Leveraging AI

172 | AI agents are taking over, Google is on fire with AI releases, using AI to talk to animals, and many more important AI news for the week ending on March 14, 2025

• Isar Meitis • Season 1 • Episode 172

Share episode

The AI revolution isn’t slowing down—if anything, it’s sprinting forward.

This week, we’re diving into the latest AI breakthroughs, including Google’s relentless AI releases, the rise of Manus—China’s viral AI agent—and OpenAI’s latest play to dominate the industry. Oh, and we might just be on the verge of using AI to talk to animals. (Yes, really.)

In this episode, you’ll discover:

  • How AI agents like Manus are redefining automation and what it means for businesses.
  • The game-changing advancements from Google, OpenAI, and Anthropic—and how they impact you.
  • Why AI-powered code generation is skyrocketing, with some startups using AI to write 95% of their code.
  • The security and control challenges of AI agents (and what could happen if they start making their own decisions).
  • Whether AI really can help us talk to animals—and what it means for communication beyond humans.

AI is evolving fast, and business leaders can’t afford to sit on the sidelines. Tune in now to stay ahead of the curve!

About Leveraging AI

If you’ve enjoyed or benefited from some of the insights of this episode, leave us a five-star review on your favorite podcast platform, and let us know what you learned, found helpful, or liked most about this show!

Speaker:

Hello and welcome to a Weekend News episode of the Leveraging AI Podcast, a podcast that shares practical, ethical ways to leverage AI to improve efficiency, grow your business, and advance your career. This is Isar Meitis, your host, and we're gonna start the show today by talking about new technologies that are coming up and emerging and how they're going to impact the world, and specifically the workforce. And then there's a very long list of rapid fire items with a big focus on Google and a lot of stuff that they've been releasing recently. And potentially our ability to talk to animals. So let's get started. last week we mentioned a new Chinese genic model called Manus that is taking the world by a storm. And that storm has grown dramatically in this past week. So let's dive a little deeper into this topic. A Chinese company named, called The Butterfly Effect, has released a new agentic environment called Manus that allows you to take action within a browser based on a agentic approach to tasks. It can perform multiple tasks at the same time. some users report 50 plus tasks all running simultaneously, and it does things like social media analysis, stock trading, travel planning, online shopping. Basically all the stuff that we can do on a browser, it does on its own. Now, in addition to the fact that it can do all these tasks, it's actually scoring very high across multiple benchmarks. So Manus claims superior accuracy over deep seek and even over open AI in a benchmark called. Gaia, which is spelled GAIA, which is a benchmark for testing agents, which is very different than the large language model benchmarks where it's a non benchmark. Here, it actually evaluates it on real world tasks and it actually scores higher than more or less every other tool out there, including the top chat GPT models. Now, while it doesn't have the human-like more conversational voice of chat, GPT, it definitely outperformance when it's getting to getting the job done in performing tasks. It is currently only available by invitation only, and it has been rumored that invitations to get access to this model has been selling in China for numbers between a hundred dollars to 15,$1,800 per invitation. I am sure that the company did not expect this kind of demand, and they are struggling and they mentioned that themselves, that they're gonna try to deal with that and ramp up their demand, but while it was very obvious even in 2024, that 2025 is gonna be the year of agents, we haven't seen anything in the immediate user market. So enterprise level agents have been deployed. There are multiple big tools for large corporations to develop and deploy and manage agents, but that was not available to the wild public or to small businesses. Manus is the first shot if you want, fired in the open to the public agentic universe. And I am 100% certain that this will not be the last, and it will probably accelerate the process dramatically. So how does it work? Well, in the backend it is using Claude 3.5 sonnet and they're also testing Claude 3.7 combined with Quinn from Alibaba to actually do the thinking process. And they're trained their own capabilities on top of that, and the agents to actually take actions on the different browsers. I don't have access to it yet, but I've seen multiple demos of people doing some pretty remarkable stuff. It's still not fully functional. There's still some sharp edges and it doesn't always get the job done. And he does some weird things every now and then. But the direction is very, very clear. And as I mentioned about mid 2024, the year of 2025 is going to be the year of agents. And again, now that Manus has done this, I expect the big players from the US and around the world to follow suit and provide us similar capabilities. So if you think about the fact that we got browser access from Claude midyear last year, and we got chat, GPT operator at the end of 2024. This was just the beginning, and I think tools like Manus are gonna be everywhere and available to all of us across multiple aspects. As I mentioned in previous episodes. The two biggest things that these things have to solve in order to really be used as a widespread business tool versus a cool gimmick that you can use to do research on the next sneakers you wanna buy is security and control. Meaning having the ability to define very clear. Guardrails on what these tools can and cannot do. In order to protect your computer, protect your data, protect your company servers, as well as stop them from doing stuff that you didn't plan to, in environments they do have access to. So let's say you allow it to shop online, you want it to shop for sneakers, and now it finds a really cool Mercedes to go with your shoes. It may go and buy it because it does its own thing. What prevents it from doing this is its own logic, but I think there needs to be another layer that protects beyond just the thought process of the agent itself that will define which websites it can visit and cannot visit, what things it can and cannot do based on personal preferences or company guidelines. And once these things will be in place, I think these tools, we'll see significantly wider spread and usage because they are incredibly powerful. Connecting this to many conversations we had in the past few months about the impact on the workforce. This can do the work of multiple employees with one agent right now for free, shortly, probably for a small amount of money, and it can actually follow a process. And the next step would be how to train these models, which will probably gonna be as simple as just showing them what they need to do once, and then they can do it again and again and again at a much higher speed. At a much higher accuracy, never sleeping, never taking vacation, and so on. And this is not perfect yet, but it's getting very, very close to the point we'll do that, which will have an even bigger impact on the workforce than large language models. And generative AI is right now. Now to show you how big the impact of AI is on the workforce. There's another very interesting headline from Y Combinator this past week. So Y Combinator, probably the most known accelerator in the world, has reported that 25% of their startups in their winter 2025 cohort have their code bases generated 95% by ai. So think about what I said. A quarter of the company in the most advanced accelerator in the world are basically creating their products that are software products by ai almost completely. That was shared by Jared Friedman, who's a managing partner over there. So it's from reliable sources. Friedman also mentioned that every one of those people is highly technical, completely capable of building their own products from scratch, and yet they're letting AI write 95% of the code. He's saying a year ago these founders would have coded the whole thing from scratch themselves, and now they're using AI to slash development time, letting startups sprints significantly faster than ever before. Now, I mentioned to you in the past two weeks the concept of vibe coding is gaining more and more attention, especially around Silicon Valley. And the concept behind vibe coding is that you just use your natural language and let the machine write the code while you're monitoring what it's doing and just working along the machine, letting it actually write the code. Now the reality, this is an amazing tool that is currently helping developers develop faster systems, but it's also giving access to people like me who cannot write code to write small applications, to do different things that were not possible before. And this means that you can do the same thing too. If you are a coder, I assume you're already using these kind of tools, but if you're not, then you can start trying them. And I'm actually planning on recording a specific Tuesday episode on the various vibe coding tools that allows you to develop applications and different capabilities from scratch on your own without knowing anything about writing code. So that will probably come in the next few weeks. An additional support to this current situation. Dario Amide, the CEO of Anthropic in an interview this week was claiming that 90% of code will probably be written by AI within three to six months, and nearly a hundred percent within 12 months from today. He shared that during a console on foreign relationship event, and we'll share a few other things that he said on that event that are really interesting, but this is a very bold and aggressive timeline, right? So we had AI coding tools for a while. Now, whether it's GitHub Copilot or Anthropic Claudes code, which they just released, that we shared last week, or cursor and many others that are gaining a lot of momentum. But going to 90% of code within the next six months and a hundred percent by the end of the year is very aggressive. Even if he's wrong, the direction is very, very clear. More and more code is being generated by coders and more and more code is generated by AI versus traditional coders. Now, Dario says that programmers are not going to vanish. They're just steer the ship. They will basically be a team lead, running a lot of ais writing code for them, that they'll be able to monitor and guide and reshape in the way they need in order to develop their products faster. Instagram's co-founder Mike Krieger, who is now Anthropic agrees and he's predicting that engineers will soon shift from writing code to reviewing AI generated code. And a lot of other people such as A-W-S-C-O, Matt Garman, are stating the same thing. And VCO, Jason Hong also shared similar sentiments. He's claiming that coding might be already dead in the water with the rapid development of these tools. And he's basically suggesting to younger professionals or people going into universities right now to go and study biology, education, manufacturing or farming as plausible and more secure future career options compared to writing code. Again, very extreme predictions, but coming from some of the most influential and relevant people in the tech industry. Staying on the same topic of pushing the limits of what AI can do in the organization, and also talking about Anthropic. Anthropic just released an upgraded console that allows to share prompt libraries and other AI developed tools across the enterprise. Which basically lets people from different departments such as marketing, customer service, and the tech team to share resources and prompts between the organization, and now I'm quoting from their spokesperson. We built our shareable prompt to help our customers and developers work together effectively on prompt development. What we learned from talking to customers was that prompt creation rarely happens in isolation. It's a team effort involving developers, subject matter expert, product managers, and QA folks all trying to get the best results, which is true, right? If you think about it, if you look at a job, it's very rarely done by one person. Different people has different aspects of the job, and if you can develop prompts or one big prompt in tandem with all these people and all their inputs in mind, you can develop better, faster, more consistent solutions that will work for everyone. And now it's gonna be built into the actual systems and processes themselves. Staying on the topic of changing the world and how it works and the whole impact of agents on our future workforce and technology in general. On March 11, OpenAI launch, a new API called responses API, which is a new tool set that allows to blend the simplicity and basic concepts of chat to advanced agentic capabilities. And the assistance aPI. So the responses, API, which is this new tool, combines three previous capabilities. One is web search, which allows you to get real-time answers with citations. File search, which allows you to dig into companies documents faster and better, and computer use, which allows you to actually use the mouse and keyboard to automate tasks on the computer. Going back to the Manus moment, while Manus has developed this as a end user tool on the open AI universe, this immediate release is a tool for developers because it runs on the API side, but it allows to do some remarkable things and develop capabilities that were not previously possible for most companies. Also in this new tool mentioned that the new web search capabilities is now showing 90% accuracy versus the 63% accuracy of the previous tool, which tells you it's still not bulletproof to hallucinations, but it's a very big step in the right direction, and I'm sure this vector will keep on moving to give us answers that are more and more reliable over time. They're also released a agent, SDK, that upgrades the swarm framework that they released previously, and it's letting developers orchestrate multi-agent workflows with ease. So configure the agents smart tasks, handoffs between agents and safety guardrails as I just mentioned previously, which are a necessary component in order to actually deploy this safely in an organization. Now, while all these things are moving very, very fast, they're not completely reliable yet. As an example, the computer using agent model, while it's really impressive scores, just 38.1% on complex tasks on the OS world benchmark. So that means it doesn't get the job done on complex tasks two thirds of the time. This means that if you're planning to use these agents in the immediate future on simple tasks, you're probably safe or safer and more complex tasks. You are not there yet, and this means that you need human supervision to monitor what these agents are doing. In my software company, data Breeze, that develops AI automation solutions for companies. This is a key component, which is allowing users to verify the automations that the agents are performing this way to reduce the errors that the system may generate over time. We're seeing significant decline in these errors as the systems learns, different use cases, but it's something that you need to be aware of. Just like you need to be aware of. Hallucinations with generative ai, these tools will not get it a hundred percent of the time correct. The flip side of that, humans don't get the job done a hundred percent of the time. Correct. So all those tools needs to do is get to human level errors and then go beyond that in the near future. And there's zero doubt in my mind that that's where we're going. Now to a slightly contradicting point of view on the direction that AI is taking right now, hugging Face co-founder and Chief Scientist, officer Thomas Wolf, is arguing that today's AI system are morphing into, and I'm quoting yes, men on servers rather than revolutionary thinkers that are needed for breakthroughs. He is specifically challenging Dario Amadei, the CEO of Anthropic in his vision of a compressed 21st century where AI drives decades of progress into years or months. And he's saying that it's very unlikely with the current tech. He's claiming that scaling today's AI models will not create Einstein's in data centers. And he's saying that the main mistakes people usually make is thinking that Newton or Einstein were just scaled up good students. And he's saying that that's not the case. Geniuses comes from questioning and not by conforming with the known facts. He's saying that today's systems are built to excel in answering known questions versus questioning the status quo and trying new things. And hence he's saying these systems will not be able to have like the Copernicus moment who found that the earth is orbiting the sun and not the other way around by questioning the norm and the consensus of his time. So it will be very interesting to see there are really smart people on both sides of that argument. But this is definitely an opinion that you need to know that is out there right now by people at the forefront of the AI revolution. So a quick summary of what we talked about so far. Agents are coming, they're coming fast. They're coming to the enterprise world. They're coming to the backend, API universe, and they're coming to the front end for small businesses and individuals to use. They are already examples and capabilities and tools out there to develop and deploy these agents. And it's a matter of time, months before each and every one of us will start using agents for our personal lives as well as in our businesses, whether small or large businesses, and the implications of that are so broad that nobody is ready for that. Whether the implications on the workforce, what exactly people are going to do, what kind of IT resource will we need? What are the safety and security measures that needs to be put in place? Again, both on the individual level and on the company level. What does it mean for websites and the internet as a whole? if there's gonna be less and less people and more and more agents crawling to collect information, what does that mean to basically everything we know on Hunt digital business is run today, and that is going to change this year or worst case scenario in the next 18 to 24 months. And we need to start thinking about this as people, as a society and as business leaders because it has profound implications on, as I mentioned, more or less everything we know. The last topic before we dive into the rapid fire is Anderson Horowitz released their six months, the top 100 Genai consumer apps. That's their fourth edition, and they released it on March 6th. OpenAI is still the king and still at the top of the list with 400 million weekly active users by mid-February of 2025, doubling from 200 million in just six months. This is an insane growth that is showing how powerful the tools and the brand are of OpenAI and Chat GPT. The surprising thing is that number two, you have deep seek. So deep seek did not exist on the previous list six months ago, and they're now ranking number two on the web usage ranking. But they're not the only newcomer to the list. 34% of the top 50 web AI products are new on the list from six months ago. This is just showing you how dynamic the AI world is and with the cap, with the ability to take an underlying model and develop something new. On top of that, from a tooling and capabilities perspective is relatively easy today and with not a lot of moats. The moats are around maybe developing the big models and the frontier models, but even that is now questionable with the ability to distill models and create things like deep seek. But again, 34%. One third of the top 50 AI applications were not on the list just six months ago. Two categories that showing big growth on the list. One is AI video tools that have evolved from, you know, glitchy experiments for geeks in early 2024 to tools you can actually use to create small business videos or even enterprise level videos for actual business use cases and definitely for personal usage. So AI video tools are high in the category, and the other, as we mentioned before, is vibe coding tools that allows you to build applications without knowing how to code. And as I mentioned, an episode is coming on that topic sometime in the next few weeks. On the mobile side. There are even bigger surprises. Chachi PT still rules the mobile side as well, but in number two you have Microsoft Edge, and in number three you have photo math. Both are tools that I A, don't really know and B, definitely wouldn't count as AI tools, but they are on number two and number three on the LI on the list, I will share the link to the entire list with you. I must admit, I don't know, a lot of the tools are in the top 50 on both web and mobile, and it's interesting for every one of you to go and take a look to see what other people are using. And now let's dive into Rapid Fire. The first rapid fire item is Ilia Sr. With his company SSI just raised another$2 billion at a$30 billion valuation, which is six times the value from just five months ago. The company that he founded just recently. And those of you who don't know the story, Ilio Kovar was one of the leading scientists in open ai. Many people are claiming he was the most important scientist in open ai. He went silent after the Clash that got Sam Altman out of the company and then back into the company. And then about six months after that event, he left the company and founded SSI. Which stands for Safe Super Intelligence. He surrounded himself with a bunch of really talented people, and just because he's who he is, maybe the most capable AI scientists in the world there, but if not, definitely one of the top five. He's able to raise whatever money he wants. Now, he hinted, and I'm quoting that he's climbing a different mountain from Open Eyes methods with early signs of promise. So what does a different mountain means and what does early sign of promise means? I don't think anybody knows, or maybe very few people knows. They're actually asking all their employees not to even mention that they're working at SSI. So nobody has SSI on their LinkedIn profile as an example. So they're trying to keep everything very hush. They have about 20 employees, which is relatively little. But because it's Ilia sca and because he has promising starts, he was able to just raise$2 billion and get a$30 billion valuation to make this more extreme. He's not planning any products and any early releases until he achieves SSI. So a very interesting investment of a lot of money by some of the leading companies in the world, just because it's Ilia and his idea to do something different than everything we've seen so far. Another interesting topic has to do with AI laws and bills that are put in place. So a staggering 781 AI related bills are pending in the US states, and. Federal level since the beginning of 2025. So to put things in perspective, the entire year of 2024 had 743 bills. And again, now we already have 781. We're not even at the end of Q1. To compare that to 2023. 2023 had under 200 of these bills. So the urgency is very, very clear. The current administration at the federal level is taking a very clear approach to let them run as fast as they can because we need that for national security. And so it leaves a lot of work at the state level, which is not necessarily good, but there are multiple attempts across multiple states to define what is ai, what is the risk, what are the limitations, what it can and cannot be done on the government level, what can and cannot be done on industry level and so on and so forth. This will probably create a whole crazy patchwork of laws that will make it harder and harder for companies to navigate. This may slow down innovation of companies in different states and may force companies to move from one state to the other maybe more than once in order to be able to keep on doing what they're doing. That is not a good thing, but again, the current political situation on the federal level, we'll probably push this even further. If you remember, middle of last year, we talked a lot about SB 10 47, which was a AI safety bill that was raised in California and actually passed a vote, but was vetoed by the governor because he thought it was slow development. It was slow innovation, which is one of the most economical engines of the Bay Area and California as a whole. What will happen in other states, I don't know, but it's definitely a story worth following. Another interesting big piece of news is Larry Page one of Google's co-founder and the world's eighth richest person just launched another company. The company is called Dy Atomics, and the goal of Dy Atomics is to harness AI in order to highly optimized design and shorter the time from idea to creation of products in the manufacturing and factory production world. I think what we're going to see is more and more such initiatives with people with background in tech and with money that will team up with other people with niche knowledge in specific area. in this case, designed for manufacturing and will create things that will dramatically revolutionize entire industries that have been running roughly the same for decades. Now. We've been talking a lot about Google in the past few months. I told you since the beginning that I think Google is gonna end up being the front runner of this whole revolution because they have everything they need and they've been at work releasing more and more stuff, more or less every single week for the past three to four months. Some of them were more significant, some of them were smaller. Right now Google actually lost the top two places that they held at the chatbot arena. So right now number one is Grok three, which I actually really like and I use a lot behind that. GPT-4 0.5, but they're still holding the next two spots and they're holding four out of the top nine ranked models, including one of their latest releases, Gemma three, which we're gonna talk more about in a minute. But first, let's talk about some cool new features that are finally coming and are putting together piece by piece, the big promise of AI within the ecosystem of a business. So the first one is they just added Gemini powered side panel to Google Calendar. It was available in Google Docs and Google Sheets and Google Drive. So now there's a side panel for Gemini in Google Calendar as well, and it allows you to ask questions such as, when was my last meeting with John? Or How many meetings do I have on Monday the 23rd? Or show me when do I have openings on Tuesday? And stuff like that. And it will give you immediate answers without you having to actually go and browse that. So think about a scheduling personal assistant that knows everything about your calendar and can provide you immediate answers. Now, this is not available on its own. You have to sign up for it on Google Workspace Labs to test it out, and then you have to activate it per user. And it's only web-based. There's no mobile support for it yet, but I see that as a very big promise. Another really cool feature that is coming to Gmail, is a button that says add to calendar. So when somebody sends you an email suggesting a meeting, so it's not a meeting invite, but you're saying, what do you think about jumping on a call on Monday at 3:00 PM. Once it sees that, it will add a add to calendar button. If you click that, it will actually open a sad panel that will fill out a meeting details with everything it knows from the context of the email. And all you have to do is click yes and it will create the meeting for you. When I mentioned earlier that this is the promise of everything, the whole point of having a unified Gemini or copilot in your workspace is that it knows everything from everywhere and can start connecting the dots. I thought that we will be there by the end of 2024. We're very far from that, so I was very wrong on that particular prediction. But these are early signs of exactly what might happen, where the system will be able to connect the dots between different aspects of your entire work ecosystem, whether you're in the Google ecosystem or the Microsoft ecosystem. And we'll be able to do these really cool things for us, like managing our time and connecting from what was said in an email to what was said and summarized in a meeting to what's available in our CRM and be able to work all these things together, turn that into agents, it will be able to actually handle a lot of the work for us as a really busy CEO that's currently running two and a half businesses, I can tell you I can not wait for this to finally be here, but early signs are here and that's very promising. Now, on the bigger, more strategic side, Google has unveiled Gemini embedding, so those of you who don't know what embedding is, the way the whole AI works is on vector databases, which is a way to store information as a bunch of numbers that define a vector. And this is how the AI works in order to search for relevant information on specific topics. So to take text or information and turn it into a vector database, use a process called embedding. So Google just announced a new embedding tool that actually does that faster and better and cheaper than current embedding tools that exist out there. And they're saying that in addition to that it can handle significantly bigger text chunks than previous tools. And it doubles the embedding language support to over a hundred different languages that it knows how to handle. So for any company or individual who wants to develop new data, this is a great new tool to go and test out. Google also rolled out its Gemini two powered data science agent on March 3rd, integrating it into CoLab. So Google CoLab has been around for a while. It is a great free environment to develop and run and test Python code for everything you need. But most people do not know how to create Python code or troubleshoot Python code, myself included. So this new version allows you to prompt Gemini, basically in pure English to create a solution for you, and it will write and debug the code. I've actually done an episode about this on how to use CoLab. And it's going to be released this coming Tuesday, showing you examples of how you can do really advanced, really cool stuff, including bringing data from third party resources, combining it with your data and coming up with whatever kind of output that you want, whether it's graphs, text charts, maps, or other stuff within minutes, not knowing a single line of code. When I told you a lot of stuff, a lot of stuff coming from Google, I was serious. A big release from Google this week is they just launched Gemma three. Gemma is their line of open source models that are smaller and faster and open source that are built on their larger Gemini models. So the model that just released Gemma three is actually very unique. They're calling it, and I'm quoting World's best single accelerator model. And the interesting thing is that it achieves similar results to tools like Deep Seeq and Meta and Lama and even some of the Open AI models running on a single GPU. Now, in addition to the fact that it's very capable in running on a single GPU, it has 128,000 tokens context window. So those of you who don't know what context windows are, this is the amount of data you can put in and out in a single chat before it starts forgetting anything. The previous Gemma model had only 8,000 tokens context window. So this one has 128,000 tokens, which is a huge jump, which it actually makes it a lot more useful. It's also multimodal, which is incredible, and it comes in three different sizes, like a lot of these models with 1 billion, 4 billion, 12,000,000,020 7 billion parameters. The idea of obviously running it on a single GPU is making it a very thin and small model that can eventually run on devices like on your phone or an wearable device and so on. It was developed using distillation, so the same ways that Deep Seek was developed by using bigger models in order to learn how they work and providing it with smaller data and yet good enough data to do most tasks much faster and much more efficient. Google also made some big updates to the models, both three and paid that was not available before. Free users of Gemini got access to long-term memory, meaning the tool can learn stuff about you, your company, your context, your job, your family, everything it needs to know in order to provide you better and better answers. And you can control that information under the saved info in the setting so you can delete stuff that you don't want it to remember about you. You can also disable it completely. Two really powerful new capabilities that are coming to the Free Gemini model is deep research and gems. So we've done an entire episode comparing the different deep research tools last week, and you can go and check that out. But deep research allows you to do well, as the name suggests, deep research and get a very detailed report. So this was only a paid functionality until this past week, and now anybody can use it on Gemini. And the other one is gems, which is Google's parallel to custom gpt. So if you want to develop custom GPT on the open AI universe, it's gonna cost you at least 20 bucks a month for their membership because it's not available for the free users. And now gems are available to free users where you can develop these mini automations with a specific knowledge base to redo specific repetitive tasks that you need to do a lot of. The only way to do this for free so far was perplexity spaces. So now we have two places where you can develop these media automations for free. And again, Google is deploying more and more stuff into this universe. Another interesting thing that Google just started doing is they're marrying Google Gemini Chatbot with your personal search history. So this new experimental feature, looks at your search history on Google and tries to learn from that about you and your pattern and what you care about and they things that are more and less important for you and use that in order to give you better and better and more personalized answers on Gemini. Now, right now you can enable it via the dropdown menu in the Gemini app, but I assume over time this will become like a backend thing that will be a given and will have very little control over, and that's not even the end of it. They're saying that this is just the beginning. And Gemini will also peek into your photos and YouTube activity to get you more and more personalized, contextually relevant content on the Gemini chat, which is again, a huge benefit that Google has that nobody else has, which will allow them to stay ahead and give us more relevant information as users. And the final cool little feature actually comes to the paid universe. I've been using it for a while on the Google AI studio, but now it is going to be available to the Gemini advanced users, which is live mode with vision, which allows you to actually open the camera and show Gemini the world, either as an image or in real time, actually video streaming, and ask it about anything that it can see and get relevant information about it. This could be sharing the camera, this could be sharing stuff on your screen. I've used it for a while now, again, on their experimental environment, and it's actually working really, really well. Every time I try to do something longer or more sophisticated with it. I found that it was struggling and in many cases crashed and I had to start from the beginning. But again, it's a very promising start that shows where we're going, whether through wearables or sharing our screen With these devices, combining with their ability to actually take action will be an incredible force multiplier for anybody who's using a screen or wearing a device that can see the rest of the world. Last piece of use on Google has nothing to do with features for their AI tools, but the Department of Justice is continuing what the previous administration started, which is to divest Google from Chrome. The only thing they've changed from the previous administration is they're not going to force them to stop investing in other companies when it comes to ai, which was part of the original ruling. And so this is good news for Google's investment arm. And it's also really good news for Anthropic that have enjoyed$3 billion from Google so far and is already have a convertible note for another 750 million as an additional investment. Switching from Google to OpenAI. So OpenAI announced that Oracle will install 64,000 Nvidia gb, 2000 GPUs at the Stargate Data Center in texas and this is supposed to go live by 2026 as part of the Stargate Initiative. If you remember, we reported about this early in the year. Stargate was announced at the White House in January, targeting up to$500 billion coming from SoftBank to create Oracle AI infrastructure for open ai. So this is just the first step in that direction. They're also building a 360 megawatts natural gas plant for$500 million. That will provide the energy required for this data center. To further showing how OpenAI invests in infrastructure. They just signed a$12 billion deal with Core Weave. So core Weave is a GPU heavy cloud provider, and their goal is to have another option to send compute and this deal will provide additional compute to OpenAI through now a third channel. Now, OpenAI also invested$350 million as an equity investor in Core Weave via a private placement. There are rumors that Core Weave are just about to go public, and so this is probably one of the last private investments they are going to get. But what this shows very, very clearly is that OpenAI is investing heavily in diversifying its access to compute for their AI models beyond their initial relationship with Microsoft, which was their sole cloud provider until not too long ago. And you'll see in a minute, once we'll move to talk about Microsoft, that they are also slowly moving away from relying completely on OpenAI for their things. So the relationship there, I dunno if it's cooling down, but it's definitely diversifying on both sides of that relationship. Now going back to the whole Agentic future of 2025. Conversation that we had in the beginning opens the eyes Chief Product Officer Kevin Whale, announced on March 11 at the Human X Conference in Las Vegas. That Chachi PT will evolve into an agent AI this year. And the quote is, this is the year it goes from answering questions for you to doing things for you in the real world. And well continues with. It's not just how efficient can you make yourself, it's also how many agents can you have going off solving problems for you. So the idea that everybody's pushing for, and again, we're gonna see this with our own eyes. In the very near future is that every person, every employee will have an army of agents that he or she will control to do a lot more tasks than we can do today. Now, I mentioned that before, if you can use these kind of capabilities to grow your business exponentially, go do that. But the reality is not every market and not every company has an elastic market that allows them to grow indefinitely. There's other players, there's other forces that define how much a company can grow. And since you cannot grow, most likely in most companies at the same pace that these tools will allow it, the only other way to make more money is to let people go.'cause now you can achieve 30% growth with 30% less people. Guess what? You're gonna have 30% less people. What does that mean to the global economy? What does that mean to individuals and their ability to find a job that will actually allow them to sustain their current quality of living? I don't know. And the problem is I know that nobody knows, not the government, not the people who's pushing this forward. If you go and listen to last week's news episode, we talked about a lot about the fact that the government knows that this is coming. That they have the leading people of the leading companies shouting from rooftops, telling them that they need to do something. And yet not much is happening. Definitely not at the speed that the technology is moving and the impact that it can have on the workforce and society as a whole. I must admit, I wish I had an idea of what I can do to change this. My role in this as far as now is to keep you educated so jointly we can start thinking about this and maybe figure something out. It's very obvious that we're gonna have a world that is dramatically different within the next 24 months for sure. And that's very, very fast. As I mentioned with regards to the relationship with OpenAI, Microsoft is also doing a lot in order to diversify from its complete dependence on open AI's models. So they just announced that they're developing a family of AI models called MAI and that it's already nearly matching open AI's performance based on an article in TechCrunch. Now in addition, is that it's testing m AI models plus GR and meta and anthropic and deep seek as potential replacements for open AI's tech driving its co-pilot products. So far all of copilot is running open AI capabilities. Do they wanna push open AI completely out or they want to test additional options, to have a plan B in case something happens with a relationship in OpenAI? Or they will allow users to pick which models they wanna run in the background for different use cases, similar to what Perplexity is doing as an example. It's not very, very clear, but it is very clear that they're steering away from being completely dependent on OpenAI and that they're developing their own models as well as looking at other companies as options. In addition, it was announced that Microsoft AI division that is led by Mustafa Suleman is also training a reasoning model based on their M AI capabilities that again, will compete with the oh one and O three models from open ai. And from Microsoft to anthropic. Anthropic, CEO has sounded the alarm when it comes to espionage on their models. ADE has warned that spies likely from China are targeting their algorithmic secrets that are worth hundreds of millions of dollars and are hidden in just a few lines of code. In an event at the console for foreign relationships, he underscored how high the stakes are while this is not a new topic, it was mentioned by multiple people in the past. This becomes a national security thing where these companies do not have the resources or the knowledge to protect their data to the level that is required from a state level espionage. Again, this was known for a year now at least, but the fact that one of the CEOs is saying that out loud in a conference just shows how significant the situation is. Now those of you who haven't listened to last week's podcast, we talked about the super intelligence strategy paper that was put together by some of the leading people in the AI space in the world. They talk a lot about that particular topic that global security and national security of the US depend of being the leader in the AI space. And it's very obvious to the other side as well. And hence the espionage enhance the need for additional investment from the US government and western governments in general to help protect the IP of these companies from foreign companies as well. The other thing that we discussed last week is AI is more or less the first advanced technology that was not developed in support or by the Department of Defense in the us and hence, they have very little control or impact on how it is being developed and kept, which also makes this topic even more complicated to protect. Now, staying on anthropic revenue, anthropic annualized revenue skyrocketed from 1 billion at the end of 2024 to an A IR pace of 1.4 billion in just March of 2025. This is a 40% jump in less than one quarter. Now this announcement comes shortly after Anthropic raised. Its most recent funding at the beginning of this month where they raised$3.5 billion at a$61.5 billion valuation, and it shows the strength and the growth of their model. We talked about this when we shared, they raise that a big drive in their growth comes from coding. Most of the coders in the world prefer to use Claude in their coding platforms because it provides better results, and that's a big part of their drive. In addition, as I mentioned, Manus, which is the new Chinese viral agent, is also running Claude in the background. So Claude is becoming not just a tool for people to use, but a very strong backend for a lot of other platforms, which fuels their fast growth. Two very interesting pieces of news from Anthropic. One of them is anthropic. Researchers train their AI clawed in an experiment to hide its true goals, to try to see if they can crack and detect what it's actually doing behind the scenes. And they had four different teams trying to figure out what the model's real goals were in order to understand how models may try to trick us in the future by showing us one thing and actually trying to achieve a different thing. In the end. Three out of the four teams which had access to the backend of the model, were able to figure out what the model was trying to do. While the fourth theme that only had access to the API could not crack the case. So three out of four may not be bad, and three out of three, when they have access to looking into the models weights and what it's doing is actually pretty good. But they themselves are saying that this may not be the case. As AI gets smarter and spotting hidden motives will become harder and harder as the models becomes smarter and smarter. That being said, this is good news on two different aspects. One, that somebody's actually investing in researching that topic, and two, that they're actually finding ways to detect the secret motives of AI agents. Will that work the same way on. Gemini, Andra and Grok and Llama and so on. I don't think anybody knows because I don't think anybody else ran the experiment, but I really hope that Anthropic will share their findings with researchers around the world to enable them to do the same thing on all the models out there to keep us all safe from a science fiction, but potentially real future. And the last interesting and cool thing coming out of Anthropic is that Dario Ade, the CEO, predicts that AI will allow us to potentially talk to animals and definitely animals like whales and dolphins that have a more sophisticated and advanced language. And he's basically claiming that since AI knows how to find patterns in a language, then predicts how to use the language, it can do it for technically any language in the world, including non-human language. I actually predicted that in early 2024 in one of the episodes, and I was talking about that, saying that I think this is the direction we will end up going because of exactly what Dario just said. If you know how to identify patterns in language and then put them together in order to put a sentence together, you should be able to do that for any language. And it will be interesting to see if that A happens and B, if it happens for which animals and how deep the conversation can be. But it might move from Dr. Doodle and science fiction to actually people able to have a conversation with maybe not their pets, but with several different smarter animals on the planet within our lifetime and maybe within the next few years, if you've been listening to this podcast for a while, I like to end on a positive note that will show that the future of us may be better on one aspect or the other by ai. So a new company, just came out of stealth, it's called Inception Labs, and they enve what they call Mercury. So it's a Silicon Valley startup, founded by some big names. And their people that founded it are from Stanford and UCLA and Cornell. And it's a new way to generate language from a language model. So far large language models, were using a sequential, auto aggressive models that basically, guess the next word, every time image generation models don't work that way. They work in a diffusion model that actually figures out all the pixels all at once, and this is what these researchers were able to do. So Mercury basically processes, entire segments of text, entire documents, all the words at once. Instead of word by word. And it's allowing it to generate a thousand tokens per second on Nvidia H 100 chips. So this is their older chips and not even the top of the line. And it's slashing the cost of generating the text by five to 10 times compared to traditional large language models like GPT-4 0.5, and Claude 3.7. Now, mercury Coder, which is a coding platform, as you can tell, there's a theme in this particular episode and in recent times, either way, which was the first model that they released. Is Async co-generation and text responses. It's ranking number one on speed and number two in quality on the co-pilot arena, and it is outpacing tools like GPT-4 oh and Gemini 1.5 flash. Why do I find this important? Because every time there's a new way to do this faster and cheaper, it means we need less power and less compute, and we're polluting the world less and investing less money in AI in order to achieve similar or better results, which I believe is good for all of us. This Tuesday we are going to be releasing the episode about how to use data in your company in order to make data driven decisions with two different tools. Chat, GPT, and the new functionality in Google collab. Making data-driven decisions is one of the most critical ways to make a business more successful. And we now have capabilities that were not available to many businesses. And even the businesses who had access to it, you still had a small department doing business intelligence and you had to wait in queue for them to develop anything for you. And now any business person can do really advanced data analysis on any kind of data on their own. In minutes. And on Tuesday, we're gonna show you exactly how. If you are finding value in this podcast, please share it with other people who can benefit from it. Literally click on the share button on your phone right now. Think about three, four, five, ten people that will benefit from knowing what we're sharing and share the episode with them. Also, I would really appreciate it if you can go and write a review and give us five stars on your favorite platform, whether it's Spotify or Apple Podcast. It allows us to reach more people, which I think is important for everyone. So thank you if you are doing that. Thank you for being a listener of this show and have an awesome rest of your weekend.

People on this episode

Head Shot

Isar Meitis

Host