Today in AI: 📚 AI Teachers Make Classroom Debut in Arizona 🖼️Qwen Unveils Powerful Open-Source Visual Reasoning AI 💼Nvidia Acquires AI Startup Run:ai for $700M 🔧OpenAI Reportedly Eyes Humanoid
11:09 Share

Happy New Year's Eve, everyone. We are ending the year with one final deep dive into some of the biggest AI developments of 2024. Yeah, it's amazing how much has happened this year. Seriously? Yeah. And you know what's really cool? We have this daily Chronicle summary that has like everything. Education, AI robotics, and even some predictions for where AI might be going next. It really does capture the whole landscape.

It does. And I think it will be really interesting to see how all this is actually already starting to impact our lives. Oh, absolutely. AI is not just a future concept anymore. It's here. All right. So let's jump right in. Our first big story is, well, imagine schools where students only spend like two hours a day on traditional academics. Wow. Right. And the rest of the day they're learning things like financial literacy, entrepreneurship, stuff like that. Get this. It's actually already happening in some Arizona schools.

They're using AI platforms like IXL and Khan Academy to personalize that core learning time. You know, that's incredible. The pilot program they ran showed some pretty amazing results. Students are apparently learning twice as much in half the time. That's what I'm saying. It's like, whoa.

Makes you wonder if traditional classrooms even have a place in the future. Yeah, it's a really interesting question. I mean, in these Arizona schools, the human mentors are still really important for teaching those life skills, you know? It's true. So maybe AI is great for delivering that personalized instruction, but... Right, but humans are still better at things like critical thinking, emotional intelligence, all those things that AI can't really teach. Yeah, it seems like the ideal scenario would be some kind of partnership

where AI and human educators each kind of play their strengths. - I think that's spot on. AI can handle the what of learning and humans focus on the how and the why. - Totally. Okay, so moving on from education, let's talk about another area where AI is really exploding. Open source development. - Yeah, this is exciting stuff.

This year, Quinn, a big player in the open source world, released a new visual reasoning model they're calling QVQ. And, you know, some people might be like, why is open source even important? But imagine if only a few huge tech companies controlled all the really powerful AI tools. That's a good point.

Right. Open source means that anyone can contribute to A.I. and everyone benefits from it. Yeah. It's about making A.I. more accessible and democratic. I like that. More democratic. And Quinn's QVQ is a perfect example of that. This model is designed to solve complex problems, especially in things like math and physics.

by analyzing images in a really sophisticated way. It's almost like an AI detective. Yeah. You know, it looks at a diagram and figures out the relationships. That's a great way to put it. And get this QVQ actually scored amazingly high on the MMMU benchmark.

which is a standard test for visual reasoning. Like it's right up there with the top closed source models, which proves that open source AI is really a force to be reckoned with. It's a huge milestone for sure. It shows that open source AI is catching up to and in some cases even surpassing the capabilities of proprietary systems developed by the big tech companies. Right.

And I think that has huge implications for who gets to use these tools and how quickly AI develops in general. Absolutely. It means more people from researchers to everyday users can access powerful AI without being tied to specific companies. It really is leveling the playing field. Exactly. OK, so let's move on to an area where AI is literally shaping the future. Humanoid robots. This is where things get really interesting. I know, right? So there's this company called Armour.

And they've developed this revolutionary perception system that basically gives robots a sense of touch. You mean like skin? Well, not exactly skin. It's actually a network of depth sensors all over the robot's body. Oh, wow. So the robot can feel its way through the world kind of like we do with our skin. Exactly. And it makes a huge difference. Their system reduced collisions by like 63.7% and improved navigation by almost 79%. That's compared to traditional robot vision systems. Yeah.

And it processes information 26 times faster. That's insane. Think about the possibilities if robots could navigate with that kind of awareness. I know it's incredible. Imagine robots working in hospitals assisting with delicate procedures or in factories operating more efficiently and safely. It's a massive leap forward in robotic perception.

It could totally change how we interact with robots and how they fit into our lives. Speaking of big moves in the AI world, we've got to talk about some of the giants like NVIDIA. They made some huge acquisitions this year, including Run.ai.

for like $700 million. Yeah, Run.ai makes software that optimizes AI workloads, and NVIDIA is planning to make it open source. It's so smart. They're not just selling hardware, they're building a whole ecosystem around AI to make it easier for developers. It's a brilliant strategy, and they're not the only ones making big moves. There's this rumor that OpenAI, the company behind ChatGPT and Delali, is looking at getting into the humanoid robotics market. Oh, wow. That's interesting. Right.

Think about it. They're experts in AI software and algorithms. What happens when they start building robots? It could lead to robots that can learn, adapt, and solve problems in really complex situations. It's really exciting, but it also brings up one of the most talked about topics in AI this year, artificial superintelligence, or ASI. You mean like AI that surpasses human intelligence? Exactly. It's a concept that's both exhilarating and a little bit scary, to be honest. Yeah, it definitely sparks a lot of debate. So what's making people think that

ASI might be closer than we think.

Well, for one, the speed of AI development is just crazy. We're seeing breakthroughs all the time and models are getting more powerful every year. And then there are the insights from leading AI researchers like Ilya Sutskever. He's been saying that ASI could arrive sooner than many people expect. So it's not just science fiction anymore? Not at all. And companies like Google and ByteDance, the company behind TikTok, they're taking it very seriously. Like ByteDance just invested $7 billion in NVIDIA AI chips.

That's a serious commitment. It is. And Google CEO Sundar Pichai has said that making their Gemini AI available to everyone is their top priority for 2025. These companies are racing to dominate the AI landscape because they see the massive potential. It's clear that AI is changing everything so fast.

But before we get ahead of ourselves, let's dive deeper into this idea of artificial superintelligence. Yeah, it's definitely a topic we want to explore in more detail. We need to talk about the potential benefits and the risks and all the ethical considerations that come with something this powerful. You're right. It's a lot to unpack. It is. But it's important stuff. Absolutely.

Welcome back. You know, before we get into superintelligence, I'm still thinking about those investments from Google and ByteDance. I know, right? Like $7 billion from ByteDance alone. It's kind of hard to even imagine that much money. Yeah. But yeah, let's talk more about artificial superintelligence or ASI. It's a fascinating concept, but I got to admit, it's also kind of intimidating. Yeah. So what are some of the potential benefits?

benefits and risks that people are talking about? Well, on the one hand, some people think ASI could be the key to solving some of humanity's biggest problems. Like imagine an AI that could help us tackle climate change or eliminate poverty. That'd be amazing. Right. Like what if we had an AI that could design super efficient energy systems or cure diseases that are impossible to cure right now, or even create economic models that ensure everyone has a good standard of living? It's like having a superpower problem solver. Exactly.

But, and this is a really important but, there are also some serious concerns about potential downsides. Likewise. Well, one of the biggest worries is something called the control problem. Basically, how do we make sure that a super intelligent AI stays aligned with human values and goals? That's a good point. If we create something that's potentially smarter than us,

How do we stop it from going rogue or doing things we don't want it to do? It's a tough question, and it's something that AI ethicists are really grappling with right now. I mean, we're talking about a level of intelligence that could outsmart us. So it's hard to predict or control what it might do. So potential benefits, but also potential risks.

It sounds like we need to be incredibly careful as we move forward with ASI. Absolutely. And that's why there's so much discussion and debate happening right now in the AI community. It's not just about pushing the limits of what's technically possible. It's about making sure that these advancements are actually good for humanity. Right. It's not enough to just create powerful AI. We have to make sure we're using it responsibly. Exactly. Wow. It's amazing to think about everything we've talked about today. You know, AI tutors,

superintelligence, humanoid robots. It feels like AI is changing the world faster than ever. It really is incredible, isn't it? And it's only going to get faster from here. Yeah, I totally agree. Before we wrap things up, though, I wanted to go back to something we touched on earlier, that idea of AI becoming more democratic. Oh, yeah, that's an important one. I think so, too. And it's something that keeps coming up, you know, like with Quinn and their open source visual reasoning model.

It just shows how open source AI is making these powerful tools available to everyone. Absolutely. It's not just about the tools, though. It's about the whole philosophy. You know, open source AI encourages collaboration and sharing, which leads to faster innovation and more diverse applications. Exactly. And that's why it's so important to have lots of different voices and perspectives in the AI field.

We need people from all walks of life contributing their ideas to make sure AI benefits everyone. I couldn't agree more. If we want AI to be truly for everyone, then we need to make sure it's being built by people who represent everyone. Okay. So as we close out this last deep dive of 2024, I have one final question for you and for all our listeners. What role do you want to play in shaping the future of AI?

That's a great question to reflect on. It's easy to feel like these huge advancements are happening out there. But the truth is, everyone has a role to play. Whether you're an AI expert, a student, or just someone who's curious about this stuff, your voice matters. So how can we make our voices heard?

What are some concrete things we can do to get more involved? Well, a great first step is to just learn more about AI. Read articles and books, listen to podcasts, attend webinars. There's so much information out there. And connect with others who are interested in AI too, you know. Join online communities, go to meetups, or even start your own discussion group. It's about taking ownership of this technology and not just leaving it to the expert. Exactly. And don't be afraid to speak up, share your thoughts, advocate for the kind of future you want to see. Every voice matters.

Whether we're writing, speaking, or just having conversations with our friends and family, we can all contribute to shaping the AI narrative. That's right. The future of AI isn't set in stone. It's something we're all creating together. It's a powerful thought to carry with us into the new year.

So as we celebrate the arrival of 2025, let's also renew our commitment to building a future where AI benefits all of humanity. Here's to a bright and thoughtful AI-powered future. Happy New Year, everyone. Thanks for joining us for this special New Year's Eve episode of The Deep Dive. It's been a great year. It really has. And we can't wait to see what amazing things AI brings in 2025. See you next year.