Leveraging AI

178 | AI is storming into higher education, Vibe Everything (not just coding), How to Learn AI, AI finds diamonds in Minecraft, and more AI news you need to know for the week ending on April 4, 2025

• Isar Meitis • Season 1 • Episode 178

Share episode

Is your business ready for a world where AI not only learns on its own—but teaches others to do the same?

From top universities using AI to teach students critical thinking, to self-spawning AI agents, to a digital brain finding diamonds in Minecraft, this episode is your all-access pass to what just happened in AI—and how it affects your business today.

If you think "vibe coding" sounds like a buzzword, wait till you hear about “vibe marketing,” “vibe teaching,” and the not-so-distant future of vibe-everything workplaces—where humans simply speak, and AI executes.

In this AI news of Leveraging AI, you'll discover:

  • How Anthropic’s Claude is revolutionizing higher education with a Socratic AI model
  • Why OpenAI’s free ChatGPT Plus for students may be less strategic than it sounds
  • What “vibe coding” is—and how it’s already bleeding into marketing, sales, and beyond
  • The fastest ways to upskill in AI (and why OpenAI Academy + business-focused training both matter)
  • OpenAI’s $40B raise, its for-profit pivot deadline, and Sam Altman’s next power moves
  • The rise of real-time AI agents and why agent orchestration is about to be your next business superpower
  • Robots in Amazon warehouses and Audi factories—labor shift or labor shock?
  • AI-generated invoices for fraud? Why metadata isn’t enough to stop deepfake accounting
  • Runway’s Gen-4 video tool that keeps characters consistent—Hollywood, beware.
  • MIT and Carnegie Mellon’s new findings that may break the “bigger = better” myth in AI model training
  • AI finds diamonds in Minecraft. No training. No help. Just... instinct?

Bonus:

🚀 Learn more about the AI Business Transformation Course (link in show notes) — now $150 off with promo code HAPPYBIRTHDAY for a limited time. Perfect for business leaders looking to implement AI step-by-step, not just talk about it.

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!

Isar Meitis:

Hello and welcome to a Weekend News episode of the Leveraging AI Podcast, the podcast that shares practical, ethical ways to leverage AI to improve efficiency, grow your business, and advance your career. This is Isar Metis, your host, and our main topics this week. The first one is going to be how the AI wars are now spreading into the world of higher education in the us. Our second topic is going to be about the concept of vibe coding and how it's spreading into other aspects other than coding and how it might spread in the future. And the third one is going to be about new opportunities for AI learning and training for employees and individuals. All very important topics. And then we have a very long list of really amazing rapid fire items, including some staggering funding rounds a lot of other cool things like AI winning in Minecraft. So let's go. Earlier this week, Anthropic launched what they call Claude for Education, and they have done this in collaboration with several different universities. The biggest one is Northeastern University that has over 50,000 students across 13 campuses, but also London School of Economics and Champlain College and some others. And the idea behind this partnership is that AI will be used in exactly the opposite of what you would expect. Instead of providing students with answers, which is what's happening right now, it's actually taking a Socratic approach where it is going to push students to provide reasonings and thinks through the progress, asking questions like, how would you approach this problem and what evidence supports your conclusion and so on and so forth. Trying to push students to reason through processes rather than get answers from ai. I love this approach. I actually do this with my kids. When they are working on tests or problems they don't know how to solve, I actually show them how to use AI to help them solve the problems, but not by giving them the answers, but by asking them leading questions to help them get to the conclusions themselves. And I'm really excited to see that anthropic and universities are taking down that path. I hope this will evolve and will take over the entire education system, providing personalized and high value learning and training to everybody. I. Now, in addition, in this partnership admin staff will be able to use Claude for different policies and trends and streaming operations of different aspects of the things they need to do today. Just making their day to day more efficient, allowing them to focus more on well teaching. So overall, I really like this approach by anthropic. Now to tell you how needed this is, as of the latest research, 75% of universities still do not have even AI policies. This is based on a recent research by Stanford, so this shows you how far behind most universities are and moves like this hopefully will push everybody to move forward. Now, if you think OpenAI would stay quiet after a move like this. So immediately after this announcement on April 3rd, OpenAI announced that it's providing its ChatGPT plus license. The 20 bucks a month premium membership is now free for all US and Canadian college students through May 31st, 2025 to help them prep for their finals. In this particular case, there's no unique approach. It's just you can get the raw power of ChatGPT for free if you are a university student. So I don't think this approach even comes close to Anthropics, more strategic and better learning process approach, but it's definitely going to provide value to students, especially when it includes capabilities like deep research and voice interactions, and it can really help students study in a more effective way. As I mentioned, I really think that the Claude approach is much better, and I assume we will see ChatGPT as well as other AI tools, potentially open source ones following the same kind of path, becoming an extension of teachers and professors in actually challenging students to think on their own versus providing answers to them. That being said, allowing students to use systems as they will use them in the workforce as soon as they're done with university. Meaning the way we use deep research and so on is not necessarily a bad idea in the big mix of things. By the way, if you want your free access to the ChatGPT plus platform, all you need is to use your sheer ID and you get instant access. No cash. No cash, no nothing. Just knowing that it's gonna expire at the end of the school year. Our next topic, we're gonna dive into the expansion of Vibe Coding. So we spoke several times on this show about the concept of vibe coating. It's now taking Silicon Valley and a lot of other places by a storm. The person who coined the phrase was Andrej Carpathy who mentioned in a post a few months ago that he feels the vibe of coding as he's just talking to the computer in English and it's actually writing the code for him and he doesn't really need to write code and it spins off and solves problems and so on for him. And this concept is now being used more and more. And recently people started talking about vibe marketing. So beyond the coding world, there were no examples of people using this concept. But the reality is you can use this for anything. So in these particular exchanges about vibe marketing, basically saying, I can be just talking to the AI about what I'm trying to do and letting the AI do the magic of understanding who my target audience is and planning entire campaigns, including the small details in them in a similar way to planning and writing code. Now, when you think about it, it makes perfect sense because for 20 bucks a month or some cases, free AI can scale entire campaigns, including hundreds of ads. Compare therefore, pros and cons against synthetic concepts and potential target audiences. Provide ideas, suggestions, create entire mail drip campaigns, and more or less, everything that marketing people do right now can be done just by talking and experimenting and exchanging ideas and brainstorming with an AI system. Now, it's not just automating what you do. You can literally go back and forth with the AI to expand what you would've done on your own. We spoke last week about the research from Harvard and Ethan Molik about how humans with AI outperform humans working alone, and how teams with AI outperform teams of humans alone. And so combine this with the ability to kinda like be in the vibe and take your knowledge and combine it with the brute force and capabilities and brainstorming and ideation capabilities of ai. And you see where this is going. I definitely see a future, a near future where it's basically vibe everything. It's vibe sales, vibe customer service, vibe accounting, vibe teaching, vibe training. And then we're just gonna drop the word vibe because this is just going to be the way we work. We are going to talk to ai, and AI will work with us and perform a lot of the tasks more and more. Of the tasks in the day-to-day that we're doing, and we'll be basically orchestrating all these AI agents to do most of the work for us while providing us feedback and new ideas and becoming a partner in the creation process. This is already happening. I am using AI today in almost everything that I do, including brainstorming, ideation, summarization, execution, and so on and so forth, across everything in my two different businesses. And I help my clients do the same thing. And it's very obvious that AI is going to be embedded into every aspect that we do. And people's ability to understand that they can work with AI just like they work with a human, just by having a conversation with it and being in quote unquote, the vibe is the future. I think everybody will win. The third topic is how can you learn AI faster and better to actually benefit from all these new capabilities that are coming? And we're gonna talk a lot about agent capabilities, in this episode in the rapid fire section. But OpenAI just launched their Open AI Academy. And they're offering free training and technical guide and community support for different AI capabilities. It's actually pretty cool. It's free to join. You just need to create an account. It doesn't have to be the same account as your ChatGPT account. It can be any account that you want. And through there you get access to online courses in short videos that they already have between, 16 to 20 minutes long videos on specific topics. They're built into multiple different categories that they're relatively easy to navigate, but they're also planning to do in-person workshops that hasn't been published exactly when and where yet. But I assume they will start with the US in major cities and then go to more and more places, potentially through partnerships with different creators and people that are already doing it, or universities or other kinds of organizations. And their goal beyond the training is to create a network of experts and developers who can share and collaborate and innovate with AI while training other people who don't know how to use it. I think this is a very worthy cause. This is what I've been doing in the past two years. I really think that AI education and literacy and training is the key to getting a better outcome with AI for individuals, companies, and society as a whole. And so I'm personally very excited about this. And on the same note, we just opened the registration for our spring cohort of the AI Business Transformation course. I will tell you a little bit about the difference between this course from the free courses from OpenAI. Our course is geared for business people who wants to understand how to implement AI in their business. The entire eight hours, which are broken into two hours per week in four sessions are step-by-step process. Taking you from your current knowledge in AI and understanding of how to use it, including a lot of hands-on and practical use case examples, all the way to a detailed blueprint on how to implement AI successfully in a business across all aspects from strategy to systems, processes, training, et cetera. So it's a different approach than open ais where it's standalone units. This is really geared towards taking business people to the next step in business operation with ai. In previous courses, we had people from all around the world, and we've been teaching this course for over two years now, since April of 2023, with hundreds or maybe thousands of business people that have taken this course and has completely transformed their businesses. In addition, it helped build our community where we meet every Friday on our AI Friday Hangouts, and we just meet and talk about specific issues that people have and how to solve them and what's currently going on with ai. And so very similar concept of building communities around AI usage and learning, but focusing very much on business use cases. So if this is something you're interested in, you can join the course right now. There's gonna be a link in the show notes, and because yesterday was my birthday for one week only, you can use promo code, happy birthday, all uppercase, one word to get$150 off, which is the deepest discount we ever give. So if you are thinking of changing the trajectory of your career and or learning how to implement successfully AI in your department, or your business, we had people from all over the world, from multiple industries take this course and make dramatic changes with ai. If that's something that's interesting for you, the next week is a great opportunity to do that at the cheapest price possible. And now two hour rapid fire segment of the podcast. OpenAI just sealed the final steps of their founding round, which was pretty obvious before, but it closed on March 31st, and they've raised$40 billion with a B. This is the largest raise of a private tech company ever, and it gives them a valuation of$300 billion, almost doubling their previous valuation of 157 billion from just October of 2024. So about six months have passed and open AI's valuation, doubled this round was led by SoftBank, which is one of their major investors in the previous round as well, who are contributing 30 billion stake in this investment, and it's joined by$10 billion by a syndicate. That includes Microsoft, which was their biggest investor before SoftBank took over. Now to show that this may not be just Hype ChatGPT now has 500 million weekly users up from 400 million weekly users just last month, so in one month ChatGPT across the various platforms added a hundred million users. Or if you want a 25% growth in its weekly usage in one month, this is staggering in any numbers, but definitely when you're talking about hundreds of millions of people and they're projecting their revenue to triple to 12.7 billion by the end of this year. Now there's a catch in this investment, SoftBank full 30 billion hinges on open AI being able to transition from its nonprofit structure to a for-profit structure, which as we know is being challenged in court by Sam Altman plus some other people, and they need to do this by December 31st or 2025, or the total value of this investment is gonna be 20 billion instead of the current amount. Not to show you that everybody's surprised, including OpenAI themselves with the rate of growth after the release of the new GPT-4 oh image generation capabilities when everybody going crazy around the internet, creating images, mostly wasting their time on Ghibli version of themselves, myself included. But I created a lot of other actual business use cases as well. Sam tweeted the following, the Cha G PT launch 26 months ago was one of the craziest viral moments I've ever seen. We added 1 million users in the last hour, so OpenAI and ChatGPT has been the 800 pound gorilla the entire time, but it's just growing faster than everybody else right now. And the latest craze just shows that again, 1 million users in one hour compared to 1 million users in five days in the chat GPT moment. So awareness is definitely there. The new tools are definitely relevant for, people are looking for, whether it's for fun or for actual business use. And this doesn't seem to be the end of it. I think it's just gonna keep on growing. And and with this raise, OpenAI will have more money than more or less any other contender other than maybe Google to continue pushing forward with the things that they want to push forward. I. Now there's a lot of people are questioning the stability or the concepts behind this. S and p global warns that SoftBank massive bet may push them into debt and potentially can risk the entire company. But it's very obvious that the bet that they're gonna keep on making and that they're all in on ai. Some experts question the profitability of AI with how much money these companies keep on losing on regular basis. But the trend is very, very clear and the investment hype is, doesn't seem to stop. Now in an interesting twist of news, Sam Altman, as well as Brad LightUp, which is OpenAI's CEO, announced that the company is gonna release an open source, open weight model. It's first basically since G PT two in 2019, sometime in the next few months. And they're talking about a very capable, very powerful open weights model. Now this is very interesting because it was very obvious that OpenAI took a different direction after GT two and went to the proprietary closed model versus open models that we've seen from other different companies, despite their name being called Open ai which is one of the sources with their entire battle with Elon Musk, which we just mentioned, but there might be several different reasons for this decision. Decision number one is maybe the understanding that really open source is not such a bad idea, and Sam was quoted a few months ago saying that they might have been on the wrong side of history when they decided to go down the proprietary versus open model. But in addition, it is very obvious that open source models are gaining huge popularity, like the latest releases by Deep Seek and lama. And so I think Open AI just wants to play in both fields and if you want, have its fingers in both cookie jars, play the open source game and allow developers and allow the community to work with their models and not just with their competitors, and continue developing proprietary models as well. I think that's probably the main reason why they're doing this. It'll be very interesting to see what they actually release, how powerful it's going to be, and how competitive it's going to be compared with the other open source models. But it's a very interesting development from OpenAI that I personally encourage. I think being able to play on both sides is important, both for the open source community as well as for OpenAI as a company. Now when exactly it's going to be released, it's unclear, but the exact quote from the X tweet was, we are excited to release a powerful new open weight language model with reasoning in the coming months. So stay tuned. I will keep you updated as soon as this happens. Now in a very interesting development, OpenAI just partner with Notion, the popular productivity platform that has over 40 million global users. Many of them are great supporters and huge believers in the Notion platform, and they're going to roll out ChatGPT Integr starting this month. If you remember in the past two weeks, we shared with you that OpenAI is building integrations with external tools. And we talked about the fact that it's going to be able to connect to Google Drive as an example, and later on they're planning to do the same thing with Microsoft SharePoint. Which again is interesting just by the fact that they are huge partners with Microsoft. And Microsoft invested 13 billions in them and they decided to use Google as their first integration. But now they're also going to be fully integrated into everything notion, which will allow users to brainstorm ideas, crunch data, and set meeting summaries, project outlines, everything that you can do in notion you'll be able to do together with chat GBT built into the platform. Sam Altman said Notion's mission to empower people to shape their own tools aligns perfectly with our vision at OpenAI. And I agree. These two companies make perfect sense to work together. Companies like Salesforce are going to develop such tools on their own, but in the notions example, they're not as big. They probably don't have as deep of pockets, and they're just gonna integrate other people's tools into it. It'll be very interesting to see A, how good the integration is, and b, what's the adoption rate versus people just being used to doing things the way they're used to doing it over time? I think it's very obvious where it's going to go, but I think that people who will figure out quickly how to use it will be able to save a lot of overhead time of just setting up their tasks and following up and doing small things, and AI will be able to do some of those things for them. Goes back to proper training. Just by getting this new capability through Notion is not gonna make your company more efficient. You will have to train your people, build use cases, and make sure they stay up to date if you want to gain these benefits. Staying on OpenAI and talking more about their new image generation capabilities, which are absolutely incredible and have done multiple, really amazing business related use cases, which I will record a separate episode about. On one of our Tuesday releases, potentially maybe even this coming week, many people are taking this to the dark side and starting to doing negative things with it. And one of the use cases that went viral this week is people are creating fake invoices to fraud basically their own organizations or third party groups by submitting them as travel expenses or other business expenses. Some people went as far as adding like smudges of sauce and leftovers of food on the invoices to make them look more legit as they're submitting them. Like any one of these AI tools, it can be used for great good, and it can really be used for negative things. There are endless number of use cases on both side of that coin, and sadly, too many people jump very quickly to do negative things with it. I don't really know how open AI can block stuff like that because everything they're going to block, people are going to think about something else, and even if they are able to block it somehow, there are going to be open source models that are going to do the same thing. There was even a statement by a spokesperson by open AI that says all of the images include metadata indicating they were made by chat GPT, but even that's not really relevant because I can generate the image with chat GPT and then screenshot it and then the metadata is gone then I can use it wherever I want. Also, I don't think organizations know how to even check for that or not even thinking about it. Especially that in most of these cases it's built into third party tools that just scan the invoices and put them as part of the expense reports. And so nobody's checking for that right now, and I'm sure there's gonna be a lot of other examples of people using these new AI capabilities to fake documents. I see this as a very scary universe We're walking into. This is not new, it's just becoming more and more real. But in episode 13 of this podcast, back in May of 2023, so almost two years ago, I recorded an episode that was called The Truth Is Dead, how AI is Putting At Risk, the Trust that is the fabric of our society. If you want to get a glimpse of what I think the future looks like when there's zero ability to know what is real and what is not, go and check out this episode again, just episode 19. Scroll back a lot on your podcast feed and you'll be able to find it. But staying on the topic of image generation and switching to a different company, mid Journey Unveils Midjourney version seven, it's the first model they released in almost a year, and it's now in Alpha testing, so it's still not available to everyone, but it's starting to roll out and it's just a week after OpenAI is released off their chat. GPT Image Generation capabilities. One of the biggest new benefits is that introduces default personalization, which means you can tailor images to your taste by showing it examples of what you like as far as styles and so on, which is a huge benefit. There's also a draft mode that allows you to create images 10 times faster, at half the cost compared to the way you could do this before. So if you're just trying to brainstorm idea or get concepts and decide on the direction you want to go, that's a really great way to move forward. If you tried the ChatGPT image generation tool, you know that it's stupidly slow, especially compared to the fast mode in mid journey. And now this new draft mode is even faster. Now, two things that were anticipated and expected in this release are not excluded in this release, which is upscaling and retexturing, but they're saying that they're coming up in the next two months. So it might be that they made an early release of this just because of the ChatGPT release, trying to say that they're still relevant and still in the game, and they're gonna add a new version following that, that will add these capabilities. Now their CEO David Holtz is saying this is not just an upgrade, it's a completely new model from the ground up, and he's urging users to experiment and try and see how much better it is in coherence and understanding of prompts. I haven't seen any people use it yet. I don't have access to it yet, even though I'm a paying user. But I am sure in the next weeks we'll start seeing examples and then I'll be able to update you more on its capabilities compared to other tools, including the previous version of me Journey, open source tools like Flux as well as the ChatGPT image generation. I can tell you that the concept of the chat chi pity image generation, that happens with context of what you're actually working on, where you can iterate in a chat back and forth, and it understands the background information and everything that you're trying to achieve, is a lot more appealing to me across the board. And the best tool for that before that moment was Gemini that had really good image generation capabilities. I think right now Chachi PITI is at the head of that race with its really amazing understanding of everything that you're doing as well as being able to keep consistent characters and understand styles and so on. But as I mentioned, there's gonna be a separate episode talking specifically about that. And now let's switch gears to talk about Anthropic on May 31st. Anthropic unveiled an update to, its responsible scaling policy, and they're pinpointing specifically which AI models need extra security. And these are models that can be used by a, and now I'm quoting moderately resource state programs in crafting chemicals or biological weapons. So basically they're not just looking at individuals, they're now looking at people with significantly deeper pockets and capabilities that can leverage these tools to do harm. Now, as a quick reminder, back in October when they updated the policy, Anthropic started sweeping their offices for hidden devices to check for people who are trying to spy on what they are doing, as they are anticipating that this will be done either by competitors or by foreign countries. And they are trying to minimize the risk of people getting their hands on their proprietary secrets. And they're doing this from two different reasons. One is obviously to keep their ip, but also from a national security safety perspective to make sure that foreign governments do not get their hands on anthropic capabilities. Anthropic are also releasing something very interesting that they announced on March 29th, which is a multi-agent research mode that has been spotted in the code by other people. It has not been released yet. This feature let's Claude Spawn its own sub-agents for tasks. Think about like web searches, deep thinking and other tasks that Claude will be able to basically generate on the fly based on the prompts and the requirements from the user and the way it understands it. This is more or less what they've promised when they released 3.7 and they said that it's going to be a unified solution where the AI will understand your needs and will decide when to reason and what to do and how to do it, and will just perform the tasks in the most efficient way. A similar promise was made for GPT five, so it's very obvious that this is the direction that everything is going. It's gonna be an agent future where you will define the task and the AI will figure out what needs to be done, will spin off specific smaller agents that can take and do these tasks and perform them for you, giving you the output very quickly after significantly faster than humans can do the entire process. Now different than, let's say, Geminis Deep research capabilities in ChatGPT. The idea behind the Claude concept is that it's a multi-agent setup that happens in real time, creating these agents ad hoc for specific use cases, which then can collaborate together to potentially outperform the other tools that are out there. That being said, I think everybody's going in that direction, so even if Claude releases that first, I think everybody will follow shortly after. From Anthropic to X. So very interesting piece of news from Elon, but I must say not totally surprising on March 28th, Elon announced that his AI startup X AI, is acquiring X, formerly Twitter for$33 billion in an all stock deal, valuing the combined entity at over a hundred billion dollars. Now, Musk announced it in a post on X, suggesting that X AI value is 80 billion and X, formerly Twitter being 33 billion for a combined value of$113 billion. This makes sense from several different perspectives and doesn't make sense from another perspective. So the reason it makes sense is this deal allows. Elon himself and his investors in X in the$44 billion acquisition he did that he didn't actually want to do and was forced into doing, it allows them an out and get value in something significantly bigger. Teasing everybody and getting rid of potentially a lot of debt and a lot of risk. The other reason is that these companies work very closely together anyway. X AI is getting fed from x, formerly Twitter as its main data source, I assume, and it's providing a lot of AI capabilities into the platform for people to be able to search and engage and understand people's tweets. And so that makes sense from that perspective as well. the reason it makes absolutely no sense is the actual valuation because it values. X AI, the creator of Grok, which I absolutely love. I use Grok all the time, and I think it's a great tool, but I believe that X AI has really low revenue, if any right now, and it's valuing them at$80 billion compared to, let's say, anthropics,$60 billion valuation when they are currently generating$120 million every month. So what does that mean? it means that Elon was able to clean the cap table, merge two companies that he's managing anyway, get them into one unified environment from both a financial perspective and it will allow him and these companies to merge their assets as well as human capital to do more with all the assets that they have. So overall, I think this makes sense from multiple aspects and it will be very interesting to see how they do in the overall race. From X AI to Microsoft on March 27th, Microsoft revealed I. Two new reasoning agents for Microsoft 365 copilot. One is called Researcher and the other is called Analyst. They're basically a wrapper around the existing chat GPT tools. So researcher is basically chat GPT deep research, but with a very interesting direction. It allows you to search through your emails and meetings and web data as well as other internal data, which I think is brilliant and it's finally starting to go in the direction that we were all hoping it's going to go in 2024, where AI will allow you to actually see what's happening across your organization. I don't think it's there yet, but I think it's a big step in the right direction. And the analyst tool is basically chat's O three mini model, allowing to analyze data across multiple data sets that you're going to give it. So that's, none of this is new, but the implementation into the Microsoft environment is new I think it's needed and hopefully Microsoft is not gonna screw it up like they have with copilot. So far I have nothing against Microsoft. I just think copilot so far has been a poor shadow of the raw tools from ChatGPT that it is based on, and I don't think I've ever saw it outperforming just ChatGPT as is. So it will be interesting to see how this works. As I mentioned, combining it with internal sources is a huge benefit. Another interesting piece of news from Microsoft is that they're stopping or delaying their data center project worldwide scrapping plans in the uk, Australia, Indonesia, and US states like North Dakota, Wisconsin, and Illinois. And that's according to TechCrunch and Bloomberg. And they are claiming that this shift is thinking that they're overbuilding infrastructure. We talked about this when we covered an interview with Satya a few months ago where he was stating that there might be overbuilt of data centers, and they're thinking that leasing that data center space might be more cost effective than building it themselves. So it will be interesting to see whether their new bet, their new direction is the right one or not. Going from the US to competition from China. So Alibaba, who recently released their latest Quinn model, now released a new subset of that called Quinn, QVQ Max, which is a visual reasoning model that can see, understand, and think. So think about a combination of a model that can see and understand images in the world combined with reasoning, capabilities, images, and videos that help it solve real world problems. They're claiming that it can solve things such as math and physics and other visuals such as flow charts and diagrams and so on, which will align it with similar capabilities from chat GPT and leading models from the west. And it adds a new thinking button that can combine with uploading visual information to allow the model to reason step by step on what it's actually seeing. Now, they're also saying this is just the first iteration of bigger upgrades that are ahead and they're planning, handling multi-step tasks and complex problems in more of an agentic approach, as well as allow it to operate on computers and cell phones and play games and generate images. Basically trying to align and stay in the competition with the leading models in the world. Alibaba, very similar to Google, has everything it needs from data distribution, resources, capital, human capital, to keep on being a really important participant in this AI race. And now to our weekly robotics section. So Agility Robotics, like by X Microsoft executive, Peggy Johnson just captured$400 million investment at a$1.75 billion pre-money valuation. And that was released by the information on April 1st. Their robot called Digit is already working on some Amazon warehouses doing several different activities. Now leading this investment was WP Global, but you will also find names there. Not surprising. SoftBank is one of the investors and that's following$150 million they raised in 2022 and a total of 320 million they raised to date before this latest$400 million raise. Now the robot is five feet nine inches tall and it's priced around a hundred thousand dollars and there's already about a hundred units studies out there performing different tasks, and they're claiming covering for labor shortages, whether that's really all it's doing or actually replacing employees, it's very hard to know. But the trajectory moving forward is very clear. It will replace warehouse employees. Now there just recently released a new version that has longer battery life and autonomous charging and better grasping of equipment and so on, and this new raise will allow them to push the limits even further. Staying in the robotic fields. One of the companies that's been pushing forward very aggressively in this past year is China's unit tree. We talked about the robot G one several times. It's the one that did the kung fu moves and did different flips and the first to do back flips, and the first to do side flips and all these really cool things. But despite all that hype, they're delaying their household robot release due to, and I'm quoting stringent safety requirements. That makes perfect sense to me. I said that in almost every show that we talked about robots before. I know for 100%, not 99.9, that this thing is safe and that it's not going to harm my kids because it's gonna go crazy or think it's a good idea. For whatever reason, none of these robots is getting into my house. Despite the fact it's really tempting because the G one robot that is about four feet tall is priced between 16 and$20,000. So a very reasonable number compared to its bigger brother. As an example, coming from re called H one that costs$90,000 and we just talked about a hundred thousand dollars price tag from a different company. So for$16,000 you can get a small robot that can do a lot of chores inside and around the house. But again, I think safety here has to be. At a completely different level than it can be in a factory where it can work in an isolated area without human employees around it. Staying on the integration of robotics into factories, Audi China facility, Audi FAW just announced that it's going to be the first that is going to start using robots. They're using UB techs, Walker, S one humanoid robot for quality inspections. The first thing that he's doing is actually really interesting and it's detecting refrigerant leaks in the air cooling system, a task that was previously done by humans, but it's posing respiratory risk because these gases are toxic. So allowing this robot to check with, its very high level of accuracy of measuring devices as well as very quick ability to see and detect things while not risking humans is a great move in the right direction. These are exactly the tasks you do not want humans to do, and there was just no other better choice right now. So I think starting with hazardous, risky operations in factories are a great way to start using humanoid robots. But as I mentioned, as this thing evolved, it's not going to stop there. I. UBI Tech is planning producing 500 to a thousand of these robots by the end of this year and beyond. Audi. They already have started deploying them at BYD and Link and CO BYD is stating 120% improvement in sorting efficiency and Link and CO is claiming 40% faster warehouse processing and 65% labor cost cuts. Using the S model that kind of tells you where this is going. Faster, better, cheaper labor at factories that will eliminate jobs from blue collar workers. That's not a good thing from a global workforce perspective or from a society perspective, or from an economical perspective, other than for the companies themselves. But if the companies will not be able to sell anything because people will not have jobs and will not be able to buy anything, then there's no point in saving money in production. I don't know where this future is going, but I think we're looking at some seriously troubling times because of the trajectory and how fast this is moving right now. Switching from robots to video generation. And another staggering raise runway ai, which is one of the most advanced and one of the longest running AI video and image generation tools on the planet. Just raised$308 million in funding, valuing them at over$3 billion in valuation, doubling its worth from the previous round. General Atlantic was LE leading investor, but you will find again, SoftBank and Nvidia and Fidelity all are investing in this tool. So you see SoftBank are more or less betting on every leading company in the AI race. Now they just recently released Gen four, which solves maybe the biggest problem that AI video generation tools had, which is the ability to keep consistent characters across scenes and generations. And this is a dramatic change in the ability to generate real movies, videos, and series with that. And if you remember back in September, they signed a deal with Lionsgate to help them create video generations to support their actual work. So with this combination of things, the ability to generate consistent characters as well as partnerships with actual studios shows you that the direction in this field is clear as well. We will need less actors less. Camera people, sound people, and so on. And more and more work will be able to be generated by ai. I assume in the beginning will just compliment the traditional way of doing things and over time we'll just replace it. And I mentioned that in 2024, I think in 2025, we will see the first series, or the first movies that will be released done completely by ai. That will be provided to an audience that will be wanting to see it and will not care, or will actually be excited by the fact it's a hundred percent AI generated. I still think this is coming and I'm just waiting for the creator to actually make it happen. Notebook. Lm, which is one of my favorite tools, just released another really cool capability, which is called Discover Sources. so far Notebook lm, you had to give it the sources that you wanted it to summarize or create into podcasts, et cetera. This could have been files that you have or links to different websites, and now you can ask it to find relevant sources for you. It's called Discover Sources. There's a button, you tell it what you wanna know about and you click the button and it gives you up to 10 relevant sources and you can pick with check boxes, which ones you want to include in your summaries, in which one you don't. This is already available for me, I dunno for everybody, but they said that they're rolling it out for everybody in April. Now they also added another functionality, which I don't really get. It's called, I'm Feeling Curious. And it basically allows you to find information on random topics that you don't really know anything about. And this is more of if you're just bored and instead of playing a game or watching a series, you wanna learn something new, I guess you can click that button and it will find you several different sources on a random topic and then you can pick and learn about it, through the platform. So I found this functionality not very helpful, but the first one is actually really great. Another really interesting feature is actually coming from Amazon. Amazon is now testing a new AI shopping agent called Buy For Me, which allows users of the Amazon app to search third party websites for stuff that is not available on the Amazon platform. So think about you're going to Amazon, you're searching for something, and instead of showing you other options, it's showing you the actual thing, but not coming from Amazon. The other cool functionality is in addition to finding it in several different sources and showing you the pricing since it has your credit card information saved on Amazon, it will actually do the checkout for you once you pick the one you wanna purchase. Now they're encrypting your billing information and obviously your credit card information. So it's perfectly safe to use per them. The only difference between shopping through the Amazon platform and using this tool is that if you need to return the thing that you're buying, you have to return it through the third party platform, and not through Amazon. But I see that as a brilliant move from Amazon to get even a bigger market share than their complete dominance that they have right now. But B, I think it's a very smart use of AI agents that allows you to shop through multiple websites from a platform you're doing most of your shopping anyway. This is a very clear move by Amazon to block or at least compete with some of the new functionality of shopping on ChatGPT that was released earlier this year. In addition, Amazon released a very interesting new capability called Amazon Nova Act, SDK, which is a toolkit that lets developer build AI agents capable of tackling over the web browsers tasks. So basically take over your browser and do multiple actions. And they release that to the public, not just on Bedrock, which is how they released our previous models. So it's an SDK that's available to anybody, whether you're using Bedrock on the AWS platform or not. This is, again, I'm sure comes to keep them relevant in this new frontier of computer control and browser control that is intensifying seems every single day. You can get access to it through nova.amazon.com if you want to test it out, and it's just available with text, images and video generation capabilities all available, not just as I mentioned on the Bedrock platform. Staying on the topic of agents augment code. Just released an AI agent solution. That wins 70% of the time over GitHub copilot Enterprise, basically allowing it to generate better code per developers 70% of the time compared to one of the leading code generation capabilities. This new tool, integrates Anthropic cloud 3.7 together with open AI oh one reasoning model, hitting very high results on various programming benchmarks. And this is just another tool in the AI coding wars that is now available to people to use. The other interesting thing in this tool is that has a$200,000 context window, which allows it to handle very large pieces of code. Now it also is SOC compliant from a safety perspective. And it integrates with tools like GitHub and Jira allowing people to just work in their current environment and not taking any extra risks while using it. And maybe the most interesting and the most exciting and the most scary announcement in the agent world is on April 1st. Emergence AI unveiled a new platform that auto generates agents in real time based on user defined goals. Basically, eliminating the need to code agents, users can put in their tasks and their goals in natural language, and the system will spin up the relevant specialized agents to do that task. Now, the CEO, Satya, Nita explained it this way, when a new task comes in, the orchestrator figures out if it can solve it by checking the registry of existing agents, creating new ones if needed, and refining them over time. What does that mean? It means complete democratization of agent generation. so far everybody hears about agents, but to create them, you needed to be a developer or have an army of developers to build them at scale. And now this is the first platform, and I'm sure others will follow. We just talked about Claude working on stuff like that. That will allow you to just prompt it and tell it what you're trying to do and explain your goals and explain your limitations, and it will create the agents in real time to do the tasks that it needs to do. This is a very significant step in the acceleration, in the creation and adoption of agents, and I think this will dramatically reduce the time to when we're gonna start seeing more and more agents in the actual workforce and in everything that we do. Now something very smart that they've done. It's built with specific guardrails and human oversight verification for the things that the agents are doing. And their CEO specifically said, you need to verify that the new agents spawned are doing the tasks that you want. So think about it as new employees that you hire, you still need to check what they're doing. You can't trust it completely, a hundred percent at least yet. And I think adding these checking steps in everything that you do with ai, and that's true from the basic things of having to write an email for you, but definitely wants to start getting into a agent world where these things could do more complex task with multi-step and taking action. We need supervision. And I think what's gonna happen over time will become more and more supervisors of tasks than performers of tasks as humans. And from practical agents to an interesting study from Carnegie Mellon and Stanford, they have found that actually over training large language models is reducing their performance. So beyond a certain size over training, an AI will actually reduce its performance. The study that was released on March 28th was showing that when they increase the number of tokens from 2.3 trillion to 3 trillion tokens in the training and huge amounts of data. They actually got a reduction of 2% in performance after fine tuning the model. And it was harder to fine tune. So they're claiming that this push to just get more and more data and more and more and more training to get better outcomes is not the right way forward. We heard rumors about this from a conceptual perspective at Q4 of last year, and this seems to be the first research at least that I'm seeing that's proving this scientifically. If you think about how long they've been training GPT five and the fact they haven't released it, even though they run at least two complete training runs to get it out there, each training run costing them conceptually half a billion dollars and yet they wasn't released potentially because this exact fact that the bigger the models get, once you get to these numbers of tokens, it actually becomes harder to make them actually work as efficient as the previous model. Forget about getting better, but that is pushing the new kinds of innovation with better and new types of algorithms and the reasoning models that we now, so like in the stuff that are getting integrated into everything that is doing and distillation, which is the concept that was used by deep seek to create their models as well as, we don't know what's coming next, but I'm sure there's gonna be new innovations that will be able to drive the scaling laws beyond just brute force regardless, we're more or less running out of new data because we use most of it training the models so far. Now staying on the research side, on April 2nd, MIT research, revealed a new framework called L-M-L-L-M-F-P, which allows to boost large language models, ability to solve long intricate planning problems, achieving 85% success rate on benchmarks, which broke those benchmarks and achieve significantly higher results than anything before. What it basically does is it guides the LLM to mimic human problem solving by breaking the tasks into smaller steps and using optimizing algorithm to call specific steps and break them down further and further, and then combine them together into the final output. And it allows turning vague prompts into actionable, detailed plans that can be executed. Now to show you how powerful this is with the framework it slashed the error rate by 50% compared to unguided large language models. And on some benchmarks such as the blocks world, this new framework nailed 88% of the task versus 33% of the underlying large language models that was used without the framework. So again, talking about innovation and new things that are being developed regularly, this is a great promise for people who are not great prompters. That basically shows that there's a path forward for just explaining in regular English what you're trying to do and having the AI understand it much better and be able to define for itself what it needs to do and achieve significantly better results. Now speaking on interesting milestones on April 3rd, futurism reported that OpenAI ChatGPT has passed the touring test. And the way they've done this is they had 300 participants run three-way chats with humans and ai. And at the end of each conversation, they had to guess which of the participants were AI and which were real. And Chachi, PT 4.5 was picked as humans 73% of the time. This is really interesting because the humans were picked 66% of the time as humans. Now, ChatGPT 4.5 dramatically outpaced other models in mimicking people with me, Lama coming second at 56% Cha G, PT four oh at 21%. And they also used Eliza Chatbot from the 1960s, with 23%. So it's showing you that these models are getting better and better at mimicking humans. And now, as I said, sounding more human than actual humans. Now, another interesting fact here that is very important for all of you who are thinking or already building chatbots for different things, without defining a persona, without telling ChatGPT, which role it's taking, its rate of success dropped to 36%. But when building a backstory, like who you are, what's your background, what you've done before, what's your knowledge is which industry you're in, that jump to the 73% success rate. That tells you that when you're defining personas for large language models, you're actually gonna get significantly better results and definitely better human-like results. We didn't talk about Manus, I think at all last week. So those of you who don't remember a few weeks ago, Manus, which is a Chinese company, rolled out a viral agent platform that really took China by a storm that allows you to run multiple operations on your browser autonomously. And now they introduced two paid subscription plans that will allow you to do a lot more with this tool and will allow them to obviously pay for some of the cost of running these tools. There's two tiers. One$39 that gives you 3,900 credits to run two tasks simultaneously, and also a$200, or if you want to be more specific,$199 per month tier that provides you with 19,900 credits and allows you to run five tasks simultaneously. They're also now introduced a mobile app that you can run this on your phone. This is just another interesting step in the direction of agents will do everything for us in our browsers and beyond in the near future. Now, there's been many tests including by tech run recently that's showing it's not perfect and that it's making a lot of mistakes. But I think the direction is clear from even just what we talked about in the last 10, 15 minutes. I love ending with a fun fact. Well, DeepMind's Dreamer, which is a new platform that they released, was able to find diamonds in Minecraft without being taught anything about the game. So the concept behind Dreamer is it's an AI that imagines future outcomes of what it is going to do and then make decisions based on that. So this way it was able to learn everything in Minecraft and be able to fly around and dig and do everything it needs to do, and was able to find diamonds within just a few days from starting to play the game again without any prior training or instructions. What this shows is it shows AI's capability to tackle really complex open-ended tasks with zero handholding and or training on that particular task versus. Pre-programming it or explaining it, the rules of the game, or in this particular case, it's a very complex game, right? It's an entire universe with rules. Now, I must admit, I'm on the fence of where this is going. On one hand, this is a really cool, as somebody who plays Minecraft with his kids and seeing my kids play a lot more Minecraft, so seeing, being able to see an AI basically learns it on its own and does very well in it very quickly is exciting. B, it shows its ability to do things way beyond playing go or chess, right? So this is how it beat humans in previous games, but this is a much more complex game. So it's very interesting. That being said, despite the fact it's a very complex game, it's still a game with very limited rules compared to the real world. Does that translate directly to the AI's ability to understand the real world? I don't really know. I think it's a step in that direction. But it is showing the ability of the AI to learn on its own and develop tasks for its own. So combine that with everything We talked about, this episode, about the new agent capabilities, and you're walking into a very different world than we know today where the AI will not need us to give its prompt. You will understand on its own what it needs to do, defining its own tasks and executing on them. This could be on one hand, really exciting, but also really scary. If you want to learn how to navigate this new future with AI and you wanna be able to apply it for your career and your business, don't forget to check out the AI Business Transformation Course. There's a link in the show notes. You can click on it right now and go and learn about our course. As I mentioned, it transformed hundreds and maybe thousands of businesses based on these eight hours of training. You can also join our community on our Friday AI Hangouts. There's a link for that in the show notes as well. We meet every Friday, 1:00 PM It's completely free, completely open to everybody, and we discuss real world AI implications and implementation. so if you wanna join us and learn more about ai, come and join us, and please recommend this podcast to other people who can learn from it. If there's one thing that's very clear that this thing is not stopping and that it's going to. Profoundly impact everything that we know, and the more people know about this, the better the outcome can be. So the role that you can take is share this podcast, click the share button right now, and share it with several different people who can benefit from it. Write them a short message. And while you're at it, it'll be really nice of you. If you also rank this podcast, give us whatever star rating you think we deserve, and write a nice message or connect with me on LinkedIn and give me any feedback you want about this podcast. I regular chat on LinkedIn with people who are podcast listeners who approach me. So thank you for all of you who are doing this, and if you want to chat with me, please do that as well and have an amazing rest of your weekend. I'll see you again on Tuesday.

People on this episode

Head Shot

Isar Meitis

Host