| | | | | By Jeff Brown, Editor, The Bleeding Edge | | |
Before we dive into today's AMA, I want to say thank you to everyone who attended my Permissionless Investor launch event on Wednesday and wish a warm welcome to our new Permissionless Investor subscribers. | | We had a great time talking about this historic moment we're about to witness in the cryptocurrency industry and the wave of innovation coming to blockchain technology. And it's all just days away. | For those who couldn't make it and have an interest, there is a replay available for a short while longer. You can access that right here. | | Now, on to the AMA… | | Cerebras | | | Jeff,
I read about a company that will likely go public in early 2025, that has succeeded in utilizing the entire silicon chip to print circuits. It is expected to make the computations of NVIDIA's chip do the work at least 50 times as fast as their current chip. The name was something like Cereberes. Do you have any information on this? | | Also, we stunned readers need some investing recommendations to participate in the development of AI before it runs away from affordability. | | – Roy N. | | | | Hi Roy, | Yes, the company that you are referring to is Cerebras. We see that you're already a subscriber to The Near Future Report. In that case, I recommend you log into your subscriber portal and look under special reports to find my research report titled How to Profit From the Next NVIDIA for more analysis. | And for any Bleeding Edge subscribers who are interested in signing up for the Near Future Report to gain access to that research, you can do so right here or call our customer support team at 1(888)493-3156. | | Optimus in the Fields | | | Jeff, as a longtime reader of your Bleeding Edge, thank you for all your insights on what's new in the world and how it will affect the market. | | I have been driving a Tesla Model Y for 2 years now and love it. I live in a very rural area and have only tried the full self-driving when they offered it for a test 2 times this last year and I would say the update this fall was almost 100% right on. | | I have a question about the Teslas Optimus robot. Do you know if they are testing them in an outdoor setting like in a field picking fruit or walking in trees? I own a tree nursery and would love to test one in our fields if they are able to work outdoors. | | – Lowell D. | | | | Hey Lowell, | | That's absolutely fantastic. I'm so glad you've been experimenting with the technology. Experiencing it in real life is the best way to understand how powerful these latest advancements are. | | As for your question, yes, Tesla is already testing outdoors walking amongst the trees. Have a look at the short video below: | | | | Source: Tesla | | The most important aspect of this video is that the Optimus shown is just walking autonomously (i.e. it is not being steered by a human). Its assigned task is to improve its ability to navigate unsteady terrain. Said another way, it is learning. | | And I love the way you are thinking. After seeing this video, how hard do you think it will be for an Optimus to walk on flat ground in an orchard and pick fruit with its "hands" that have 22 degrees of freedom? | | Not hard. This is why I believe that Tesla is sitting on a $25 trillion market opportunity. | | The Future of Cybersecurity | | | Hi Jeff. | | As a longtime subscriber, I enjoyed your year in review and outlook for 2025 in The Bleeding Edge. Best wishes for a happy and healthy 2025! | | I would like your valuable insight into the telecommunications industry. Specifically, how are they going to address cybersecurity with the recent and ongoing hacking by nation-state-sponsored actors? It was serious enough for the FBI to issue advisories last month and before with the latest one asking them to improve their IT security since there is no way of knowing if the hacked telecommunications system is still compromised. | | What a sad state of affairs. Two things are rather ironic with this national security challenge. They are recommending that we use encrypted phone calls and text messaging software, and the compromised systems are partially due to the backdoor entry that the government asked them to put in after 9/11. | | The other area is the semiconductor industry. You have written and shared extensively about Intel and AMD and, to a lesser degree, about Qualcomm and other ARM-based semiconductor companies. Concerning AI and the new NPU architecture, what is your insight on the leader or top two leading companies and their architectures? | | I'm looking forward to reading more about the exciting world of technology in the new year. | | – Raymond F. | | | | Thank you, Raymond. Our Brownstone community is in for what I believe is going to be an incredible year in 2025. I'm very excited about what I'm seeing in the near future. My team and I are going to be very busy this year as a result. | | The topic that you raise is so critically important to national security. Not just the U.S. but all Western countries and even developing countries. Very few understand the severity of the problem and how bad the security breaches have been. | | China-based Huawei is one of the largest telecommunications equipment manufacturers in the world. Its business was literally built upon the theft of Cisco's routing software that powers Cisco's telecommunications business. | | It jumpstarted Huawei's telecommunications business and saved tens of billions in research and development costs by doing so. Huawei was then able to undercut Cisco, Juniper, and Alcatel-Lucent on bids, growing the company into a global player. | | The problem was that Huawei was caught surveilling all data traffic running through its routers and sending that data back to China. | | The industry has known this for a long time, but things came to a head in late 2022 when the Federal Communications Commission (FCC) voted unanimously to ban all China-made telecommunications equipment from Huawei and ZTE. | | Most Western countries have followed with bans, especially for the use of national wireline or wireless networks. | | But it's not just the telecommunications equipment. Just this morning, the Supreme Court voted to uphold a ban on China's TikTok application in the U.S. The U.S.-based app must be sold or shut down. | | Those opposed to this have been screaming about "anti-China hysteria" or "racism" or "hate," but they clearly don't understand what has been going on. | | TikTok is a mass surveillance software application that collects data on both adults and children, monitors all their activity on their phones and computers, and sends that information back to China. | | The app also pushes data, video, and propaganda to manipulate U.S. consumers to achieve their political goals. This is a fact. China is an adversary of the Western world. TikTok is a tool for surveillance and psychological warfare. The smartest thing we can do is to make sure that TikTok is not on our phones or computers. | | I get it. It's handy, useful, fun, helps kill time, and some make money on the platform, but it is dangerous and should be banned. TikTok is a foreign adversary-controlled application and is a national security threat. | | The reality is that all of our messages, e-mails, and calls over internet protocol are encrypted, but if we're giving access to our phones to a foreign adversary, it defeats the purpose. | | The telecommunications equipment has been banned, we need to remove the surveillance apps from our phones and computers, and, when possible, make our calls over secure channels (for example, over a service like Facetime or Signal). | | And hopefully, to your point, over the next several years, the deep state will be dismantled and there will be a lot less unlawful surveillance of U.S. citizens by the U.S. government. | | The semiconductor industry – when it comes to artificial intelligence (AI)-specific chips – is very dynamic right now. The GPUs designed by NVIDIA and AMD are more general-purpose chips widely used for AI training. I refer to them as the workhorses in AI. This is largely because there wasn't much else to choose from five years ago. | | ARM-based chips (ARM is a semiconductor technology licensing company) are rising in importance, especially for applications at the edge of networks – where electronics products are being used in the field. Think smart-grid energy management in cities, autonomous vehicles, smartphones, industrial sensors, and automation in manufacturing. | | Qualcomm is pushing hard to try and enter the race. Intel continues to fail. And Apple has been designing its own NPUs which are manufactured by TSMC. | | But beyond these giant companies, there are some extremely promising fabless private semiconductor companies building semiconductors for both training and inference applications. And I expect we'll see both acquisitions and IPOs related to some of these companies in 2025. | | Naturally, I'll continue to provide more detailed research on AI-related companies and specific recommendations in The Near Future Report(large-caps) and Exponential Tech Investor(higher-growth small-caps). | | | | | | "When President Trump takes office on Monday, January 20, there is going to be shock and awe with executive orders… a blizzard of executive orders." That's exactly what Jeff Brown just predicted in this new interview he recorded in Washington, DC. He believes some of these executive orders will help trigger the biggest crypto boom ever… In these artificial intelligence coins. (Click here to get the details.) | | | | The Cost of the Next Great Leap in AI | | | Jeff, I'm a lifetime member and want to thank you for your solid research. | | Multiple headlines put forward the notion that AI is just completing its first phase (hardware), with promises of much more investments needed to come for the training and refining of AI models. | | In contrast to this, in the December 20, 2024, edition of the Wall Street Journal, Deepa Seetharaman wrote a very long headline article in the Exchange section of the paper, "The Next Great Leap of AI Is Behind Schedule and Crazy Expensive," discussing multiple difficulties in the 'training' phases of the various AI models… the lack of accurate information required to train AI systems, new approaches to how the training will be conducted, etc. | | As your time allows, can you share your take on the impacts on AI costs, schedules, and success if these problems become impractical or financially insurmountable? How great is the risk that the AI training will reach a dead end or the usefulness will be capped in some way? | | – Peter B. | | | | Hi Peter, | | There has definitely been a bunch of FUD (fear, uncertainty, and doubt) being promulgated about AI recently, almost all of it from the media and journalists, and almost none of it from the industry. | | One of the first things that I do whenever I'm reading/researching is understand who the author is and whether or not I should bother reading. Sometimes I'll read someone's work for a couple of years to determine if the analysis is worth following for now. | | The article that you refer to was authored by someone who has a BS/BA in journalism and history and zero experience in technology, semiconductors, or artificial intelligence. Some recent journalism from the author was about " Instagram's effect on teen girls and how AI systems struggle with racist and hateful content." | | This is not a source that I would go to for accurate analysis of artificial intelligence technology, the latest advancements in related semiconductor and software architectures, and projections on the future direction and/or success of the industry. | | Case in point, even before that article was published, OpenAI had already released some of the performance results of its o3 model in December. | | | | We can see measurable progress in solving the training challenges in the impressive performance improvements on related AI benchmarks, especially those related to reasoning. | | None of the major players are worried if they are spending $1 billion a year or $1 billion every six months to run their models. | | OpenAI has raised almost $24 billion to date. The company is projected to generate $11.6 billion in revenue this year compared to $3.7 billion last year. It's one of the contenders for being the first to artificial general intelligence (AGI), which I have predicted will be achieved by one of the major players before the end of 2026. | | I have never seen a more prolific environment for advanced research than what I've seen in AI over the last five years. Every major obstacle and bottleneck has been overcome. There is nothing but acceleration. And the industry is years ahead of what even experts were predicting just five years ago. | | Any legit AI team has access to whatever capital they need to advance this technology. And as long as they are demonstrating clear progress, they have had no trouble raising additional capital. And that's exactly the environment we'll be in during 2025 and 2026. | | So to answer your specific question about the risk that AI training will reach a dead end, or the usefulness will be capped in some way… | | Zero percent chance. | That's it for today's AMA. If you have a question or comment for me or the team, you can reach us right here. | | Have a great weekend. | | Sincerely, | | Jeff | | | | | | | It's been an incredible week for the private space economy… | | | | | | Jeff's predictions for President Trump's Day One executive orders... | | | | | | It's the catalyst we've all been anxiously waiting for… | | | | | | | | | | | | | | | To ensure our emails continue reaching your inbox, please add our email address to your address book. This editorial email containing advertisements was sent to riku221199@gmail.com because you subscribed to this service. To stop receiving these emails, click here. Brownstone Research welcomes your feedback and questions. But please note: The law prohibits us from giving personalized advice. To contact Customer Service, call toll free Domestic/International: 1-888-512-0726, Mon-Fri, 9am-7pm ET, or email us here. © 2025 Brownstone Research. All rights reserved. Any reproduction, copying, or redistribution of our content, in whole or in part, is prohibited without written permission from Brownstone Research. | | | | |
