can at least upload images, right? So that leaky, leaky pipe, you want to, you know, proof, prove you're a handyman or handy woman. Go ahead, upload it now to the O1 model, right? Any, any photo image. If you're trying to, you know, read something in a different language, a complex problem. If there's, you know, something in a deck or a sales presentation that you don't understand and you can, you know, snap a photo of it, a meme that's going over your head, right? You can use the L1 model now and you can upload a photo. But what you, what people don't know is, well, just take a photo of a spreadsheet. It actually does an extremely impressive job. And I'm, I'm double checking the data here because I do want to make sure this is correct. And it is correct, right. At least on, on this one. I'm going to go through and check it later. But by, by first kind of count, it doesn't look like we have any, any hallucinations. So it's giving me some good stuff, right? So look at some of these. I'm not going to read them all. Let's see. Okay. Okay, here's, here's a good thing, right? So on number four, it said, turn high Engagement shorts into. Let me make sure I'm sharing my right screen here. There we go. So number four, it says, turn high engagement shorts, like metrics into a series. And then it says Data point perplexity, AI overview and what's the best PDF plugin. So those are two different YouTube video titles, I guess, in this screenshot from the spreadsheet. And it says, have viewed versus swiped away rates of 182% and 200% respectively. So it actually went through and created its own metric by combining two different metrics. That metric does not exist. So it combined two different metrics. And then it said, it implies a hook strong enough to capture viewers beyond a single impression, possibly shorts or very short form content. Then it says, so it gave me a data point. It said why it matters. So it says exceptional viewed versus swiped metrics indicate you're grabbing and holding attention quickly, something crucial in short form formats. That's so freaking impressive, right? It literally, I didn't give it any information. I didn't say, hey, you know, this is a YouTube channel. Here's a, this is a screenshot of metrics. I. Nope, I, I Uploaded a photo. I gave it a very specific task. It, it's literally going through there in 18 seconds. And it not only quickly analyzed everything, right, it converted a screenshot where the fonts like seemingly like two, because I'm trying, I'm doing a little work around here. It literally understood it that quickly and started to create its own metrics based on the data that it observed to then create a plan. All right, so then it gave me the data point why it matters and then how to implement. How to implement. So it says, one, identify if these videos are short form content. If yes, because I didn't get all of the stats in that screenshot. And then it says, if yes, produce a mini series of similar short form explainers. Two, use identical opening hooks from these top performers, such as starting with a pressing question or a bold claim. And then it says three, compare the viewed versus swipe map metrics over two to three new uploads with the same hook style to confirm improvement. It's freaking good. That's good, right? Like I've been in different forms of marketing, advertising, communications for 20 years. That is good stuff, right? I've hired expensive consultants who would have to take 18 days to come up with that kind of thing. Right? Essentially creating your own metric or your own way to measure something and then giving me step by step ways to pull play it out. Friggin amazing. Love to see it. So, all right, I see, I see a couple questions. I'm gonna try to get them here at the very end. So what do you guys think of this? You know, and if you're listening on the podcast, we do always put our contact information. Let me know if you do want to see more content around this new 01 model. Once we get our hands, hands a little dirty with the O1 Pro, I'm sure we'll do a show or two on this, but it's pretty impressive. So now what I'm going to do. So I'm back in this spreadsheet, right? There's a ton of data. It, I mean this is, this is a. I think it's like 13,000. It's like 13,000 data points. So I don't even know how much I can upload. I didn't see context on this yet to see how much I can just copy and paste in there. Right? So let's see, let's see if we can break 01 quick. Right? All right, so let's go in and I'm just gonna say, I don't know, give me 10 trends from this data, I, I doubt I'm gonna be able to paste in like 13. Let's see how many. This is exactly 14, 000 cells of information. Okay. Looks like it at least took it. I didn't break it. All right, so There we go. 14, 000 cells of information. I just said, tell me 10 trends. Okay? And there we're done. It didn't even need to tell me any details of how it thought. So that's, that's, that's pretty impressive, y'. All. And I'm, I'm just double checking some things, right? So it says, okay, we compare them all 18, 111 shares on what? On the best PDF plugin. So I'm just double checking this because I don't want to, you know, be giving information that's not correct. So I'm checking over here on my other monitor, and surely enough, that video has 111 shares. So it's, it's not like it just made up a bunch of information. It literally, literally live here gobbled up 14,000 cells of data and gave me 10 insightful trends in seconds. In seconds, yeah. Brian saying, I bet the consultants you hired cost more than 200amonth, though. Yeah, that's a good point, Brian. And I, I think that brings me to, you know, as we kind of wrap up here, that brings me to the question now, is this right for your business? Is this a revolution or a ripoff? Right. There's no one answer. I don't think I will say I'm going to reserve my judgment until I go through the pro. But this new model, off the bat, very impressive. Very impressive. Right. And, and there's. There's a. I think there's a lot to still explore. I'm impressed, personally. All right. Gareth asking Jordan, do you think this new large pricing hike is to help show that they can get close to being more profitable as a company or a sign of other financial issues? I can't, I can't speak to that. I don't have any inside intel on that. Right. I, I do hear from people at Open AI that say, you know, oh, they listen to the show and all that, so. So thanks to them for their support. I don't have any intel on that. My. My own take is, I mean, this has to be right. So OpenAI just raised, I, I think $7 billion a couple of months ago. I mean, this has to be part of their plan to profitability. Right? It makes sense that they were, you know, losing billions of dollars reportedly every single year. I said this all along, right? I said eventually companies, especially startups, it's different for Google and Microsoft and Amazon, right? Amazon with their very impressive Nova model which we haven't even talked about I think because it's so hard to use. But I said very early on I said Anthropic and OpenAI are going to have to choose. You can either burn through money, right, and give people everything or you have to be extremely restrictive, right, and worry about profit right away. Profit over users. It seems to me OpenAI said hey, let's just capture the market. They were first there, they were crushing everyone because there was no competition right in November of 2022. But you know, as they started to see competition from, from Google, from Anthropic, from Mistral, from Mata, right they just kept just busting everyone over the head with the best models still for so cheap. So I'm not sure but you know, I do believe that a big part of their long term strategy was capture the market first and then offer premium products later. Nathan saying, do you think the $20 will become the old free version with dial back capabilities compared to what is currently available on the $20? No, I don't think so because there's other things, right? This whole 12 days of ship, ship miss right. You know, open. So we're going to be getting a lot of releases. So I'm not going to put out my whole line of predictions but I've been saying all along sometime after the US election to the first quarter of 2025, we're going to get a GPT 4.5. I personally think that's what we're going to see over the Next, you know, 11 days of this little series. I think we're probably going to see Sora Sora released at least two. Maybe it's just the Chat GPT Pro. I'm not sure. I, I do also think that we will eventually see these tools start to be rolled out to other modes. Here's what I mean by that. I just showed you all. 01. You can upload photos now, right? That's an example of a quote unquote tool but it doesn't have use of all the tools. Advanced voice mode, web browsing, you know, whether you want to call it chat GPT search or browse with Bing, it doesn't have all of those things. The same with advanced voice mode. So I don't necessarily think the 20amonth tier is, is going to get stripped away from. I just think you're going to see all these newer features right like oh it as soon as advanced voice mode you can have the ability to upload files and it can browse the Internet literally. Game over for like 80% of medium sized consulting businesses if they don't pivot hard. Because then I can instantly upload all my information, get real time access to literally. And if you know how to speak to a model, that's. That's bananas. That's bananas, right? So I don't necessarily think that the 20amonth version is, is going to get worse. All right y', all, that was a lot. But to recap a lot of new stuff coming from OpenAI. So they just kicked off their kind of 12 day event with a new version of O1 and a 200amonth new tier coming in with the chat GPT Pro which gives unlimited usage as well as access to that 01 Pro model. So I hope this was helpful. Thanks to our live stream audience who joined us twice in a day. Y' all are amazing. Appreciate the support. I'd also appreciate it if you tell someone about this. If this was helpful, y', all, please click that little repost button. Right. It's, it's the holiday season. I don't ask you for a $trillion or $7 billion of fundraising. I just ask you share this with your network. We put in a lot of time to make sure you stay up to date with the realest news, the realest thing in artificial intelligence. Please repost, share this with your friends and if you haven't already, please go to your everyday AI.com Sign up for the free daily newsletter every single day, Monday through Friday. We not just do this podcast in this live stream, but we also recap the the show for the day in the newsletter. So thank you for tuning in. Hope to see you back tomorrow and everyday 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 your everydayai.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.
More from LARRUZ


Ep 882: Rogue AI Agents: Why Breakouts are Happening More and How Companies Should Prepare (Replay)


Ep 881: Faster AI Agents, Fewer Human Coworkers: The Overly Productive Future of Managing Agents? (Replay)


Ep 879: New: OpenAI Presence. Has The AI Customer Service Takeover Finally Arrived? (Replay)




