The Three Tiers to the AI Boom

In partnership with

Welcome to offtheticker - thanks for following along. WOW what a good last week in the market again. We love a booming market! If you have been following us for a while you will know that we had some great wins the past couple of weeks with $NBIS, $HOOD, and $RGTI. Congrats to our subscribers that have been following along. Every week we like to break down upcoming earnings and how we would be trading around these earnings. Due to the small amount of companies reporting earnings this week we are bringing a special edition of the newsletter, where we are breaking down the timeline that we think will take place over the next 10-15 years in terms of the AI boom and how we would invest for each term.

Tier 1: Software and the integration of AI

Tier 1 represents the infrastructure era of artificial intelligence, which is the stage where the foundation for the entire AI economy is being laid, and we believe is still in the early stages. This tier has already produced clear winners: chipmakers like NVIDIA and AMD, as well as a host of supporting hardware and software firms building the backbone for generative AI and large-scale machine learning. NVIDIA’s GPUs have become the staple standard for training and inference, driving fiscal 2025 revenues into the tens of billions as hyperscalers scrambled to secure compute capacity. AMD, meanwhile, has surged with its MI300 accelerator line and high-performance EPYC CPUs, capturing share from data-center expansion and AI-specific deployments. Together, these companies power the digital engines behind the Mag 7, which include Google, Microsoft, Amazon, Meta, Apple, Nvidia, and Tesla, the firms driving the AI revolution forward.

On the software and cloud side, Microsoft, Google, and Amazon are the key consumers and commercializes this Tier 1 compute. Each is embedding AI deep into their ecosystems, Microsoft through Copilot integrations across Office and Azure, Google through Gemini-powered search and workspace tools, and Amazon via AWS AI APIs and custom Bedrock foundation models. These companies using AI to make products smarter, monetizing intelligence itself, and generating entirely new revenue streams from APIs, model-as-a-service platforms, and enterprise AI deployments.

This Tier 1 wave is the enabler, the industrial stage of AI, where physical infrastructure, data-center power, and silicon design define competitive advantage. It’s the reason global tech capex has exploded beyond $350 billion annually and why investors see these firms not only as software companies, but as the new industrials of the digital age.

How we are trading around tier 1:

As we look ahead, our view is that NVIDIA’s explosive stock growth is likely to cool from its breakneck pace, but what comes next should be a long period of sustained, durable expansion built on the company’s position as the core facilitator of the AI revolution. NVIDIA’s has integrated itself into every major AI platform making it the supplier for a technological gold rush that shows no signs of slowing. While these year over year gains may moderate from the triple-digit surges of 2023–2024, we expect steady compound growth over the next 10–15 years as global demand for AI compute, training capacity, and inference efficiency continues to rise across industries. Dont get our tiers confused, as we still think AI is still in its infrastructure-building phase, and NVIDIA remains the indispensable hardware and software backbone of that movement.

Similarly, Taiwan Semiconductor $TSM and other major chip fabricators are positioned for long, sustained growth as the semiconductor supply chain scales to meet the needs of AI, cloud, and advanced computing (what we will discuss in tier 2). These companies’ consistent margins, capacity expansions have been providing us with reliable growth. Semiconductor manufacturing is becoming the new digital infrastructure and we dont see this slowing down any time soon.

At this stage of the AI boom, the major winners have largely been established. Companies like NVIDIA, AMD, TSMC, Broadcom, and ASML have already secured their positions at the heart of the AI economy and barring any major regulatory or geopolitical intervention such as export restrictions, subsidy shifts, or nationalization efforts similar to what we’ve seen with Intel with the Trump administration, we do not expect this leadership group to change meaningfully.

In short, we are going to be holding Nvidia for the very long future and it is a company that we could see consistently beating the S&P 500 over the coming years, as it is hard to match the growth we are seeing with Nvidia at such a large market cap. This is truly like nothing we have ever seen before and we think a nice blend of these major companies could help to beat the S&P 500, which has been dragged down by retail, over the coming years.

Tier 2: The continued scalability and scarce data centers and power

As the first wave of AI infrastructure solidifies around the Mag 7 and the major chipmakers, Tier 2 represents the enablers of scale, the companies that will power, cool, and construct the physical backbone of the AI era. This is the tier that I think that we are currently blending into and are still in such an early stage of really being able to make money from this opportunity. I want to break this tier up into two categories that are both very important and would be considered their own industries but are currently playing out together hand and hand and are somewhat relying on each other. These two tiers are energy and full stack data center capabilities.

Energy

One of the largest and most overlooked bottlenecks in this boom isn’t software or silicon which have been well established within tier 1, it’s energy. The massive compute demands of training and running large AI models have pushed U.S. data-center energy usage toward record highs, with estimates showing power demand will double by 2030. Data centers from Microsoft, Amazon, and Google are already straining local power grids, forcing utilities to delay projects or build entirely new generation capacity. This is driving a renewed spotlight on nuclear energy, long seen as politically complicated but now being reconsidered as a clean, scalable base load solution. Early signs from Washington, including new Department of Energy fast-track initiatives and bipartisan support for small modular reactors (SMRs) suggest a regulatory environment increasingly favorable to nuclear deployment.

This major shift positions companies like Constellation Energy (CEG), NextEra Energy (NEE), and Vistra (VST) as pivotal Tier 2 winners, operating or investing in nuclear and renewable capacity that could become essential to powering the next generation of AI infrastructure. Below is the link to our Vistra break down that is out now and explains why we think the stock is set up for success.

There are a ton of ways to play this bottleneck but we want to highlight to our readers the importance in finding companies that are going to position themselves to be a reliable outlet for these large data centers and are not just companies with goals to power very day housing and non-ai types of projects. Constellation, now the largest nuclear operator in the U.S., is particularly well-placed to benefit as policymakers signal openness to extending reactor lifespans and investing in modular nuclear systems. Meanwhile, electrical equipment providers like Eaton (ETN) and Schneider Electric (SBGSY) are becoming critical suppliers of grid management and data-center power systems, offering investors exposure to the physical infrastructure needed to sustain the AI revolution.

Full Stack Data Centers

Beyond power generation, the construction and physical creation of data centers has become its own emerging industry. With global data-center capacity set to triple over the next 15–20 years, companies that build, own, or operate these digital factories are entering their golden era. Firms like Nebius AI, Iren, and other specialized computing-infrastructure providers are quietly becoming indispensable. Nebius has been our go to pick and we don’t see any signs of this company stopping. We recently wrote about it in our weekly stock picks and the stock is continuing to soar. We called this stock back in the 50’s and the stock is now sitting at around 120$ just a couple months later.

Iren is another company that we are planning to write a weekly stock pick article on soon. Originally known for cryptocurrency mining, has successfully pivoted its expertise in high-density computing and power management toward AI and cloud workloads, leveraging its access to low-cost, renewable energy and robust cooling technology.

These firms, alongside major data-center REITs and developers such as Equinix (EQIX) and Digital Realty (DLR), will remain essential for decades, providing the physical real estate, network fiber, and cooling systems required to keep the AI ecosystem running. Even mining and materials companies like Freeport-McMoRan (FCX) and Albemarle (ALB) will play an indirect but critical role, as copper, lithium, and rare-earth demand accelerates to meet data-center electrical and battery storage needs. Together, these players form the industrial side of AI, less flashy than software but just as vital. The next 15–20 years of growth in Tier 2 won’t be driven by code or chips, but by steel, wires, and watts, the foundation upon which every AI model depends.

How we are trading around tier 2:

As we previously mentioned in the above sections, energy and data centers, we are currently playing in both sectors and think that these are the two major bottlenecks that could make or break the second leg of this Ai boom. After our article that we posted about Nebius, we saw Microsoft invest in a 10 billion dollar deal with the company, which helps our thesis that these major companies are going to need to rely on the continued build of full stack capable data centers to continue their massive scale up, and the winners will come from the companies that can provide this service to the major tier 1 players. We are expecting that over the coming months we will see other companies begin to invest into Nebius and Iren and force these companies to push the limits of capital expenditure to meet these needs. Overall, the more deals the major tier 1 players sign, the more wealth that will get distributed to the smaller tier 2 players.

Tier 3: Robotics and how far away are we really?

Tier 3 represents AI’s hands and feet; the moment intelligence meets physical reality. This is when we could really begin to see a scary but potentially trillion-dollar industry form that could take the job of millions of people but allow these major companies to profit an infinite about of money. I personally do not think that we are close to seeing this become a reality but over the next 10-15 years I think this all could change in terms of seeing meaningful job displacement and a worry if we have gone too far. We’re entering this very early stage, where advances in machine learning, computer vision, and edge computing are finally merging with maturing hardware to create humanoid robots capable of performing real-world industrial tasks. This stage is being led by companies like Tesla, with its Optimus robot, and Figure AI, which recently partnered with major enterprises such as BMW to begin real-world testing of humanoid robots on assembly lines. While many of these machines are still in limited deployment, the rate of progress has been rapidly increasing.

Figure AI, meanwhile, is emerging as the independent play in the humanoid space, leveraging large language models and reinforcement learning to train its robots on flexible task execution, a critical differentiator from older, pre-programmed industrial robots. Other players like Agility Robotics, Apptronik, and Sanctuary AI are advancing rapidly as well, often with backing from major tech companies seeking to integrate robotics into logistics, manufacturing, and even consumer environments.

At the heart of all of this is AI as the central nervous system, which is once again why I think we will see continued growth in the tier 1 and 2 category thanks to tier 3. Unlike traditional automation, these humanoid systems don’t just follow instructions; they learn, adjust, and improve through the same neural architectures that underpin tools like ChatGPT or Gemini. For investors, Tier 3 represents a frontier where hardware meets intelligence, with multi-trillion-dollar implications across logistics, manufacturing, healthcare, and defense. The winners of this phase will be those who can most effectively fuse AI cognition and also have the scalability to mass produce these robots.

How we are trading around tier 3:

One of the most significant potential moments in Tesla’s history, and perhaps the clearest signal yet of what Tier 3 innovation truly means, is expected to arrive on November 6th, when many believe Elon Musk will unveil the Optimus 3 humanoid robot. For months, Musk has hinted that Tesla’s long-term value may not be driven by cars or even energy storage, but by its humanoid robotics division, which he describes as the company’s future cornerstone. This marks a dramatic pivot, and possibly the largest shift in Tesla’s narrative since the Optimus 2 launch, toward the physical manifestation of artificial intelligence.

We know as investors that Tesla’s competitive advantage lies in scale. The company already possesses the manufacturing capacity, supply chain expertise, and engineering needed to mass produce complex hardware, something few robotics startups can even begin to match. The challenge has always been the software and intelligence behind the machines. Both the original Optimus prototype and its successor, Optimus 2, have been seen as early-stage proofs of concept that are far from capable of meaningful work in industrial environments. If Musk can demonstrate that Optimus 3 represents a true leap forward, capable autonomy, it could spark major debates about the true value of Tesla. The market will focus on the idea of a subscription-based model, where industrial companies “hire” robots on a monthly basis to perform labor-intensive or repetitive tasks, a concept that could transform the economics of labor across manufacturing, logistics, and service industries. We are considering adding double or triple leverage shares in the stock right before the event but want to make it clear to investors that this is not financial advice and we are doing our own research and with how Elon Musk operates our plans could change over the next month.

The free newsletter making HR less lonely

The best HR advice comes from those in the trenches. That’s what this is: real-world HR insights delivered in a newsletter from Hebba Youssef, a Chief People Officer who’s been there. Practical, real strategies with a dash of humor. Because HR shouldn’t be thankless—and you shouldn’t be alone in it.

Sign Up Free

Buying Cannabis Online Is Now Legal, And Incredibly Convenient

For years, buying cannabis meant taking a trip to a dispensary, dealing with long lines, limited selection, and inconsistent pricing. But thanks to changing laws and innovative online retailers, buying high-quality THC products is now 100% federally legal—and more convenient than ever.

And when it comes to quality and reliability, Mood is leading the way…

Because they’ve completely flipped the script on cannabis shopping. Instead of memorizing hundreds of confusing strain names – like “Gorilla Glue” and "Purple Monkey Breath" – you simply choose how you want to feel: Creative, Social, Focused, Relaxed, Happy, Aroused, and more.

Each gummy is formulated with the perfect blend of Delta-9 THC and botanicals to deliver the perfect mood.

Want a great night’s sleep? Try the Sleepytime gummies. Need laser focus Mind Magic gummies have you covered. Hotter sex? Try the Sexual Euphoria gummies.

It's cannabis shopping that actually makes sense for “normal” people.

20% OFF your first order!

The information provided in this newsletter is for informational and educational purposes only and should not be considered financial, investment, or trading advice. While every effort has been made to ensure accuracy, we do not guarantee the completeness or reliability of the information. Past performance is not indicative of future results. You should conduct your own research or consult with a licensed financial advisor before making any investment decisions. We are not responsible for any losses that may occur from the use of this content