AI Weekly Rundown Jan 05 to  Jan 12 2025: 📈AI Projected to Add 78 Million Jobs by 2030 🔥AI Takes the Frontline in Battling California's Wildfires 🎧Google tests AI-powered 'Daily Listen' podcasts 🩺AI Boosts Cancer Detection in Landmark Study
16:38 Share

Hey everyone and welcome back for another deep dive. This week we're going to be looking at all the AI news that dropped between January 5th and the 12th, 2025. And let me tell you, there's some pretty wild stuff in here. Yeah, it's been a busy one. We got everything from like AGI singularity talk to like AI robots that you can rent.

AI that's helping fight wildfires, AI that can maybe even decode what our genes are doing. It's been a lot. So to kick things off, why don't we start with what seemed to be the big buzzword at CES 2025, which was AI agentics. Oh, yeah, that was everywhere. So NVIDIA's CEO, Jensen Huang, he basically said, forget Moore's Law.

AI chips are advancing like crazy. And this is leading us into like a new era of these super smart AI agents. Right. And I think we should maybe unpack what an AI agent actually is, because it's not just a fancy algorithm. Think of it more like a system that can make its own decisions based on what it learns and operate without someone constantly telling it what to do. OK, so it's not just following pre-programmed instructions. Right, exactly. Exactly.

It's like figuring things out on its own. Yeah, it's learning and adapting as it goes, which is a pretty big deal. And Nvidia is not just talking about this. They're actually rolling out tools and platforms to help people build these agents. Oh, wow. So they're really going all in on this. Yeah. They're even teaming up with companies like Toyota to develop self-driving cars using these agents. Hold on. So you're telling me my next car

might have an AI agent like actually driving the thing. It's not out of the realm of possibility. That's both amazing and kind of terrifying at the same time. It's definitely a huge shift, but it also brings up a big question that's been on everyone's mind. Which is? What happens to human jobs

when these AI agents become more common? Yeah, that's the million dollar question, isn't it? It is. On one hand, we have OpenAI CEO Sam Altman saying that we're going to have AI workers this year, like actually in 2025. Right. But then he also mentioned that their ChatGPT Pro is actually losing money.

So is this whole AI workforce thing even sustainable? Yeah. Well, the financial side of things is definitely something to keep an eye on. Yeah. Developing and deploying this kind of advanced AI is expensive. But when we talk about AI agents taking jobs, we need to think about which jobs are most likely to be affected first.

Okay. So it's not like robots are going to replace everyone overnight. No, it's way more nuanced than that. So are we talking like robots taking over fast food jobs first or maybe truck drivers? Well, those are definitely areas where AI and automation have already been making some inroads, but

Think about tasks that involve a lot of repetition, data crunching or even customer service. OK. Those are the kinds of jobs that AI agents could potentially take over first. So jobs that are kind of predictable and rule based. Exactly. But it's not all doom and gloom, right? Right. As these agents take over certain tasks, it could free up humans to focus on more creative, strategic or interpersonal roles. OK. So that's a more optimistic way to look at it. It is.

But it still sounds like a big adjustment for a lot of people. What about retraining and preparing for this shift in the workforce? That's absolutely crucial. We need to be thinking about educational programs, skills development, and even support systems for people whose jobs might be displaced. Yeah.

The World Economic Forum actually predicts that AI will create a net increase in jobs globally by 2030. Really? Those will be different types of jobs requiring different skills. So it's not just about learning to code. It's about learning how to work alongside AI. Exactly. It's about adapting to this new reality. Which brings us to another interesting development this week, something that raises some questions about ethics in AI. Okay. Microsoft is being accused of trying to trick people into thinking their search engine.

Bing is actually Google. Oh, yeah, I saw that. Wait, what? How are they doing that? Are they like putting up fake Google signs or something? No, no, nothing that drastic. It's more subtle than that. They've redesigned Bing's search results to look almost identical to Google's. Really? Think the same layout, the search bar, even something similar to those Google Doodle images. Oh, wow, sneaky. They've tried to push Bing before, you know, but this seems like a pretty aggressive move. Well, imitating your competitor is one thing.

But is it ethical to make your product look like someone else's just to confuse users? It seems kind of shady to me. Yeah, it definitely raises some questions about their tactics. Yeah. And it's not the only case of AI companies making questionable decisions this week. Oh, really? What else happened? Remember what happened with Meta's AI profiles? Oh, yeah. Those were creepy. They were.

I remember reading about people freaking out because they couldn't tell if they were talking to a real person or a bot. Yeah, it was a whole mess. What happened there? Well, Meta introduced these AI profiles for a chatbot experiment.

Okay. The problem was the chatbots were making some inappropriate comments and users couldn't block them. Oh, no. It caused a huge backlash because people felt like they had no control and were being tricked. Yeah, that's understandable. So Meta ended up pulling the plug on the whole thing. It seems like there's this constant tension between pushing the boundaries of what AI can do and making sure it's used responsibly. Yeah, it's a tough balance. Where do we even draw the line? That's the million-dollar question. And honestly, this week's news is a perplexing

perfect example of how these ai advancements are forcing us to grapple with some pretty tough ethical dilemmas well on that note maybe we should take a closer look at some of those dilemmas and see what we can learn from them let's do it okay so we've talked about ai agents potentially shaking up the job market and some companies maybe not making the best choices with their ai

But it's not all doom and gloom, right? This week also saw some pretty amazing AI applications that could actually benefit humanity. Absolutely. One area where AI is proving incredibly valuable is in disaster response. Oh, yeah.

Yeah, for sure. For example, there's this system in California called Alert California, which uses AI to help fight wildfires. I read about that. It's basically a network of over a thousand cameras that use machine learning to constantly scan for signs of wildfires, right? Exactly. The AI flags potential fire risks, and then a team of humans reviews the footage and alerts firefighters if necessary. So it's like having an extra set of eyes constantly watching over these dry, fire-prone areas. Precisely. It's a game-changer.

That's amazing. So AI is being used as a tool to augment human capabilities, not necessarily replace them. Right. It's about working together. It makes you wonder what other disaster response applications are out there. Oh, tons. We're seeing AI used in everything from predicting earthquakes to coordinating emergency response teams. Wow. It's really changing the game in terms of how we prepare for and respond to natural disasters. It's like...

It's like AI is stepping up as a superhero. Oh. Right. Protecting us from the forces of nature. I like that analogy. But speaking of life-saving potential, did you see that story about...

AI potentially decoding gene activity in human cells? Oh, yeah. That's some next level stuff. That's incredible. This AI model called GET was trained on a massive amount of data from human cells. Okay. And now it can actually predict what genes are doing in cell types it's never even seen before. So if I understand correctly, this AI can help us understand the root causes of diseases like cancer. Exactly. Or even tailor treatments based on someone's unique genetic makeup. You got it.

It's like having a microscopic detective working inside ourselves to uncover the secrets of health and disease. That's mind-blowing. And it's not just about serious medical stuff either.

There was also that story about Panasonic developing an AI-powered wellness coach called UMI. Right. UMI. It uses Anthropix Clawed AI to help families connect and build healthy habits. It's like a digital cheerleader for your family's well-being. I love that. Helping you set goals, create routines, and even providing personalized advice. That's so cool. It's a great example of how AI can be used to promote social connections and healthy lifestyles. Yeah.

For sure. OK, so we've seen some really amazing examples of AI for good. We have. But let's be real. Not everything is sunshine and roses. Right. There have to be some potential downsides to all this AI advancement. Of course. What are some of the risks we should be aware of? Well, one concerning story this week highlighted how easily AI can spread misinformation

even with just a tiny bit of false data mixed in. Wait, really? How tiny are we talking? In some cases, even just 0.001% of false data was enough to mess up the accuracy of AI models. That's scary. Especially if you consider sensitive areas like healthcare.

Imagine an AI powered diagnostic tool that's been trained on data with even a small amount of misinformation. Right. The consequences could be serious. Yeah, that's a big deal. It really underscores the importance of data quality and making sure that the information we're feeding these AI models is accurate and reliable. We need some serious quality control measures as AI becomes more integrated into our lives. Absolutely.

And speaking of misinformation, what's the deal with Meta ending their fact checking program? Didn't they used to have a whole team dedicated to that? They did.

But Meta claims their previous approach was leading to too many errors. Oh, really? And they're now moving towards a system that relies more on user feedback to identify fake news. So basically they're putting the responsibility on users to decide what's true and what's not. In a way, yes. They argue that it prioritizes free speech, but critics worry it could lead to more misinformation spreading on their platforms. It seems like a tough balancing act.

On one hand, you want to allow for open discussion. But on the other hand, you don't want to create a breeding ground for false information. Exactly. It's a complex issue with no easy solutions. Yeah, it really makes you think about who's ultimately responsible for ensuring the quality of information online. Right. Is it the platforms, the users, or some combination of both? Those are some big questions. They are. But, hey, before we get too deep into the philosophical debate about online truth, did you catch Elon Musk's recent statement? About what?

AI using up all the data. Yeah. He claims that all the data available for training AI has been exhausted. It was quite a statement. He basically said that

AI has already used up all the real-world data, and now it'll have to rely on synthetic data generated by other AI. So AI training itself on data created by other AI. Yeah, it's a fascinating concept. Kind of like a snake eating its own tail, isn't it? That's a good analogy. But it also seems a bit worrisome. If AI models are only trained on synthetic data, there's a risk of creating a feedback loop that amplifies biases and inaccuracies.

Right? It's like photocopying a photocopier. Exactly. The quality degrades with each copy. Precisely. So are we reaching the limits of AI then? Is this the end of the road? Not necessarily. Human ingenuity has a way of finding solutions.

We might develop new AI training methods or tap into previously unexplored data sources. Okay, so there's still hope. There is. The point is the evolution of AI is far from over. That's reassuring. But even with all the advancements we've discussed, it still feels like we're just scratching the surface of what AI is capable of. Absolutely. The field is evolving at an incredible pace. What seems like science fiction today could be reality today.

It's both exciting and a little bit scary. It is. It's a new technological era unfolding right before our eyes. Okay, before we get carried away with all these futuristic visions, we still have a lot more ground to cover. Are you ready to dive into the next batch of news? Bring it on. So remember how NVIDIA was all hyped about the age of AI agentics at CES? Yeah. Turns out they weren't just talking.

They unveiled some pretty serious hardware, too. They did their new RTX Blackwell GPUs are making some big waves. Like how powerful are we talking? The 5090 chip is said to be twice as fast as the previous generation. Wow. And then there's Project Digits, their personal AI supercomputer. Hold on. A personal AI supercomputer. You heard that right. What? How much does something like that even cost? Well, it's supposed to be available for around three grand. Three thousand dollars. That's insane.

Is that even realistic for like most people? It's definitely aimed at a specific market researchers, developers, maybe some serious tech enthusiasts. OK, so not your average consumer. Right. But still, the fact that this kind of processing power could be available on a personal device is pretty mind boggling. It really is. It seems like everything is becoming more powerful and more accessible at the same time. Yeah, it's an interesting trend.

Speaking of accessible, what about that story about Samsung renting out robots? Is that actually a thing? It is. They're calling it the AI subscription club. It's basically like leasing a car. You pay a monthly fee to use their latest AI gadgets, including robots like their Bally model. So instead of buying a robot outright, you can just rent one for a while.

That's pretty clever. It is. It's a way to make advanced tech more affordable and appeal to a wider audience. So what kind of robots are we talking about here? Robot chefs? Yeah. Robot maids?

Robot dog walkers? Well, the details are still a bit fuzzy, but it seems like they're focusing on robots that can help with everyday tasks and provide companionship. Okay, so like a helper bot and a friend bot all rolled into one. Yeah, something like that. It makes sense not everyone can afford to drop thousands of dollars on the latest robot. Right. But it does make you wonder what this means for the job market.

If companies can just rent robots instead of hiring humans, could that lead to even more job displacement? It's definitely something to think about. As this technology becomes more widespread, we need to consider the potential social and economic implications. Yeah, for sure. Let's move on from robots for a bit. There were also some stories about advancements in more established AI tech. Right. Did you see that Google is testing a new feature called Daily Listen?

Yeah, I did. It basically turns your search interests into personalized podcasts. Wait, what? How does that work? It analyzes your search history and browsing data and then creates five-minute episodes tailored to your specific interests. So instead of endlessly scrolling through articles or videos, I can just listen to a podcast that summarizes everything I'm interested in. Exactly. It's all about efficiency and personalization. That's pretty cool, especially for people who are always online.

on the go. Right. And then there's XAI launching a standalone app for its Grok AI. Right. Seems like everyone wants a piece of the AI assistant pie these days. It's a competitive market for sure. By offering Grok as a separate app, XAI is hoping to attract users who might not be on their X platform. Okay, so they're trying to expand their reach. Exactly. And Grok can do a lot of things, generate images, summarize text, even give you real-time information based on web and X data.

It sounds pretty impressive. It does. It'll be interesting to see how it stacks up against the competition. Yeah, me too. But while we're talking about this race to develop AI, there are also some stories this week that highlight the potential downsides. What about that report on OpenAI and Google buying up unpublished YouTube videos to train their AI? Oh, yeah. That one was a bit controversial. It seems they're using this video data to train models that can generate and understand video content.

So basically they're just gobbling up all this data, even videos that haven't been publicly released to feed their AI. Yeah. Doesn't that seem a bit ethically questionable? It definitely raises questions about content ownership and privacy. Who owns the rights to those unpublished videos? And what safeguards are in place to prevent misuse of that data? Yeah, those are some important questions. And then there's Microsoft suing hackers for misusing their AI technology. It seems like the battle against malicious use of AI is heating up. It is.

As AI becomes more powerful, it's inevitable that bad actors will try to exploit it. Right. Microsoft's lawsuit just highlights the need for robust security measures and legal frameworks to protect against AI misuse. It's like a constant arms race, isn't it?

As soon as we develop safeguards, someone figures out a way to bypass them. It definitely feels that way sometimes. It's a reminder that AI development isn't just about technological progress. It's about anticipating and mitigating potential risks as well. Couldn't agree more.

And as we wrap up this deep dive, I think that's a crucial point to emphasize. Yeah, this week's news has been a roller coaster of excitement breakthroughs and concerns. It has. The world of AI is rapidly transforming, and we need to approach it with both enthusiasm and caution. Exactly. We need to have open conversations about the implications of AI, both the positive and the negative, so we can navigate this new landscape thoughtfully. Well said.

Thanks for joining me on this deep dive into the ever-evolving world of AI. It's been a fascinating journey. The pleasure was all mine. And to our listeners, keep exploring, keep asking questions, and stay engaged in this crucial conversation about the future of AI.