EP 240: AI News That Matters - April 1st, 2024

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. Is Microsoft really working on a 100 plus billion dollar AI project? And what's going on with Apple? And you probably won't believe this new OpenAI voice cloning technology. We're going to be talking about that today and more on everyday AI. What's going on, y'? All. Thanks for joining. If you're new here, let me tell you what this is. My name is Jordan Wilson and I'm the host of Everyday AI. We're your daily live stream podcasts and free daily newsletter helping everyday people learn and leverage generative AI. So if you haven't already, make sure you go to our website@your everydayai.com Sign up for our free daily newsletter. And we've had the best and largest resource for free generative AI education on the Internet on our website now more than 230 episodes covering every category. No matter what you want to learn in generative AI, you, Your EverydayAI.com is the place to do that. All right, so thanks for joining us live for our live audience. You could start your week however you want to, but people like Tara and Brian and Mike and Juan, they're choosing to start their week with everyday AI because we do this usually every single Monday and we bring you the news that matters because there is so much AI news, you could spend hours every single day trying to keep up and still be like, what the heck is going on? So let us do all the hard work for you. We do this, like I said, almost every Monday. We cut through the fluff, dissect the marketing from big tech and we say, hey, here's what's actually going on and here's what actually matters. All right, so regardless of where you're joining us, whether it's South Florida like Rolando or at the airport, Douglas, thanks. Appreciate that. Let's get into the AI news that matters for the week of April 1st. I'm not, I'm not one of those, I'm not playing any tricks on you for, for April Fools. Let's just get straight into the AI news. All right, so Apple is set to unveil its AI strategy at the June 10th developer conference. So Apple is set to reveal its long anticipated artificial intelligence strategy at its annual worldwide developer conference. So wwdc starting on 6-10-10. So now we have some official reporting on the start date as well. And it is expected to heavily center around AI technology. So while Apple itself has not officially disclosed specific details on the unveiling, a new Bloomberg report shows that there will be significant emphasis on AI. We've been talking about this for months, but as some of the new reporting comes out, we're definitely going to talk about it every Monday on the news that matters. But the biggest reported news is that Apple will not be debuting its own large language model reportedly. And Apple, according to other reports, was spending millions with an millions of dollars a day working on various AI models internally. And instead of debuting an internal model, Apple will reportedly team up with either OpenAI or Google. And most reports are suggesting it will be the latter in working with Google and use their AI models in Apple's next iPhone. So here's some some more information from the Bloomberg report showing some of the new AI features what we might be expecting in Apple's next hardware device. So Those features an AI powered version of Siri AI integrated into the Message app and as well as AI in Apple's iWork apps and that's your documents, spreadsheets, key keynote, etc so this move by Apple signifies a strategic shift toward integrating AI into their products and services, potentially impacting the future direction of the company's tech offerings. You know and hey for, for our live stream audience, I'd love to hear from you what you think of Apple's moves. You know obviously Samsung has already brought edge AI or you know, large language models to to their latest phone in the S24. I think this is extremely powerful. We already know that on device AI or edge AI is the future of how we will all interact with AI. But yeah, like Douglas is saying here, given the pace of AI announcements, is Apple too late to the party? Douglas hey, my, my two cents is no, right? I've actually been saying this for probably about eight months that Apple will definitely be last to the the party but they are usually the coolest person there. You know, kind of. I think what's worth noting in all this reporting is, you know, about six months ago we were hearing that Apple is spending millions of dollars a day on working on its own internal AI. We heard about a month ago and we always share all of this in our newsletter so make sure you're reading that every day as well. Right? It's not just Monday we dive into the news, but we talked about two months ago how Apple was shutting down its car division and shifting a lot of those resources toward AI. So you do have to think that this will be a huge AI announcement from Apple. But again Reportedly in lieu of working on or debuting its own model, it seems like they're going to be partnering with either OpenAI or Google to use either OpenAI's GPT or Google's Gemini. Personally, I would love to see them go with GPT, but it does seem like, according to the most recent reportings, that they are leaning toward using the Gemini model. And again, nothing here is confirmed. Although, you know, this showing a tweet here from Greg Josewak, the VP of Marketing for Apple, kind of goes, it doesn't exactly say it, but kind of teasing it here. In a tweet that just came out, he said, mark your calendars for WWDC 24, June 10th to 14th. It is going to be absolutely incredible. And he capitalized the A and absolutely. And the I and incredible. You know, it's no, no secret there that that is a, a nod to the big AI announcements that they have coming. So they, that actually might be the, the first, I guess super official hint at what is to come from Apple. But yeah, I, you know, to get back to Doug's initial kind of question here, I don't think they're too late, right? I don't think they're too late. I think, you know, you can tell from their, their earnings calls that they are working toward, you know, integrating artificial intelligence more into all of their hardware itself, software. You can tell by their acquisitions over the last 18 months. You can tell by, you know, a lot of these reports like the Bloomberg one that we're talking about here, that Apple is going all in with AI and there is no other way. But like everything else, Apple's never first, right? Apple's never first. There's always, I feel a joke amongst, you know, those Android users, Samsung users, etc. You know, that says, hey, you know, Apple users enjoy, you know, this new feature. We've had it for two years, right? So Apple obviously playing the, the Apple blueprint here, I won't say it's the, the safe route, but definitely the safer than most, right? Because all the other, you know, big tech names from, you know, Amazon and aws, Microsoft, Google, et cetera, they've already kind of planted their flag down, you know, with their flagship AI products. And all we have from Apple so far is rumors, reporting and rumbling. But I do think, and I've said this, go back, I'll try to find this, this episode. But I said probably seven or eight months ago that Apple is obviously not going to be first. They will probably be last the party, but they will be the Coolest kid there, right? And it will work. I think it'll work well. So definitely something to, to keep an eye on. Hey, great point from Juan here talking about the, the tweet kind of Gemini esque colors. Juan, I do agree, although it does just seem now that is the, that's just the new color scheme for AI for whatever reason. But great observation here from Juan talking about this tweet that we have on screen here from Apple's VP of marketing that it does appear to have some of those Gemini colors, you know, the cool blue and the, the pinkish orange. So great observation there. But yeah, it does. At least to me it does seem like this is just the, the color scheme of AI. I don't know why that's a thing, but apparently, you know, if, if you announce, you know something AI heavy, it seems like you're using those colors for, for whatever reason. All right, let's keep it going. More AI news that matters. A huge one. How did this one get bumped to the second slot this week? A hundred plus billion dollar partnership. But Microsoft and OpenAI are collaborating on a $100 billion plus data center project with their Stargate supercomputer. All right, let's break down what this is and talk about why it matters. So here's the facts first. So Microsoft and OpenAI are partnering on a groundbreaking project, project that could cost up to $100 billion featuring an AI supercomputer named Stargate that is set to launch in 2028. So about four years out. So Stargate could host millions of powerful AI chips. And this project aims to meet the surging demand for AI data centers capable of handling more advanced tasks than traditional centers, driven by the rapid adoption of generative AI technology top to bottom. Right? So real quick, I'm going to take a break from kind of my bullet points here and talk about why this actually matters. Right? So all the generative AI, so even what we talked about with, you know, whatever comes out in Apple's next iPhone, you know, set to be announced in June already, you know, we have the Samsung S24. You know, when we talk about, you know, different, you know, generative AI products all the time here, you know, Google, gemini, anthropic, Claude, chatgpt, you know, midjourney, OpenAI's Sora Runway, right? So all of these kind of software that, you know, hundreds of millions of people use, it needs powerful GPU chips, right? So, and right now there's not enough. Right. We've talked about some recent announcements from Nvidia and their new Blackwell chip might kind of change that and might give us the compute that we need. But right now, getting, getting back to details on this supercenter. So the, the proposed US based supercenter is part of a series of supercomputers planned over the next six years and is expected to be 100 times more costly than existing data centers. Yeah. Wow. So for context here, right, we just talked about Apple. Apple's campus reportedly cost about $6 billion and is the most expensive construction product in the U.S. so most expensive construction project. So 6 billion versus more than 100 billion. Right? Obviously a lot. The cost of that is associated with the reportedly millions of AI chips. All right, so people are talking about, okay, how is this even feasible, right? So Microsoft is likely to finance the project according to reports with significant costs focused on acquiring these AI chips which are crucial for AI computing and are often priced at premium rates right now. So obviously the cost of this could change depending on cost and availability for all these AI chips. But it's more than 100 billion, so expenses for the project could surpass 115 billion, which is more than triple Microsoft's capital spending on servers and infrastructure in the previous year. Yeah, that's wild, right? So the new project is designed to accommodate chips from various suppliers, right? So this isn't just the new Microsoft chips that they announced and it's emphasizing the importance of infrastructure innovation to achieve advanced AI capabilities. So this collaboration highlights the significant investments being made in AI infrastructure, signaling a major advancement in AI computing that could have far reaching implications for various industries and technological advancements. So after all that, yeah, I had to get out all the details there because this is, this is huge, right? So we're talking about a literal hundred plus billion dollar project between the actual infrastructure and, you know, all of the reported millions of kind of AI chips that would be housed there within. This is pretty big, y'. All. Like you've, if you listen to the show, you hear, you know, me and other guests talk about that all the time about the importance of compute, right? So it's no, no secret that compute is the next currency, right? When you have the most valuable companies in the world, look what they're doing with their money, right? They're investing all the biggest companies in the world. You know, even if we look here at the us, the magnificent seven, which are kind of the seven biggest tech companies, you, your Google, Microsoft, Meta, etc, they're all investing in compute power, right? Kind of the same way that, you know, we said data is, is gold for the last 10 years and oil and all these other things. I mean we're literally going to be looking at a gold rush for GPUs or whatever the, the future computing chip looks like that can help us achieve all of this generation generative AI. But I mean you cannot, I think, take this news lightly and you have to understand the importance of it. A 100 plus billion dollar kind of project that we're talking about here. And like we said, one of the most expensive kind of capital or construction projects here in the US to date is Apple's campus, which is 6 billion. So I'd love to hear what, what our live stream audience thinks about this, but I'll, I'll keep it short here so we can get to more AI news that matters. This won't be the last, I don't think this won't be the last $100 billion investment that we see. It won't, I would assume to see similar infrastructure projects that are just as big, probably from aws, maybe from Google, potentially from other, other companies as well, which we're going to talk about here in a second. A new kind of group of companies teaming up against Nvidia. But yeah, I'd love to hear, you know, even if you're listening on the podcast, I always put an email and you know, even my LinkedIn, I'd love to hear what you guys think about this, this new race and if compute is the new currency, seems to be what most smart people in the room are talking about. So I'd love to hear from you as well. All right, our next piece of AI news that matters. OpenAI has introduced its Voice Voice engine for text to voice generation. All right, so OpenAI has launched Voice Engine, a text to voice generation platform that creates synthetic voices based on very short voice clips. All right, so this technology allows for the creation of AI generated voices that can read out text in multiple languages, aiding industries such as education, healthcare and communication apps. And reportedly it only requires about 15 seconds of source audio, which if you've done any work with AI voices, we talk about companies such as eleven Labs, well Said Labs, you know, and most of the big companies are creating their own kind of voice AI voice platforms. Having only 15 seconds, needing only 15 seconds of reference audio is kind of wild, especially if it produces results similar to what we're seeing here from Voice Engine. All right, so for our live stream audience, hopefully y' all can hear this. Let me go ahead and we're going to share, we're going to go ahead and share a Quick example here, if I can bring it up and hopefully everyone can hear this. So we'll have this in our newsletter. So if you're on the podcast, you can make sure to go to your everydayai.com Sign up for the free daily newsletter. We'll be sharing a link to where you can go kind of listen to this yourself. So here we go. In the first clip, these are just two very short clips I'm going to play. So we have the reference audio. So I'm going to first play the reference audio. So in theory, whenever people get access to this, this is not publicly released. Similar, similar to Sora, OpenAI is just releasing this kind of internally right now and letting us all, you know, kind of drool over the capabilities. But here is the reference audio. 15 seconds. Let's take a listen. Force is a push or pull that can make an object move, stop or change direction. Imagine you're riding a bike down a hill. First, the push you give off the ground is the force that gets you going. 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 G gen AI. 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 of or just need help building your front end AI strategy, you can partner with us too. Just like some of the biggest companies in the world do. Go to your everydayai.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 genius. All right, so hopefully and hey livestream audience. I'd love to hear, just to confirm that we could hear that. So now here is the generated audio. So this is reportedly from OpenAI's kind of platform here. This is the generated audio based off that 15 second reference clip. Let's take a listen. Some of the most amazing habitats on Earth are found in the rainforest. A rainforest is a place with a lot of precipitation and it has many kinds of animals, trees and Other plants, y', all, that's wild. All right, here we go. I'm going to do about a five second clip. So here we go with the again. Reportedly this is all we have, but here is a very quick listen to the reference audio. First force is a push or pull that can make an object move stop or change direction. And then we have the AI generated version here. Let's take look a a listen. Some of the most amazing habitats on earth are found in the rainforest. A rainforest is a place with a lot of y'. All. That's wild. So we've used here at Everyday AI, we've used so many different voice cloning technologies. A lot of times we just do little tutorials and we just like to see where the technology is headed. This is. Wow. Yes. Okay, so I, I agree with what Josh just said here. Just wow. Tara saying, love the voice. Reminds me of a PBS type show. Yeah, pretty, pretty good voice there. Right? But y', all, I cannot state this like that is amazing. It. The quality and the ability to replicate that voice, which, you know, the, the source audio or kind of the source voice there was pretty unique voice. Right. And the ability to replicate it kind of shocking. Right? Especially if all it does take is 15 seconds of reference audio. And y'. All. All right, I'm going to save some of my takes on this for, for a second, but let's get back to kind of some of the details of this. Right, so we teased it. Let's talk a little bit more about what this means. So right now there's going to be partners who are apparently going to be getting early access to Voice engine. And, and those partners include, and this is according to OpenAI, include Age of Learning, Hey Gen, Dimaji, LiveOx, and Lifespan, showcasing diverse applications from pre scripted voiceovers to real time student responses. All right, so a little more about OpenAI's voice engine. It's trained on a mix of licensed and publicly available data with only about 10 developers currently having access to the model. Geez. Only 10 people. All right, so OpenAI emphasizes ethical usage policies in the blog post that we'll be linking in the newsletter, prohibiting impersonation without consent, requiring explicit speaker consent, and adding watermarks to trace audio origins. All right, so we also have to talk here about the US government because the US government is taking steps to regulate AI voice technology with recent bans on AI voice cloned robocalls to prevent misuse. So OpenAI suggests that they're taking a lot of measures to mitigate risks associated with this new AI voice technology, including phasing out voice based authentication for sensitive accounts and developing tracking systems for AI content. So yeah, this should be pretty interesting with only 15 seconds needed for reference audio and with all of the privacy data, deep fake concerns. Right, so, so here's the thing. It's not hard, it's not difficult right now to clone someone's voice even without their consent. So it is going to be interesting when you, I mean, presumably this technology is going to be able to be used by hundreds of millions of people. Right. So we have that list of, you know, partners who may be getting early access to, to this voice engine technology, including Age of Learning, hey, Gen, DiMaggi, LiveX and Lifespan. All right, so is this scary? Absolutely. Right. The tech, it's one of those things. Is the technology almost now too good? Right. Voice cloning is not new. It's been around for many years, or just AI voices have been around for many years, but for the most part it's either required just a lot of source audio and then the quality of the voice might not be there. It might be 80%, 90% at least with this one example of what we have here from OpenAI. I mean, again, I'm not a professional audiologist, I don't even know if that's a thing, but it sounds pretty accurate. It sounds pretty accurate. I would say, I don't know, about 90, 95%. My hearing's not great. I'd love to even hear from our live stream audience. What percent would you say that replicated? Because it sounded almost identical to me. Yeah, great, great point here from Rolando saying if you are an author, you do not have to read your own Audible version anymore. Yeah, obviously, you know, we talk about even job displacement. Right. Will voice actors have a future? I, I don't know. Right. That at least seems to me one of those kind of areas where, you know, I think the jobs are going to very, very greatly changed pretty quickly. All right, let's keep going on with the AI news that matters. All right, our last big story of the day here. Nvidia is facing challenges as tech companies are uniting to break up Nvidia's AI dominance. So a group including Intel, Google, arm, Qualcomm and Samsung is developing an open source, open source software suites called the Unified Acceleration Foundation. So that is for short, uxl. And this is to counter Nvidia's AI market domination. All right, so this new project aims to prevent AI developers from being locked into Nvidia's proprietary tech, allowing code to run on any machine and chip. So this new UXL collaboration, it includes the Open or sorry, the one API open standard developed by intel to eliminate dependencies on specific coding languages and tools specifically tied to Nvidia's CUDA platform. All right, so Nvidia recently became the first chip maker to reach a $2 trillion market capitalization driven by its focus on AI hardware like their chip, the H100 and the upcoming H200 GPUs. Also Nvidia's CUDA architecture C U D A. It currently outperforms all competitors and it's led to a high demand and chip scarcity. All right, little more quick here about the UXL project. So it initially is aiming to broaden options for AI and high performance computing applications with plans to eventually support Nvidia's hardware and code. So pretty, pretty interesting there. They're going full open source, including the ability to collaborate or to use Nvidia's hardware and code as well. So they are seeking uxl, is seeking collaboration with additional chip makers and cloud computing companies such as Microsoft and Amazon to ensure broad deployment possibilities. All right, so this is an interesting one. I'm sure this has happened before, but I was trying to even think when was the last time that you had competitors, right? So as an example, Google, Intel, Qualcomm, arm, these are all companies, multi billion or trillion dollar companies teaming up. You don't see this often. You don't see this often. I think everyone is finally catching up to what I said. Literally almost a year ago, I think I had an episode here on Everyday AI. So hey, if you've been here since that original episode, thank you. But about nine months ago or so I had an episode saying why Nvidia is the most important company to the US economy. I don't think people understood then and I still don't know if people have even fully understand it now. Right. Like back when that episode Nvidia, I don't even think it was a top 20 company in the US and now they're a top three three company by market cap. They are so far ahead of everyone else. It's not funny. Right? Yes, I was lucky enough to partner up with Nvidia and went to their GTC conference, you know, about now, 10 days ago. Feels like I'm still, you know, jet lagged from, from that trip and you know, great conversations that we were able to have out there. But Nvidia is so far ahead. Even their most powerful chip right now, now that is currently on the market, is being replaced, right? So their Hopper chip is being replaced by this Blackwell chip that is, depending on what metric you look at, it is extremely not only more powerful, you know, on some reports and benchmarks it's 20 times more powerful, but it's also 10x more energy efficient. Right? So not only is Nvidia ahead of literally every single company in the world that has products, but their new Blackwell system is exponentially more powerful and more energy efficient than their current chip that is already leading the world in computing. All right, so very interesting development here, and it's something that I think we should all be keeping an eye on is this group of technically competitors, right? Google and intel are competing with each other. Samsung and Qualcomm, our competitors competing with each other, but they're all coming together to form this UXL just because essentially they've noticed that Nvidia is so far ahead and they don't see any other way to keep up or to catch them except to come together, right? This to, to me has that like what, the 2013 Miami Heat, right? If you're an NBA fan, when you know all these, you know, the LeBron James and, and you know, who was that? Chris Bosch and Dwyane Wade. When you know, essentially all these superstars came together and formed a super team and everyone's scratching their head and it's like, oh, is this even legal? Is this even fair? Well, I mean, maybe we're going to see something like that with uxl because Nvidia is so far ahead. It is literally requiring a super team of companies to come together and to form together to say, hey, this is the only way that we feel that we can catch up or keep up with what Nvidia is doing. All right, so that is it for today for the AI news that matters. We appreciate you tuning in and if you haven't already, please go to your everydayai.com Sign up for that free daily newsletter. If this was helpful, we'd really appreciate telling your friends Tell someone about this Share this episode so many of these episodes take 5, 10, 15 hours to plan and we do it every single weekday. So it takes you about 5, 10, 15 seconds to go ahead, repost this to your network. If you're listening here on LinkedIn, retweet it, re exit, whatever it's called. We appreciate your support. Please tune in tomorrow for our Hot Take Tuesday. Thank you for tuning in today and we hope to see you back tomorrow and every day 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.