Source: Smart Investor
Cathie Wood, known as 'Wood Lady,' recently released 'Big Ideas 2026' with her team and enthusiastically delivered a presentation to highlight the key points of the report.
"It does feel like a hype cycle at the moment." However, Ark's Cathie Wood shifted her tone, arguing that the massive capital expenditures on AI today are not 'dark fiber' but are being genuinely consumed. "AI is still in a very early stage and has a long way to go."
Once again, she used the term 'Great Acceleration,' which stirs excitement.
'Wood Lady' acknowledged that capital expenditure in the technology and telecommunications industries as a proportion of GDP has approached the peak levels of the previous bubble. However, unlike the unused fiber optics back then, GPUs today remain in short supply.
Want to stay updated on Cathie Wood's latest investment moves? FollowFutubull Celebrity Tracker, your guide to tracking the activities of major investors and decoding investment signals.
From a broader perspective, she compared this cycle of investment to an 'infrastructure-level' phase—similar to railways, automobiles, and electricity—that could push the share of capital expenditures to 12% of GDP.
Signs of this 'acceleration' in structural changes are already emerging in multiple directions: first, the rapid expansion of AI infrastructure, with data center spending increasing 2.5 times since the advent of ChatGPT and expected to reach $1.4 trillion annually by 2030; second, the ongoing restructuring of fintech, with the total scale of stablecoins surpassing $300 billion, posing 'dislocation and turbulence' for traditional finance; third, productivity itself is being reassessed, such as Tether achieving an output per employee of over $50 million last year. Behind this extreme efficiency lies the iteration of new asset structures and system efficiencies.
Therefore, while the outside world worries about whether AI will 'replace humans,' ARK's response is more like a counter-question: in today's world, asking ChatGPT a question might just start a new business—isn't this the best era for entrepreneurship?
From multi-omics and gene editing to nuclear power and reusable rockets, from Robotaxis to automated logistics... What 'Wood Lady' describes is no longer a single-threaded narrative but the outline of a systemic reconstruction. When these technological platforms begin to interconnect, growth may not extend gradually but leap forward in steps.
As 'Wood Lady' herself said, ARK’s research leans more toward an investment banking and primary market perspective, making early-stage industrial ideas and concepts seem somewhat distant when viewed from the secondary market angle.
But how can we put it? When we see a 70-year-old investor still tracking, studying, and trying to understand the most cutting-edge technological trends with almost 'entrepreneurial enthusiasm,' perhaps this alone is the most compelling interpretation of 'long-termism.'
Many of her predictions may not be fully embraced by the market, but her sensitivity to the idea that 'change is happening' and her courage to invest capital into technological trends are extremely rare.
We admire this curiosity and the drive to understand the world. Here is a speech by the wise investor 'Cathie Wood' shared with everyone.
01. AI remains in a very early stage
I would like to begin with what we call the 'Great Acceleration.'
The key point we want to convey is: AI is still in its infancy and has a long way to go.
Of course, it does feel like a hype cycle at present because capital expenditure in the technology and telecommunications sectors as a percentage of GDP has almost returned to the highs seen during the tech and telecom bubble.
But let us not forget, that bubble laid down vast amounts of fiber optics, much of which became 'dark fiber,' sitting idle for years. The driving force behind this cycle is GPUs. To quote Brad Gerstner of Altimeter: these GPUs are not 'gathering dust'; they are in short supply and being consumed actively.
We fully agree.
When we examine this wave of capital expenditure through a historical lens, it resembles the handover between two 'infrastructure-level' mega-waves: railway construction in the 19th century pushed capital expenditure to 5%-6% of GDP; the rise of the automobile industry in the early 20th century brought it to 3%-4% of GDP. Later came the tech and telecom bubble, where combined investments in software, hardware, and communications equipment reached roughly the same scale as the automobile wave.
Today, with simultaneous advancements in AI, robotics, energy storage, multi-omics sequencing, and blockchain, our current capital expenditure cycle has already reached the equivalent magnitude.
However, we believe this time will not stop here.
Our assessment is that the share of capital expenditure in GDP may eventually rise to 12%. I will elaborate on the reasons later, but the core concept can be summarized in one word: productivity. It is the most fundamental yet robust variable in economics. Due to the logic of productivity, we believe we are still in the early stages of this cycle.
Before continuing with our observations on the improvement in productivity, I would like to present a broader conclusion: driven by this wave of investment boom and rising productivity, we believe that by 2030, the U.S. real GDP growth rate has the potential to accelerate, approaching levels above 7%.
This is not baseless optimism. Technological revolutions often lead to step-like increases in GDP growth. For example, from approximately 400 years ago until around 1900, the long-term average global GDP growth rate was only about 0.6%. However, by the late 19th and early 20th centuries, the maturation of technological platforms such as railways, telephones, electricity, and internal combustion engines drove a profound productivity revolution, causing global real GDP growth to leap from 0.6% to 3%, a level that persisted for about 125 years—a fivefold increase.
Our current assessment is that there may be an increase of approximately 2.5 times, though I even consider this estimate to be relatively conservative. This time, it is not a single technological platform advancing alone; rather, five innovation platforms—robotics, energy storage, artificial intelligence, multi-omics sequencing, and blockchain—are simultaneously evolving and accelerating their convergence.
The superposition and coupling of these platforms are what we refer to as the source of the 'Great Acceleration.'
02, Position Yourself on the Side of Change
Next, I will take you through some of the most astonishing key data points and pivotal moments from our research this year.
Let us start with AI infrastructure. The scale of this cycle for data center infrastructure is currently around $500 billion, which is approximately 2.5 times larger than before the advent of ChatGPT. However, this is just the beginning.
We project that by 2030, this figure will climb to $1.4 trillion annually.
To be honest, we are fortunate to have had the turning point of ChatGPT. It was a quintessential 'moment of epiphany' for all of us: you suddenly realize that it can do so much, evoking an almost incredulous feeling.
Upon further analysis, you will see in 'Big Ideas' that the subscription cost at ARK—$20 per seat per month (strictly speaking, paid by the company), and $200 per month for a small number of heavy users—is essentially offset by productivity gains within about half a day.
In other words, the payback period is almost negligible. This is because it directly elevates research efficiency to a new magnitude, akin to leveraging time.
We have another judgment that may be controversial: software stocks have actually faced a tough period during this wave of AI prosperity. Ironically, the reason lies precisely in AI's 'disruptive' nature, which will particularly impact SaaS and more broadly reshape all traditional software.
We believe that the growth rate of the software market could reach approximately 19% in a bearish scenario and possibly even 54% annually in a bullish scenario, compared to about 14% annually over the past five to ten years. However, this is contingent upon positioning yourself on the right side of change.
Many software companies that are 'on the right side of change' have yet to go public, such as OpenAI and Andro X AI. In the public market, the most typical winners tend to emerge at the 'Platform as a Service (PaaS)' level of the technology stack, with the most representative example being$Palantir (PLTR.US)$。
Ultimately, it comes down to the same point: align yourself with change.
This does not mean that all SaaS providers will be crushed. Instead, a more likely outcome mirrors what we often see in traditional industries: accelerated market consolidation, eventually leading to the dominance of a few giants or even 'super leaders.' For instance, in the retail sector,$Walmart (WMT.US)$is the reference we use for comparison.
When it comes to the impact of financial technology, blockchain, and DeFi (Decentralized Finance) on the traditional financial system, JPMorgan is highly likely to remain unscathed. It will integrate new technologies into its own system while potentially driving further industry concentration.
What we've discussed so far pertains to the enterprise side. Shifting the focus back to the consumer side, 'Big Ideas' predicts that 65% of future advertising spending may occur within 'AI-generated' spaces, while about 25% of shopping and e-commerce will be driven by AI.
03. In the future, Bitcoin will become a very important store-of-value tool.
Finally, we come to $Bitcoin (BTC.CC)$ this line. The reason why the trend in 2025 is highly controversial is because after the 'flash crash' on October 10, the market experienced a significant pullback.
Our assessment is that this deleveraging was mainly related to a software malfunction at Binance, which triggered an automated deleveraging mechanism, causing a chain reaction resulting in approximately $28 billion worth of forced liquidations. However, we believe that this round of impact has largely been absorbed.
Thus, the most pressing question for Bitcoin holders now is: Have we finally broken away from the traditional rhythm of 'once every four years, cyclical deep corrections'?
We are starting to lean toward the view that yes, structural changes may be occurring. Even if there is a period of consolidation between $80,000 and $90,000 in the short term, we still believe that the direction of the next key turning point is more likely to be upward.
We also discussed how stablecoins would 'replace' some of the roles we originally thought Bitcoin would fulfill. In fact, we had such expectations ten years ago, but stablecoins did not exist at that time. Now, in use cases related to remittances and emerging markets, we do see stablecoins taking over that functionality, replacing what we once believed Bitcoin would provide.
However, this does not alter our judgment regarding Bitcoin's long-term outcome.
Bitcoin still has a long way to go as 'digital gold.' Interestingly, throughout Bitcoin’s entire lifecycle, its correlation with gold has been almost non-existent, standing at only 0.14.
Nevertheless, we still believe that because it represents an entirely new technology, Bitcoin is not only a risk-on asset but is also becoming a risk-off asset. It can hedge against inflation.
Its supply growth rate is even lower than that of gold. I would bet that gold mining companies are now desperately trying to extract more, given the higher price incentives currently offered by the gold market.
But Bitcoin cannot do this. Bitcoin miners are mathematically constrained, with a total cap of 21 million coins.
Therefore, we believe that Bitcoin will become an extremely important store of value in the future, especially over the next 5, 10, or 15 years, during which we anticipate a generational wealth transfer that will further highlight its importance.
To my knowledge, the current scale of global stablecoins has exceeded $300 billion. Meanwhile, we have also observed a significant acceleration in asset tokenization on public blockchains, reaching a scale of $19 billion. This increase is remarkable, having approximately tripled or even more over the past year.
From a longer-term perspective, we believe the tokenized market will eventually reach $11 trillion, primarily driven by three types of assets: publicly traded company equities, sovereign debt, and bank deposits.
This will be an enormous new market. We have also begun publishing quarterly DeFi reports to complement our Bitcoin quarterly reports. Many changes are accumulating and intensifying as the financial system undergoes a 'complete migration' to new technologies, leading to significant dislocations and volatility within the industry.
04, Possibly the Most Important Entrepreneurial Window
Another point we wish to emphasize is that the productivity of the new generation of companies has reached an astonishing level.
Take Tether (a cryptocurrency company whose main product is the stablecoin USDT) as an example. We estimate that it achieved a profit or output of 'US$50 million per employee' last year, which is almost unheard of. We believe that once enterprises fully adopt these new technologies, the resulting productivity gains will far exceed what most people currently imagine.
Thus, while many worry about 'large-scale job displacement' due to advances in robotics and AI, we disagree. On the contrary, we believe this could become one of the most important entrepreneurial windows in history.
Because today you can directly ask ChatGPT or Grok: Is there a huge unmet demand in the market? If so, how should products or services be designed? How to set up the business?
Just imagine, if we had such a starting point when we founded ARK back then, how much trial and error could have been avoided.
You can even treat it as a co-creation partner and work with ChatGPT, xAI, and Grok to build out your business plan step by step. Even if you are job hunting, you can reverse this approach: first identify a real unmet need, then consider whether to build a business around it.
Starting a business is thrilling and exciting, but the risks are equally significant.
Remember, 90% of startups fail, so stay clear-headed, confirm that you are solving a genuine pain point, and try to incorporate AI, blockchain, and other available new technologies into your toolkit.
05, The Integration of Biotechnology and AI
Speaking of 'multis (multi-omics and multiplex sequencing, internally abbreviated as multis at ARK),' frankly, I used to zone out whenever I heard about it, but this time I feel it’s completely different. 'Multis' refers to the direction of life sciences: the integration of multi-omics and sequencing technologies, combined with artificial intelligence and new tools like CRISPR for gene editing, which are pushing many advancements forward.
For example, in cancer diagnosis, we are moving the detection window to stage one, or even before stage one. Even polyps leave signals in the blood, and these can all be detected through blood tests.
At the same time, the cost of drug discovery and development is being systematically reduced. We believe that the comprehensive cost of bringing a drug from research to market, which currently stands at approximately $2.4 billion, may drop to around $700 million within the next four years. This figure naturally includes the costs of failed projects along the way.
Our colleague Shea has done thorough work quantifying exactly how much a 'cure' is worth.
Our estimate is that for a condition as rare as HAE, the systemic value of a one-time 'cure' could be as high as $11 million, with just a single treatment required. This solution is currently in Intellia's trials and involves in vivo treatment, eliminating the need for extremely painful pre-treatment procedures.
The reason it is worth $11 million is because it can save an equivalent amount of long-term costs for the entire healthcare system.
Of course, companies may ultimately only set the price at around $3 million. But I believe insurers will have no objections, as the calculation is straightforward.
06, 'Two New Worlds' and Distributed Energy
Next is reusable rockets, which are equally astonishing. SpaceX is at least a decade ahead of any competitor. It achieved the first landing of a recoverable rocket in 2015. Today, it controls roughly two-thirds of the global satellite launch market and about 85% of the orbital payload mass.
More importantly, we are witnessing a new application scenario that would be nearly impossible without reusable rockets: 'space data centers.'
This is a new direction we have only recently incorporated into our research framework, and it was indeed inspired by Elon Musk. A completely new world is being developed.
When many people talk about automation, AI, and unemployment, I often say: don't jump to conclusions. We are actually witnessing the simultaneous formation of two entirely new worlds.
One is in outer space, where numerous new industries and jobs will emerge; the other is in the digital world. With blockchain technology, we are seeing immutable property rights in the digital realm for the first time.
One of the most effective ways to lift individuals and nations out of poverty is to establish reliable, clear, and non-arbitrarily erodible private property rights.
We are beginning to see its雏形, and it is just the beginning, so we are very excited about it.
Next is distributed energy. Over the past year, we have conducted extensive research on nuclear power.
You might be surprised to learn that regulation in the mid-1970s essentially 'paused' nuclear power at the most critical phase of its cost curve. If regulation had not interrupted the downward trend in construction costs for nuclear plants, electricity costs would likely be about 40% lower today.
This is why we hold high expectations for nuclear power. As nuclear power scales up again, it has the potential to reduce electricity prices once more and become a significant 'hedging force,' offsetting the increasing load on the grid from AI and data centers with more stable and lower-cost supply.
07, Autonomous Driving and Automated Logistics
Of course, we also cover other autonomous vehicles.
We have robots, rockets, and other autonomous vehicles, among which the most typical example is undoubtedly Robotaxi. We believe we are in the process of witnessing the birth of the robot taxi market.
This year, we also updated our analysis. According to our estimates,$Tesla (TSLA.US)$has expanded its competitive advantage over Waymo. Previously, we believed Tesla's cost structure would be approximately 35% lower than Waymo’s. However, as we refined our analysis, we now estimate it to be 50% lower.
In addition, look$Uber Technologies (UBER.US)$It is also interesting to observe other companies attempting to 'enter' this space. They understand its significance, but we have consistently stated from the beginning and still believe now: Uber Technologies and other platforms are more likely to serve as 'customer acquisition channels,' directing traffic to Tesla and Waymo.
As they expand across the United States, Waymo could become the second-largest player.
One of the most intriguing studies we conducted this year involved estimating how much urban mileage Uber Technologies currently covers in the United States. The answer is approximately 1%.
We then worked backward: How many vehicles would Tesla or Waymo need to deploy to cover this 1% of urban mileage with autonomous vehicles? The result was 140,000 vehicles.
What if the goal is to cover all urban mileage? The answer is only 24 million vehicles, which is less than 10% of the current vehicle stock in the United States.
This estimation is based on the U.S. market. We also know that annual car sales in the U.S. are approximately 15 million units. This comparison provides an intuitive sense of how significantly the automotive market size may contract in terms of 'how many vehicles need to be produced.' Nevertheless, even with this contraction, we can still achieve coverage of 100% of today's urban mileage in the U.S. using autonomous vehicles.
The reason is straightforward: the utilization rate of autonomous vehicles is much higher than that of private cars. Our personal vehicles are typically in use only about 4% to 5% of the day, whereas autonomous vehicles might achieve a utilization rate of 50% to 60%.
We continue to believe that the autonomous driving market will generate a market capitalization of approximately $34 trillion by 2030.
By 2030, the global ecosystem is estimated to reach around $30 trillion, encompassing not just vehicle providers and traffic-directing platforms, but more critically, platform providers. The platform providers will capture the majority of the economic benefits.
Finally, there is automated logistics. Autonomous trucks can reduce distribution costs by 60%. Drones and ground-rolling robots will also significantly lower delivery expenses.
Did you know? There are already 4 million drone deliveries annually. Zipline, a company in our venture capital fund, accounts for roughly half of that volume, we believe.
We believe that in the future, the cost of drone delivery will reduce the current$DoorDash (DASH.US)$cost of approximately $15 per order to less than $1. This implies a long-term reduction of around 90%. Therefore, this represents another vast emerging market.
08, We Have Now Entered the "Golden Era"
The reason we are so excited is that we believe we have now entered the "golden era," which was precisely why I founded ARK.
I founded ARK Invest in 2014 because I believed that ARK's research could address an unmet demand in the market—research similar to what investment banks conducted whenMergers and acquisitionsmarkets were more active.
However, over the past four years, the M&A market has been far less active, nearly halted under the regulatory environment of the previous FTC (Federal Trade Commission).
Therefore, our research is indeed meeting and filling a gap. What makes us even happier is seeing many startups quoting our research in their pitch materials.
What are startups doing? They are heads-down, sprinting as fast as they can to innovate.
And what are we doing? We are helping them estimate how large the market they are chasing really is.
I believe that the 'wall of worry' surrounding innovation over the past few years is actually a good thing. I think it’s positive because innovation certainly comes with risks.
But we believe that now this flywheel has been built, and the era of innovation will be unstoppable.
Futubull[Opportunity Page]The tracking feature for celebrity portfolio opportunities has been officially launched! Explore a variety of portfolios held by renowned investors, follow their strategies with a single click, precisely target high-quality assets, and invest with greater confidence!
Editor/Rice
