Hello and welcome to 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 have two fascinating topics to start with today, and then we have a long list of short items to cover. The first topic that I wanna dive into is Sam Altman interview on the Invest Like The Best podcast. It is a fascinating interview that revealed a lot on the mindset and the thoughts of Sam Altman as they are right now in the midst of all the craziness that's going on. So that is the first topic. We're gonna cover that, and the reason we're gonna cover that is that it has a lot of seeds that are going to be planted to lead us to the second conversation, which are two big letters that were signed by all the leading people in the industry. One is asking the government not to intervene and not to block open source models, and the other, which is a lot more interesting, a lot more important, and also a lot more dumb in my own personal opinion, is to asking the government to help the leading labs to slow down the AI development or pacing Or like they titled it, Pacing the Frontier. And so both of these are important, and we're gonna talk about those a lot. And then we have a lot of smaller items, and yet some of them are very important that we are going to cover. So let's get started. As I mentioned, let's start with Sam's interview. It was published on July 28th, so just a few days ago, and it covered many, many different aspects of what Sam believes right now. But the main critical things is what Sam thinks about where we are in the journey and what kind of impact that is going to have or can have on our lives on this planet One of the main things that Patrick, the host, asks Sam is about the incident that we covered last week with the OpenAI model breaking into Hugging Face's most secured environments, and Sam said the following: "We were evaluating one of our unreleased models, and it was supposed to be working in a sandbox, and it figured out it could basically cheat on the test by chaining together multiple zero-day exploits to break out of the sandbox, get access to the internet, and then break through multiple systems on the Hugging Face side to kind of get the answers to the test and look really good on the eval." He also added, and I'm quoting, "This is the first sort of security incident that I have felt very viscerally." Now, Sam said that as a result of this incident, they paused training. But he said something that is more interesting, which again will be tied to our second deep dive of the episode today, and he said the following: "We may have to pace the rate of development to give ourselves enough time for society to harden around some of these new capability levels, and trying to figure out how we do that in a way that does not feel like regulatory capture for anyone and also does not feel like collusion among the frontier labs. That's going to take some work, and it's important to get right." I agree 100%. Again, we'll see in a minute the letter that came out roughly at the same time, and we'll see what's not getting it right looks like, again, in my own personal humble opinion. But this is the first time I think I heard Sam clearly saying out loud that he thinks that the labs need to pace themselves on the rate of development, or this may lead to really bad outcomes Now they also talked a lot about AGI and what that means for jobs, and Sam said that he was completely wrong about jobs. And he said a few sentences. I wanna quote a few of them because I think they all paint the picture when we look at them together. So one of the things he said, "If we could go back to 2019 and show people our latest model, not only would they say it's AGI, they would say that the economy would have been Completely upended and that has not happened. And I think just from a kind of intellectual humility point, anytime you're that wrong and that confident, which I think we were as a field, you have to update Sam gave a few thoughts on why he thinks nothing really extreme happened, with the economy despite these models being so, so, so good. He said: "AI is just very jagged," and we've heard that many time before. And he continued: "It's like superhuman genius on some ways, like a dumb toddler on others." He also said: "You can go hire an AI consultant right now, or talk to an AI sales rep right now, or hire an AI engineer or whatever. Somehow, most people seem to still really prefer interacting with a human." Now, I wanna counter that in a second, but I will read a few more sentences first. He said: "Human values have value because they're human. We are deeply hardwired to care about people. We're going to care about what people care about." And I do agree with that. Now I wanna go back to the previous point and say why I don't agree with interacting with AI. I have lots of scars and lots of horror stories with working with customer service agents. One of my worst nightmares, literally, is having to talk to a customer service agent in Spectrum. Spectrum has the worst customer service on the planet, most likely, but definitely the worst one that I ever encountered. Another slightly smaller nightmare is AT&T. In both these cases, you have to talk to six or seven people in different departments. Each time they're gonna put you on hold, you're gonna wait. The people will not know what you said before, so you have to repeat that sentence again, and you'll spend an hour trying to solve a problem, which in many cases you just get tired in the middle and just not solve. And so you're gonna keep on paying them more money or suffer through the thing you're suffering just because of the pain of working with customer service. On the flip side, I had to do a procedure with my company this week. Doesn't really matter why. I'm not gonna dive into the details, but I thought the same thing, that it's gonna be really complicated. But I said, "Okay, start on the website. Don't call anybody." I couldn't find a way to make it the way I wanted to make it on the website. I called the number and got answered by a bot, and I'm like, "Okay, this is gonna be a nightmare until I can get a person on the line and get this resolved." The bot asked me what I wanna do. I told it exactly what I wanted to do. It said, "Oh, so you wanna do one, two, three?" I'm like, "Yes." Like, "Okay, how much do you wanna pay?" I said, "This and that." I'm like, "Do you wanna pay it from the regular account that's connected to our account?" I said, "Yes." It said, "Okay, so let me confirm. You wanna do this, you wanna do that account, you wanna pay it for this purpose, and you wanna do it today?" I said, "Yes," and it was done. The whole call took less than five minutes. The call was answered in about three seconds. This is by far the best customer service I had over the phone from any large company ever in my life. I was completely shocked. So what I want to say about humans wanting to deal with humans, if humans can ever be as efficient as a bot that is connected to every company system, knows exactly the regulations, what it can and cannot do, and can act across multiple departments versus talking to multiple people, I will take that bot over a human any day of the week for anything I'm trying to resolve online. Now, I may not be the average person, I don't know, but I think if people had the trade-off between dealing with the nightmare of customer service in more or less every large organization versus solving the problem in five simple sentences in less than five minutes, the answer will be very, very simple. I said that multiple times on this podcast. I think the world in which we have contact centers, both calls and chat, are over. Now, they're not over yet, but it's a dying breed that of people and companies that provide this as a service. This will stay only in white glove solutions that will be ridiculously expensive and will be able to provide very unique services that the AI may or may not be able to, uh, deliver, or in very secure, problematic environments in which nobody wants to connect the bots to all the different systems. Other than that, I don't see any scenario where people will want to deal with humans in these scenarios. So this is my personal disagreement and my personal, uh, example about what I think about what Sam said. But let's continue Uh, one of the other reasons he thinks that jobs are not going away or that he was wrong that jobs are gonna be significantly impacted by AI, he said, and I'm quoting, "A year ago, people said software engineers are cooked. It's done. It's over. That didn't happen. What did happen, though, is that the nature of a software engineer, how much they would do changed quite a lot. And you don't really write code in the traditional sense, but you do something that is very recognizably software engineering." And I agree with that 100%. Now, what I also can tell you is now I am somewhat a software engineer, even though I've never learned how to write code or how to design software. I did ran several different software companies. I'm a CEO of a software company right now in addition to running Multiply, so I understand software. But I think that is also changing, like the fact that people who do not have specific degrees or specific backgrounds can now do the things, not at the same level. By no means I'm saying I can compete with a software engineer on writing software. But I can tell you, I'm creating pretty sophisticated pieces of software that connect to multiple different things and have different layers of security and access and users and deployed across complex environments, and I can do it myself. I would like to share with you something important that you need to know, and it is we just opened the registration for our September cohort of the multi-agent orchestration course. We have been teaching these cohorts since May of this year, and I'm also teaching it privately to companies. Each of these cohorts delivers extreme value in teaching people and companies on how to build agent orchestrations that can automate practically every knowledge work in your business effectively by building the right infrastructure, by learning how to build the basic building blocks, learning how to combine them together, and learning how to connect them to your entire tech stack of your company. This is dramatically changing the way companies are working. And if you're still using ChatGPT or Claude by giving it instructions and then going step by step and really thinking that this is the top of the line, you have no clue what you're missing. And this course will take you from using AI like it's 2025 to what you need to use AI for, which is fully automating and building swarms of agents that actually run and do things for you instead of you having to run them. And because you're a listener to this podcast, you can get $100 off the course by using promo code LEVERAGINGAI100. Don't miss this opportunity. All of these cohorts are sold out, every single one we've done. And so if you wanna learn how to do this, if you want to completely transform your personal life, we have several different people who have changed careers. We have two people who found a new job. We have people who made personal changes that are very dramatic beyond just business changes. Don't miss out because this will sell out as well. Come and sign up right now before this one sells out too. And now back to the news. So that also changes, uh, the whole concept. I think the most profound thing that Sam said in the interview, he said the following: "In some sense, we are about to create a genie that can grant any wish. I think it is very important that the first wishes that we, the world, ask the genie to do benefit the world as a whole." I think that Sam deeply believes that. I just think in the current scenario of things, this may not be the path that he can take because of multiple other pressures that happen around him Now, one of the things that he said that is related to that, and again, will lead us to the next conversation, another seed that I'm planting, he said, Concentration of power with AI is a terrifying thing. I think a lot of talk about safety concerns is well-founded, and then a lot of it is about people that just really, even in a slightly subconscious, want to concentrate power. I'm terrified of a world where the very real fears of AI are used as a way to say only this small group of people can have it because it's too dangerous and only they understand it. But don't worry, they're going to make the right decisions for all of us. I don't believe in that. And yet, him and his company signed a letter asking the US government to intervene and take control over the rate of development, which again will lead us shortly into the next segment Now, another very important aspect of it that I tend to agree with is he talked about robotics, and he was asked about when will robotics, like the physical AI, will have the, quote-unquote, "ChatGPT moment." And he said, "It's not 20 years. I would say we get the ChatGPT moment of robotics in the next, like, two or three years." So he's talking about robots being able to do impactful work across everything that we do in the next two years, which is basically a few seconds from now on the big scheme of things. We're completely not ready for that. But he said something that I think is really true and really interesting despite all the negative aspects, and he said, "If the role of people in the world is to be like the actuators of AI in the cloud, very bad. So I think it's much crazier if we don't get it than if we do talking about robots that can do the physical work." So basically what he's saying, that in the worst case scenario, AI controls what needs to happen, and we actually do the work for AI in the physical world versus robots do the work for AI in the physical world, and we enjoy the benefits. I think this is a very interesting point of view that I haven't heard anybody else say before. And he was also asked, does he think that intelligence will become a commodity like refined oil? And the answer was, and I'm quoting, Intelligence itself, I would say yes." So let's do a quick summary, and will put the link to this interview. I highly recommend you listen to that. There's a lot of other stuff in there. He was asked about the first days of OpenAI and how did they get the first investments and the compute challenges and a lot of other stuff, uh, that I think is interesting to listen to. Again, not every day we get to listen to what these people think or what they went through or what they consider when they make decisions. But what I think the summary of all of this is, is that even Sam himself is a little scared, and maybe a lot and not sharing it, with the current capabilities of AI. He believes that AI can generate good in the world and that that good needs to be distributed. But right now he is, and a few others, are running this as a very, very small closed group. And despite the fact he thinks it's bad, this is what's coming, he also thinks that we should probably slow down. But on the other hand, he thinks that the benefits and that intelligence as a commodity is something that is already here so before we dive into the second segment, I want you to remember a few things. One, Sam thinks and believes that the pace of AI should slow down. B, he thinks that OpenAI should be and can be, and in his eyes is, still competitive in cost compared to Chinese open source models, which is definitely debatable. Uh, and also that he believes that concentration of power is really, really bad, especially when it comes to AI capabilities. So before we dive in, I want to give you a quick timeline of the things as they evolved in the past few weeks, so you have the big context, and then we're going to dive into the details and where we are right now. So on the days between July 9th and the 13th, the OpenAI model breached the, uh, closed sandbox and then reached Hugging Face production systems and hacked into them to get information out of them. Again, without any malicious intent, but it still did it. On July 24th, roughly 25 companies led by NVIDIA and Microsoft published a letter defending open-weight AI, trying to influence the government not to pause or stop and put any regulations on open-weights AI. We need to remember Microsoft is now all in on open-weights AI and the benefits. NVIDIA has been in this, uh, role for somewhat they are on the fence because they have both closed and open source models. Um, so we're gonna talk a lot about that and who did sign and who didn't sign and how are the responses related to it. Uh, July 26 to 27, Claude shared conversations that people shared by clicking the share button on Claude started appearing in Google search results. So not a real breach, but yet people were really surprised with that happening. July 27th, uh, Dario Amodei publishes Anthropic position on open weights, and the China Commerce Ministry accuses the US of AI hegemonism. On July 28th, the Pacing the Frontier letter is published with at that day 1,100 signatures. Now there's already over 1,300 signatures on that letter. We're gonna focus mostly on that in a few minutes. July 29th, Hugging Face and OpenAI releases a full incident postmortem, uh, that they investigated together On the same days, Anthropic discloses three further incidents that found their models as they were tested, has done their own cybersecurity breaches as it was being evaluated, broke out of sandboxes, and reached places it was not supposed to reach, including hacking three different companies. and August 1st, which is today, is the deadline that the government has set in the June executive order for the AI safety and testing frameworks. So this is kinda like where we are. So let's start with the open weights letter on July 24th. And again, I think the trigger for that was that deadline for the US government to come up with whatever, and it was the last week to influence any of that. So on July 24th, roughly 25 companies published an open letter urging the administration not to impose premature restrictions on open weight AI models. The effort was led, as I mentioned, by NVIDIA and Microsoft and Meta as well. OpenAI signed it just a couple of days later, and Anthropic and Google didn't sign it when it was released, and not even shortly after. Now, this came as reports started showing up that the administration was weighing in on restrictions on Chinese open source models following the release of the latest DeepSeek model and Kimi K3 and GLM 5.2, like all the latest models that are very, very close to the frontier and yet are open source with open weights that anybody can massage and change to their needs And as I mentioned, the August deadline for creating a US safety testing framework was potentially the trigger to push this letter out faster. On Monday, July 27th, China's Commerce Ministry accused US government of AI hegemonism. Basically, he's saying that recent sanctions, uh, related to model distillations and stated that for any action causing substantive harm to Chinese interests, China would take all necessary measures to safeguard its legitimate rights and interests. Now, what does that mean? It means that China is completely not on board with anything the US government wants to do that may harm the pace in which the Chinese labs are moving right now. Even if what they're doing is, quote-unquote, "stealing IP from the US companies," the Chinese government is going to protect them and protect the pace in which they're running right now. Why is this important? You will understand in a few minutes once we get to the second topic of the pacing the frontier letter. They took a few more days, and then Dario Amodei released his statement in a full blog post saying why he took his time to relate to the open source letter, and he said the following, and I'm gonna quote a few quotes Anyone who has read my past writings should know that I don't regard such bans as useful measure. But let me state it clearly so that there is no doubt. Anthropic has never advocated for a ban on open weight models In a different section, he says the following: my primary concern is the risk that authoritarian governments build AI models that are more powerful than those built by the US, and use them to achieve permanent military superiority or or perpetrate incredibly deep repression of their own people. My secondary concern is the risk that powerful AI models may be misused to carry out cyber attacks or biological attacks and may have serious alignment problems And he continues now connecting the dots, and he's saying banning the use of these models by US businesses does nothing to address these risks because bad actors are unlikely to be legitimate US businesses. It would protect US AI companies from competition, but that has never been my goal So what is he suggesting instead? He's saying we should not sell powerful chips or chip-making equipment to China, and we should crack down on the rampant smuggling and workarounds used to obtain access to such chips. We should crack down on industrial-scale distillation operations And he said all sufficiently capable models, open and closed, should go through mandatory safety testing So now we know very clearly, because he took the time to actually go deeper into this than just signing a letter, what Anthropic stand is and what Dario Amodei's personal stand is on this. So he's saying safety is a critical concern. We should figure out what these models do, and we should stop Chinese companies or any other companies from distilling US models and not give them access to chips in any way possible, because that allows them to close the gap and use it for things that he's afraid that may be allowed to use, either intentionally by the Chinese government or by bad actors that will get access to open weight models. I tend to agree with everything he is saying So now let's connect it back to some things we already talked about last week and talked about earlier today, the Hugging Face intrusion. So we had a ChatGPT model that was not released yet that was able to break out of its cage, out of its sandbox, get access to the internet and hack Hugging Face, which probably has pretty decent protections, and was able to get access to all the information it wanted in their servers, all of that without anybody telling it to do it. So there was no bad actors telling it to do this. It was just trying to get answers. It wasn't trying to do anything bad. It was just trying to achieve a goal it was set to achieve To put things in perspective, those of you who missed last week, approximately 7,600 attacker actions grouped into roughly 6,200 clusters were used in order to make this attack successful. This is nothing any company in the world is prepared for right now By the way, the problem was solved by using an open source model. So eventually, Hugging Face was using GLM 5.2 to stop the attack and understand what was going on. And the reason they did not use the Western closed source models from OpenAI and Anthropic is because these have limitations on what you can use them for, and the models themselves block them from addressing the attack in an effective way. So they've used an open source Chinese model to do this, which again gives you an idea that maybe the current situation and making it more extreme is not the right way to go Now, what we've learned in the past few days is that Anthropic had its own evaluation incidents beyond what we've learned in the past. So Anthropic disclosed that retrospective review of 141,000 cybersecurity evaluation runs surfaced three further incidents in a similar class, and I'm quoting, "We found three incidents in which Claude model reached the internet from within or while interacting with a third-party evaluation environment and then gained unauthorized access to real systems of three different organizations." I don't know if OpenAI is doing it just to show that their models are badass as OpenAI or to really share their concern about the current risks of the current new level of models. Either way, we now know that this has happened as well Now they've shared three incidents, and I wanna address just one of them, 'cause in this one, the argument that they made is the following, and I'm quoting: "Our evaluation prompt stated explicitly that Claude had no internet access, but misconfiguration left the machines that Claude accessed as part of the evaluation with live internet access." Which means the model didn't try to break out. The model assumed that it cannot get out and that it is acting with an environment that is a staged environment, and it didn't think it's accessing the real internet, but a simulated internet access because that's what it was told. I find this maybe even scarier that the model that is breaking out because it means that these labs, as advanced and as aware as they are, have no clear protocol on how to do this right, on how to actually have a safe environment in which they can test the next models before they're doing it. Now, think about what would we think if a lab that develops biological weapons or chemical weapons or atomic weapons, each lab that does that will figure out random ways to decide how to do their next test versus having extremely clear, well-defined guardrails, processes, procedures that are unbreakable because humanity is at risk. And people have been developing biological weapons and chemical weapons and atomic weapons for decades, and yet nobody died so far from something like this, which nobody really knows if it was really breaking out of a lab or it came to us from monkeys or whatever other animals, so can believe whatever you wanna believe. But that would have been one example of how bad this can look like. But think about how many labs around the world are actually doing this and how this one incident in I don't know how many decades has actually happened. This is now happening all the time to these labs, again, to both OpenAI and Anthropic and I assume other labs as well. We just didn't hear about it yet. And so we have to have a collaborative effort to make a solution that will keep this future development safe. I don't think we can stop the future development. I think we should find a way to continue that development in a way that will be the world, with very clear guardrails that actually work. Now, is this doable? I don't know. I'm not an expert, but there are many, many, many experts in the world, and if they collaborate, they can potentially agree on a solution that they can develop together and apply together to keep us all safer. Now let's continue On July 28th, a group of people submitted a letter to the US government, an open letter called Pacing the Frontier. That letter asked the US government to support an international effort to develop the technical and governance tools needed to deliberately pace the AI development. I'm gonna read the letter to you so you know exactly what it is stating. "AI could help create dramatically better future, but that outcome is not guaranteed. The world-leading AI companies believe they could be close to automating AI research. It is hard to predict exactly how much this will accelerate AI progress, but there is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems. To realize AI potential, industry, government, and society at large may need the option to buy time to address emerging risks, development security measures, and strengthen oversight. But each company and country is under intense competitive pressure not to unilaterally slow that acceleration. And today, the world lacks the technical and governance tools to deliberately pace frontier-wide progress." Building on work already underway to monitor frontier model releases, we request that the US government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development Now, as you know if you've been listening to this podcast, I strongly agree with the motion, right? I've said multiple times for a very long time on this podcast that I'm praying for the day that this will happen. Not that the US government will take control, but that there will be an international collaboration of everybody that needs to be involved in order to figure out where this can go and how can we get there safely so we can enjoy the benefits and avoid or at least reduce the risks and the negative outcomes. So I will start with that. I strongly agree with the general direction. Let's see who signed the letter. So initially, there was just over 1,100 signatures. Now there are over 1,300 signatures that include some of the top leading people in the top leading labs in the world. So the list includes chief scientists from the top labs of, uh, Jakub, uh, Pachocki, who is the chief scientist of OpenAI, Mark Chen, who is the chief research officer over at OpenAI, Jared Kaplan, the co-founder and chief science officer at Anthropic, Shenjia Zhou, the chief scientist at Meta AI, Shane Legg, co-founder and chief AGI scientist at Google DeepMind, John Schulman, chief scientist of Thinking Machines, previously an OpenAI co-founder so Dario Amodei, uh, CEO of Anthropic, Ilya Sutskever, the CEO of Safe Superintelligence, again, previously one of the co-founders of OpenAI, founders of Senior Safety Leadership, so Zarimba, I hope I don't butcher his name, who is the head of AI resilience at OpenAI Foundation. Jack Clark, co-founder and head of public benefit at Anthropic. Chris Olah, co-founder and interoperability research lead at Anthropic. Benjamin Mann, the co-founder of Anthropic. Jan Leike, Anthropic, formerly co-lead of OpenAI superintelligence team And so on and so forth. There's a very, very long list of names of people who are definitely heavy hitters in the industry who knows what they're talking about. So the most important people in AI research in the world today in the Western Hemisphere, at least most of them, signed this letter and agree with the motion Now, a few people that are clearly not there, at least as of right now, are Sam Altman, Greg Brockman, Mira Murati, Demis Hassabis, and Yann LeCun. So again, not everybody signed the letter, but people from their organizations that are senior people from their organizations signed this letter Now, going back to connecting the dots from things we talked about in this episode, Sam Altman said in the interview that was released on the day the letter was released, meaning it was recorded before, that he believes that we need to pace the progress, and yet he did not sign the letter, so that's a little weird to me I'm also surprised that Demis Hassabis, the CEO of DeepMind, did not sign it because he clearly stated exactly these things multiple times before Now, the letter drew immediate substantial reactions across the board on multiple platforms, from Reddit to X to different articles and so on. And there are arguments to both sides. I must admit that I felt really negatively when I read the letter, and I'll explain why in a minute, but I will give you some examples of other people who think, uh, like me The main arguments against the letter are the following, With the most common objection, which I completely agree with, that is saying that the people that are building these systems, that leading labs are asking the government to impose constraint on what they are doing. You can just stop it yourself, right? If you think that what you're building puts humanity and this planet at risk, just stop. Don't ask the government to stop you. Don't ask for regulatory approach. Just say, "What we are doing is really, really dangerous. Let's stop now, and let's convince other people that what they're doing is dangerous, and potentially get political pressure to then help do this." But stop yourself. Don't ask other people to stop you Another big argument was regulatory capture, basically saying that if the leading labs are going to have the government slow everything down while they're already ahead, it is mostly going to slow down the smaller competitors and new entrants into the market, keeping the leading labs ahead because other people cannot do one, two, three, four because these are new regulations The third argument against it is that regardless of what regulation and handshake from an international perspective is gonna put in place, it is not gonna put any real constraints on China or other actors that would not be willing to share exactly what they're doing behind closed doors. And hence, it is going to leave the US behind in the race, putting us at risk of having the Chinese or somebody else in the world with superior AI capabilities that could dramatically change the balance in the world The other interesting thing is that there are more than several different people who signed both letters. One letter is saying, "Don't put any constraints on open-weights models," and the other letter says, "Let's slow everything down." And that obviously doesn't make any sense Now, on the arguments supporting the letter, there were a few also very good arguments. One is the labs are on framing that saying that they have a serious coordination problem and they cannot have an agreement just between the labs. They need somebody to help them get to such an agreement They also clearly mentioned the US-China factor and how that puts pressure, as well as they said competitive pressure, which may mean different things. You will hear my opinion in a minute The other clear arguments for this was the incidents that were just reported about. So you had three, four different incidents in which recent AI models and not previous AI models doing stuff that just didn't happen before. And now we have a very real problem on our hand, at least from a cybersecurity perspective. But a cybersecurity perspective means it can hack into any real infrastructure system. So think about our power systems, our water systems, Everything that you know as modern life can be hacked right now and potentially influenced or completely shut down or destroyed by an AI model, whether if it's used by bad actors or without any malicious intentions like we just saw, but just trying to solve another problem and the electrical power in a specific county or state or the entire United States would be in the way of it achieving the goal, and it will find a way to shut it down So here are my personal thoughts on this after I had a few days to think about this. The first thing is one of my favorite quotes from any movie ever From the good, the bad, and the ugly, if you have to shoot, shoot, don't talk. Or in this case, if you need to stop, stop, don't send letters. It blows my mind that these people literally think that they are putting the future lives of themselves, their dear ones, and the humanity at risk. And what they're doing in order to potentially try to stop is they write a letter and just sign it, and they send it to the government. You are grown-ups. You run a company. You run some of the most successful, capable organizations in the world. You can make your own decisions. You can stop. Yes, there is very significant pressure. We're gonna get to that in a minute, but you can stop. And by stopping, I think you're gonna a much bigger impact than by signing a letter. For individual people in these companies, you can resign, right? If you think the company you're working for is doing something that puts humanity at risk or that does not align with your moral values, leave the company. Find another place to work that does align with your moral code and values and where you think you should be and investing your time and your genius brain that you're now being used for something you believe is really, really dangerous. So this is number one. Number two, I really think that people need to grow up. And when I say people, I specifically mean Sam Altman and Dario Amodei. Those two that hate each other's guts, that won't shake hands in pictures with other leaders in the industry, and that will avoid each other and will go at each other throats at any given point. If we get to the point that, again, both of them believe, and they both clearly stated that the future of humanity and the world as we know it is at risk, grow up. Sit in the same room, and by doing so, you will very easily attract other people to participate in the conversation. I would make a serious bet that Demis Hassabis, who multiple times in the past said that he thinks the current race is crazy and it's not what AI should be used for, will gladly join them. By having these three people in the room, that will definitely attract more or less anybody else in the AI industry to join them, most likely including Chinese labs. Now, the fact that these Chinese labs will join the organization means by definition that the Chinese leadership will join the conversation because they are in and a part of these labs, which means that we still don't know what's gonna be the outcome, but at least we know who's involved in the conversation, and does not require the US government to participate in that. I actually think having the US government participate makes it a lower chance that the Chinese leadership will participate. I will get into more of that in a second. I also think that it is not a government thing. I think the government might be the last organization that needs to be involved. I think everybody else needs to be involved. The research companies, the leading labs, academia, people who are experts on social and political and education and economy, and every different aspects of how our world operates today needs to be a part of the conversation. And in the end, including governments, because they are the ones that needs to eventually put the infrastructure and the budgets and the people and the resources in place in order to actually, once we get to a conclusion on how to do this, can actually do this right, rather having to figure out how to make this into laws and regulations and resources and so on inside multiple countries. But I think they are potentially the last ones after a more deeper, less political conversation actually happens. Now, the only argument that I can think about why the US government is required to participate is influencing China. I strongly believe that in the current situation, the US government has zero leverage over the Chinese government, and we've heard that before, right? The Chinese government, if anything, will retaliate to anything the US government is trying to impose on them, even if their labs that they're protecting are doing stuff that is currently stealing from the US companies. So I think that having the US government in this particular case is not valuable to get China to participate, but if anything, the other way around Now, speaking about China, nobody knows what China would do, right? They may say yes, of course, but do something completely different behind the scenes. They may say absolutely not, we completely disagree with this motion, and we don't think that what you're doing is the right thing, and we're not willing to participate, and we won't let our companies participate either. I have to assume that the Chinese governments do not want rogue AIs doing whatever rogue AIs can do, potentially more than people out in the West. And the reason for that is being the type of country they are, they want to control everything. And having a rogue AI do things that are not within their control, it is not something they want. Again, my personal interpretation. And so I do think that they may participate and definitely allow their companies, like the Chinese leading AI labs, to participate to make sure or at least try to control these AI capabilities from going rogue and doing the things that they want versus the things that the Chinese leadership want. Again, what they will do behind the scenes, nobody knows, but we also don't know what the US governments will do behind the scenes when it comes to military development or stuff like that And then there's obviously the money side. Hundreds of billions of dollars was already invested in AI development and in this AI race. And there are hundreds of billions and potentially trillions in potential revenue that is supposed to come in as an outcome of these investments. This is very, very serious pressure for these companies not to stop. This is the only real argument that I can think of. You have a board of directors that represents your investors, and they are the ones that can decide whether you stop or not, and they are representing the investors who invested hundreds of billions of dollars in order to make this thing come to life. So they're not just going to say, "Let's stop and forget about all the hundreds of billions of dollars we invested." This is the only issue that I see here, right? At the end of the day, there are huge sums of money, nothing like anything we've ever seen before, that are involved in this process. However, if we do not figure this all out, it may all come collapsing anyway. It may actually may be a much bigger collapse, colossal collapse, it will come to a very bad event that plug on everything or will lead to whatever catastrophe that nobody wants to see or even imagine. And so slowing down collectively might be the right way to actually recoup and recover these investments instead of hoping and rolling the dice or flipping a coin to an outcome that is the right outcome, not just from a moral future of the planet perspective, but also from a financial perspective So bottom line, I think the direction is the right direction. I think signing a letter and sending it to the government is pathetic. I don't have a better word to describe it. I think that these-- the leaders of these labs should grow up. I think they should take responsibility, and I think they should just say, "We're stopping. Let's get in a room and talk about this," and then see what happens. I think they need to get the buy-in of their investors. I think they will get the buy-in of their investors if they tell them the risk of what happens if they don't. So let's say something really bad happens out of OpenAI, and then the government or governments around the world ban OpenAI. Now you say, "Oh, they will never do that." Well, they just did it to Anthropic just a few months ago on something much, much, much smaller. And so this is a real risk. And so I do think that there is a way to get the investors to be on board with this as well, which then may hopefully eventually lead to what I've been hoping now, which is wide international collaboration to evaluate what this may do to the way of life as we know it and figure out the way to progress in a pace that we can control versus that will control us And now to some rapid fire items. I'm trying to be quick. There's a lot of really interesting things to cover. Uh, one of them is that OpenAI Chief Finance Officer Sarah Friar announced to employees that the company's annualized recurring revenue in July of 2026 exceeded its total for the entire second quarter. So one month, July, made more revenue to OpenAI than the three preceding months before that combined together What this means is that the very dramatic changes that OpenAI did early in this year in response to Anthropic opening a growing gap on the business and enterprise revenue were the right measures to take, and closing stuff, things that were more of shiny objects, and focusing on developing codecs and GPT work, basically copying what Anthropic was doing, is yielding the right financial fruits. This is really critical because they're planning on an IPO either later this year or early 2027, and having that growth and that strength in their financial capabilities is critical for that being successful In parallel to that, OpenAI is slashing the rates of their two smaller models. So GPT 5.6 Tera and Luna prices are cut up by up to 80%, and that is amid surging competition and the whole enterprise cost scrutiny that is currently happening. So effective July 30th, which is yesterday, the time this is recorded and the time you're most likely listening to this or some of you, um, just weeks after its public release, they are going to dramatically cut the prices. So the fastest, smallest model, GPT 5.6 Luna, is gonna be cut in prices by 80% to 20 cents per a million input tokens and $1.20 per a million output tokens And the mid-tier GPT 5.6 Tera model saw a 20% price reduction, now costing $2 per million input tokens and $12 per million output tokens. This is still significantly higher than the Chinese competition, but is significantly lower than it was just a few weeks ago, and to be fair, two days ago Now, the interesting thing here is that Luna, now again the smallest model, now delivers performance that is comparable to frontier class models from just a year ago at roughly six cents per task, nearly nine times faster and cheaper than previous versions of a year ago Now the top-of-the-line model, the GPT 5.6 Sol model price remains unchanged, but it now also offers a new fast mode that delivers up to two and a half X faster speed than the previous processing speed, and it gives it to you at at the same level of intelligence as before What does that tell you? It tells you that the competitive pressure is on, and that if you want the frontier model, you will pay for that a lot of money. But if you don't, you can find cheap, faster alternatives, and I mentioned that multiple times on this show. I don't think they have a choice. And the reason I think they don't have a choice is because Chinese models are there too, and they're providing significantly cheaper capabilities that are not at the frontier, but that are good enough for most tasks. And I think that's what we're going to continue to see moving forward, meaning more and more cheaper open source models fulfilling gaps, and then the frontier labs having no choice to provide similar competitive models to that. Another company that is pushing very hard in that direction is Microsoft. They are also focusing a lot on releasing cheaper models that are specifically customized to specific tasks. They just released a bunch of new AI models that are specializing in specific things such as image generation, voice generation, transcription, coding, data security, uh, et cetera. And all of those while reducing GPU costs by 50% to 90%. So I'm not gonna dive into all the different models they released, but all of them are optimized for very specific tasks. Some of it is a task that can be used by anybody. Some of to specific industries or specific clients that can tailor the models to their own specific needs in order to reduce errors and increase speed and reduce cost all at the same time And these new models that are designed for specific things by OpenAI handle up to 90% of tasks efficiently while reserving expensive frontier models like GPT 5.4 for 10% of the overall tasks. Which again proves my point that I've been talking about for a very long time. You do not need the frontier models for most of the tasks that you're doing today if you know how to use the existing models. In the Microsoft case, it's specifically trained, but you don't even have to go that far. You can use the existing models without training them, and maybe not do 90% of the tasks, you can probably do 75% of the task with models that are significantly cheaper than the top-of-the-line models out there. Staying on the same topic, DeepSeek just released a new Flash model of their latest model that matches OpenAI's performance at 60% lower cost. So DeepSeek has released version 4 Flash 0731, that tells it when it was released, which is a to its already very competitive AI model that scores nearly identical to OpenAI's GPT 5.6 Luna, again, the smallest model, while dramatically undercutting it on price. So the fact that ChatGPT cut that price makes no difference. I mean, it makes a difference. It now makes them a lot closer in price, but still, DeepSeek is winning by a big spread Another company that is playing the same game is Thinking Machine Labs. So Thinking Machines has released its Inkling model just a couple of weeks ago, and we shared that on this podcast. But now they're releasing Inkling Small, which is a quarter of the size of the previous model, but it matches it or almost matches the full version performance And so similar approach, build a leading model and then work your way backwards from there to develop a faster, cheaper model that doesn't give you a big decline in capabilities and outcome, but that runs faster and significantly cheaper and can run on more devices. And this has been the trend in the last couple of years, and it's just accelerating dramatically right now with everybody's push to save money on tokens Now speaking on model releases, there have been a bunch of other model releases that do not focus on price optimization. The first one is Black Forest Labs has launched Flux 3. Those of you who don't know Flux, Flux is one of the most advanced, uh, image generation capability tools out there. I love it. I use it a lot across multiple things that I'm doing. Uh, and now they've released a new model that is a new multimodal foundation that is now already available for early access It also now has top-tier video generation performance, and its early evaluations is showing it generating videos that outperforms Runway Gen-4.5 and Luma Ray 3.2, and it's capable of generating 20-second videos and native audio from text, from images, from existing footage They're also integrating their models into robotics and physical AI with what they're calling Flux Mimic, which is a tool that allows robots to predict the actions they need to do and enable them to run in a more effective way So a highly capable AI model in the visual space that started as a model that only generated images, now generates images as the best of them, generates video better than most of them, and is now fully focused and integrated into physical AI in order to do things beyond just generating images. We talked about a similar path from Midjourney with their body scanners that they're planning to build. So again, these are companies who started with very highly capable visual capabilities that are now merging them into other applications in the real world in order to make more money and provide more revenue sources beyond just image generation, which obviously is very hard to compete in. But Flux is open source, so again, you can do a lot of cool things with it, uh, with just the image and video generation models Staying on the visual side of things, MiniMax just launched MiniMax H3, which is a unified multimodal video generation model that consolidates five separate AI tasks into one compared to the previous models that they had. It unifies text, image, video, and audio into a single unified cohesive system that generates 2K video outputs in four to 15 seconds And again, it can do it all beginning to end, including the sound effects, including the speech, including everything, including references of images and so on. And from a price point based on third-party evaluations, it costs about thirteen cents per second, uh, generation, which is actually really, really cheap. It's about two dollars for a fifteen-second full video clip in a native two K resolution. This is very competitive Now we mentioned a model for robotics. Google also is playing in that field, and now they released Gemini Robotics ER2, which is aimed to be integrated into advanced embodied reasoning model designed to act as a high-level brain for robots, enhancing their ability to understand, plan, and collaborate in the physical world. So again, Google, while not releasing Gemini 3.5 Pro that everybody has been anticipating for a very, very long time, is continuing to release models in other aspects, either smaller models, open source models, and now we see a robotics specific model that is supposed to play in the robotics world. Going back, connecting it to what we said in the beginning is saying we're going to get the ChatGPT moment in robotics in the next two years, you're seeing more and more advancements in the models, huge investments in this field across the board, all around the world. So I tend to agree with Sam that we're going to start seeing very capable robots, uh, roaming in different places that we're not expecting to see within just a few years Now I should probably tie this to the article where we're talking about the letter to stop the frontier, but another thing that Anthropic has released this week to discover mathematical flaws in cryptographic algorithms which are being used to protect a lot of really important stuff. So one of the tools that are being used in the cryptographic world is called HAWC. I don't know what the acronym stands for, but Claude Mythos preview reduced the effective key strength of HAWC by half in just 60 hours. It's cutting the expected cost of a full key recovery attack by two in the power of 64 to just two in the power of 38. That is still a really, really, really large number, but it is a significantly smaller number, and it did this by finding a novel mathematical property called nontrivial automorphism. I don't have a clue what that stands for. But what it means, it means that the next step might be able to crack really, really sophisticated cryptographic codes and get into basically any secure data on the planet without having a supercomputer work on it for the next seven years And I'm gonna end with two items that relate to home usage of AI with two interesting, and I must admit surprising, or maybe not so surprising, uh, developments of things that we will be able to start buying in the near future. The first one is Apple is set to start selling a major smart home overhaul of seven new products that are built on the new advanced Siri AI. So this is a complete overhaul of the Apple Home series with all its different components and beyond. It will include things like a refreshed HomePod, Apple TV models, a smart display, and other capabilities, all powered by the new Siri AI, which, as you know, is running on Gemini in the background. They're talking about imminent product releases that is planned for the very, very near future, including a new Apple 4K TV The HomePod, a refreshed HomePod mini, and all of these is expected in the very near future running on iOS 27 And then the next home component is even more surprising, and it is going to come from no other than Martha Stewart So according to its own website and recent reports, Hint, which is the name of the new system, uh, or Hint Home Intelligence in its full name, is an AI-powered home management platform that has been co-developed by Martha Stewart and has officially launched nationwide on the Apple App Store. The platform is aiming to transform home ownership by integrating human expertise with artificial intelligence to offer personalized proactive guidance for maintenance, cost savings, and project management So the idea is to track everything that's happening in your house, everything that you're doing in the house, everything the house needs in order to help you maintain it in a more effective way, all of that under the Martha Stewart brand. I find this, A, interesting, B, exciting to see how this is gonna work out. I know of several different startups who are in this field of reducing maintenance costs and then connecting it to, uh, insurance companies in order to reduce your insurance costs and stuff like that. So a big name like Martha Stewart getting into this field will be very interesting to track. That's it for this week. There's a lot of other news into the episode that you can still read by signing up for the newsletter. There's a link for that in the show notes, and you can then get the newsletter every single week. It has everything that we're talking about during the episodes, but a lot more than that in additional links that you can follow and explore on just the topics that you're interested in. I wanna remind you for our September cohort of the multi-agent orchestration course. Uh, we started teaching this course back in May, and we had several extremely successful cohort studies truly, truly transforming, uh, people's lives and their businesses. The things that people are doing immediately, uh, during the course, while the course is still running after every session just blows my mind because they are completely changing people's lives and putting businesses on new trajectories by applying what they're learning every single session. And as I mentioned, you can now sign up for the September cohort if you haven't done so yet. And as listeners of this podcast, you can get $100 off by using the promo code leveragingAI100. That's it for today. We'll be back on Tuesday with another fascinating how-to episode that will teach you how to implement something with AI, the exact step-by-step process. And until then, have an amazing rest of your weekend.