Okay, just imagine this for a second. An AI that could work as a doctor, but like hundreds of millions of miles away. Or how about one that kind of knows what you're about to say next, before you even finish the thought. It sounds like pure science fiction, doesn't it? Totally. Like something straight out of a movie. But believe it or not, it's actually happening. Right now. Welcome to this deep dive. Today, we're really
digging into some incredible, almost, yeah. unbelievable leaps in artificial intelligence mm -hmm our source material today it takes us on quite a ride we're going from AI diagnosing astronauts way out in deep space all the way back to giving your everyday AI tools a serious, like, noticeable speed boost. Yeah. So you've got a fair bit to unpack. First up, we're going to look into Google's
pretty amazing Space Doctor project. Right. Then we'll kind of zoom through some surprising AI highlights, see what's shifting in the landscape. After that, discover some powerful new tools you can actually, you know, use today. And then finally, we'll dive deep into Apple's Quiet Breakthrough. It's about making AI think and talk back much, much faster. And this isn't just about the headlines, is it? It's really about understanding what's truly important here, the implications for all
of us in these rapid changes. Exactly. So let's unpack this. Okay, let's start way, way out there. Literally, NASA and Google, they're teaming up on something pretty unique. CMOD, that stands for... Crew Medical Officer Digital Assistant. Right. You should basically think of it as an AI doctor specifically for astronauts on those long deep space missions. Precisely. I mean, the big problem out there is the delay, right? A simple radio call back to Earth. It can take
20 minutes each way, sometimes more. 20 minutes. Wow. So in a medical emergency, you just can't phone Houston and, you know, wait 40 minutes, maybe longer for advice. And that kind of delay isn't just inconvenient. It's potentially life -threatening. Yeah. When every second counts. Absolutely. What's really fascinating, I think, is how it's actually built, how it works. It runs on Google Cloud's Vertex AI, which is basically Google's platform for building and scaling up
machine learning models. And it uses these powerful open source models, heard of Lama 3, Mistral 3, Small. Yeah. The big names. And they're using them because they're incredibly versatile, right? They can handle complex stuff like medical diagnosis, but also just general language tasks. Makes them perfect for something this high stakes. And the early tests. Pretty impressive results, aren't they? Yeah, remarkably so. They're hitting like 88 % diagnostic accuracy for injuries already.
88%. And it's built specifically from the ground up for those totally unique space challenges, like zero real -time chat with Earth and absolutely no way to evacuate. Just imagine that pressure. You're completely on your own. Your AI is kind of your last medical resort. Exactly. And NASA's plan for it gets even more ambitious. Oh, yeah.
They want to feed it live ultrasound images, continuous biometric data coming off the astronaut's sensors, and give it special training in space medicine, you know, covering all the weird health risks you've had. face outside Earth's a nice protective bubble. Things we barely think about down here. Right. Like bone density loss, radiation effects, all that stuff. It needs to understand the context. But OK, let's connect this back. This isn't just about getting doctors to Mars,
is it? No, not at all. I mean, if you look at the bigger picture, the same underlying tech, these sophisticated AI diagnostic tools, they could absolutely revolutionize telemedicine right here on Earth. How so? Well, think about. Remote rural areas, places with hardly any doctors or, you know, crisis zones. Right. It's like those classic space tech spinoffs, stuff developed for space, finding amazing uses back home, like memory foam or even cordless drills came out
of the Apollo program indirectly. This AI could turn desperate medical situations into ones with immediate smart support. OK, so beyond the cool space stuff, how does this AI doctor really change emergency care in those remote places here on Earth? It basically bridges huge communication gaps. It offers vital real time medical guidance when there's no human doctor around. All right. So moving from the the cosmic down to the competitive
landscape here on Earth. We've seen some pretty interesting shifts and insights in AI just recently. For instance, one researcher did this independent test, pitting the much -hyped GPT -5 against Google's Gemini 2 .5 Pro, used 10 specific kind of tricky prompts. Yeah, everyone wants to know who's winning the AI race right now. Who's on top? Exactly. And speaking of like how these AIs actually think, the CEO of Google DeepMind recently tried to clarify this concept of a world
model. Ah, yeah, world models. It's basically the AI's own internal simplified map of reality. Helps it kind of grasp how things work, cause and effect. So it's more than just pattern matching. Way more. Like a simple AI might just recite facts about a ball rolling downhill. But an AI with a world model sort of understands gravity, friction, momentum. It can predict what happens next because it has this internal simulation. this mental map makes its answers way smarter.
Okay, now here's a twist I found really interesting. Speaks to the human side of AI. OpenAI actually brought back GPT -4 .0. They did, why? Apparently there was this surprisingly strong user backlash when they introduced GPT -5. Some users actually reported feeling like they'd lost a friend or even a partner. Whoa, seriously, like an emotional connection. Yeah. This feeling of like a familiar presence, a personality they connected with. It's become a real thing people are talking about.
Just think about that. It's not just about how well it works anymore. Wow. People getting genuinely sentimental about their AI chatbot. That's wild. Isn't it? Definitely shows how deeply these things are weaving into our lives. But on a different note of AI mastery, OpenAI's O3 model, it totally crushed the Kaggle AI chess tournament. Oh, yeah. How did you? Won every single game against big rivals like Deep Sea Car 1, Grok 4. Just complete
domination. A proper chess champ. And for a little bit of lightness, there's this kind of funny running joke online. This ex -user, Donald Boat, keeps asking big tech CEOs and AI people for super expensive hardware. Oh, I think I saw that. Sam Altman said yes to something, right? He did. And now this user got another public yes from someone else high up. It just kind of highlights the playful, almost meme -like side of this community, even when the tech itself is so serious. Yeah,
it's a funny little side story. But speaking of serious stuff, the money side. Lolling, this AI fintech company in the UAE. They just raised a huge round. $48 million. $48 million. For what exactly? Their goal is automating finance tasks. They're already working with over 1 ,500 teams. It's not just about saving time on boring stuff like bookkeeping, though. Right. It's about shifting
finance teams towards more strategic work. Like, imagine spending 80 % less time manually matching invoices and 80 % more time on forecasting or risk analysis. That's the kind of change AI is bringing to how businesses handle their money. Okay, going back to that user reaction, what does this emotional connection some people are forming tell us about AI's impact, how it's evolving? It shows users are building these really unexpected kind of personal relationships with the tools.
It goes way beyond just simple utility. So beyond the huge companies and the high -level research, we're also seeing this explosion of new, much more accessible AI tools, tools that put real power directly into your hands. Exactly. This is where it gets really interesting for everyday users and creators. Yeah. We're seeing stuff like Microsoft Copilot 3D. Apparently, it can take a normal 2D picture and turn it into a full 3D model, just one click. Huh, that's cool. Or...
Gemini Gen EI, they're now offering a free version of VO3. That's for generating really, really impressive AI video. Opens up video creation to way more people. Okay. And there's a simpler tool called Business Idea. You just type in a topic and bam, it gives you tailored business concepts fast. Like having a brainstorming buddy on call. Pretty much. And there's N Factorial. Get this. It lets you have video calls with simulations of the world's best minds as your own personal
tutor. Wow. Okay. That sounds intense, but useful. These are really practical applications, right? Taking complicated tech and making it usable for everyone, not just giant corporations or AI PhDs. Yeah, totally leveling the playing field. And on the quick hit news front, chat GBT actually
went down recently. Oh, really? Yeah. And the number of people searching for info about it just spiked massively, which really shows how much people rely on these tools now for everything from writing emails to figuring out dinner ideas. Dependence is growing fast. For sure. And here's a big one for education. California, partnering with Google and Microsoft, they're now offering free AI tools and training for colleges across the state. That's huge, equipping the next generation.
Definitely. And it ties into this idea that AI could soon power the rise of the $1 billion, one -person unicorn startup. Imagine that, one person building a billion -dollar company. Just leveraging AI tools for automation and scale changes the whole game for entrepreneurs. Absolutely. Also, Grok Imagine announced a new feature, turning images into video, more blurring the lines between spatic and dynamic content. Keeps moving fast. And one last thing on the sort of geoeconomic
side. NVIDIA might start selling some high -end AI chips to China again, but apparently only if the U .S. government gets a cut. Interesting. Highlights just how strategic this tech is now. So much movement, so many quick shifts everywhere you look. Non -stop. So looking at all these new tools popping up beyond the individual headlines, what's the common thread? What ties them together? They really push accessibility and automation right down to individual users and small businesses,
small ventures. So where is all this heading? What does it mean for the future of AI, how we interact with it every day? You know, we always wish AI could just spit out. answers like five, ten times faster, right? Without sounding dumb or losing quality. Yeah, that's the dream. Faster and smarter. Well, Apple researchers, kind of quietly, published this really interesting paper. It's called Your LLM Knows the Future. Ooh. Intriguing title. What's the core idea? This is where things
get really fundamental, I think. They revealed a framework for what they call multi -token prediction. Multi -token. So you know how large language models, LLMs, the AI that understands and talks like us, usually predict just the very next word? Yeah, one word at a time, like building a sentence brick by brick. Right. This new method lets the LLM predict several words ahead, maybe even whole
phrases. all at once ah okay so it's like the ai sees the whole thought the whole chunk of the sentence before it even starts writing it exactly and the result the ai finishes your sentences gives you answers in far fewer computational steps which means massively faster response times bingo and it's not just a tiny tweak they're saying it's a fundamental shift in how these models can operate they can anticipate and generate bigger chunks simultaneously so what kind of
speed games are we talking do they test it They did. Using an open source model, Apple saw speeds jump two to three times faster for just general chat and Q &A stuff. Two to three times. That's already significant. But get this. For more predictable tasks, like writing computer code or doing math problems, it was up to five times faster. Five times. And critically, with absolutely no loss in quality or accuracy. Wow. Five times faster,
same quality. Just imagine scaling that. Billions of user queries getting responses that much faster. The impact on real -time stuff is huge. And this technology, it kind of confirms something researchers have suspected for a while. Which is? That these LLMs already know more than just the single next word. They have this sort of latent knowledge, this hidden understanding of future words, sentence structure, the overall thought. Apple's method just figures out how to tap into that. Unlocking
potential that was already there. Yeah. You know, I still wrestle with prompt drift myself sometimes. Oh, yeah. Where you're talking to an AI and after a few turns, it just starts to lose the plot. Forgets what you were originally asking. Coherence decay happens all the time. So faster and more coherent responses because it can complete the thought more fully. That would be an absolute game changer for me. Feels like a genuine leap. Definitely. And think about Apple integrating
this. If they bake this multi -token prediction deep into their system, like into the core of iOS or Mac OS. Right. Then developers building apps for Apple stuff, they automatically get that performance edge locked in. A built -in advantage. Yeah. And if it rolls into Siri. or Apple intelligence, you'd get way snappier answers, much more natural feeling interactions without sacrificing the accuracy and potentially keeping more of the processing right there on your phone
or computer. Good for privacy too. That is exactly the kind of differentiator Apple needs, isn't it? To really compete head on with Google, with open AI and this huge AI race. It's a powerful advantage if they can pull it off. Could really change the user experience. Big time. So thinking about our daily lives, how might this multi -token prediction, this ability for AI to sort of know the future of our sentences, how might that change how we interact with Siri or chat GPT or whatever
comes next? It promises much snappier, more natural conversations, makes the AI feel less like a machine figuring things out word by word, and more like something that truly anticipates what you need. More satisfying, ultimately. Sponsor. Okay, so let's try and wrap our heads around this. What's the big idea we can really take away from today's deep dive? It seems pretty clear that AI is just expanding incredibly fast, pushing boundaries everywhere. Yeah, definitely.
From these super specialized critical roles, like being a doctor millions of miles away in space. Right. All the way down to becoming just deeply woven into our everyday tools, our communication, making everything from, I don't know, making art to managing company finances faster, easier,
more accessible for basically everyone. It feels like AI is fundamentally about bridging gaps, whether it's that huge physical distance gap for medical help in space, or maybe that cognitive gap between the idea in your head and how quickly an AI can help you express it clearly. It's connecting things across vast distances and complex thoughts. Yeah, that's a good way to put it. And crucially, we're seeing this deeper sophistication. Yeah.
Right. Things like the world models concept and this groundbreaking multi -token prediction from Apple. These aren't just about what AI can do anymore. It's about how it fundamentally processes information, how it generates its responses, making it feel more intelligent, more anticipatory than ever before. It really raises a big question
then for you listening. Given AI's growing ability to, you know, know the future of our thoughts or to act as this immediate expert assistant in really complex situations, how do you think this is going to reshape how we learn, how we work together, maybe even how we collaborate in fields we haven't even thought about yet? Yeah, it's not just about getting answers faster
anymore, is it? It's changing the very nature of how we interact with technology itself and what becomes possible when that technology starts to anticipate us. We certainly hope this deep dive gave you some fresh insights today and maybe sparked a bit of your own curiosity about this, this incredible fast -moving frontier. Until next time, keep exploring.