Betting on the Tech Revolution: What "Cathie Wood" Said in Her 163-Page Annual Report

Author: iBloomberg

Wall Street star fund manager and CEO of Ark Investment Management, Cathie Wood, has released the report titled "Big Ideas 2024" as scheduled, led by her ARK research team.

Since the beginning of 2024, Wood's flagship fund, the ARK Innovation ETF (ARKK), has risen by 68%, ranking in the top 1% among its peers.

In this 163-page report, "Wood" continues to focus on the field of "disruptive innovation technologies," predicting that technological convergence, AI, digital wallets, precision therapies, and 3D printing will change every aspect of the world, with technology expected to accelerate the global economy to 7% by 2030.

Disruptive Technology Innovation Platforms

The report points out that the convergence of disruptive technologies will define the development of the next decade. Five major technology platforms—AI, public blockchain, multiomic sequencing, energy storage, and robotics—are merging to change global economic activities, with economic growth potentially accelerating from an average of 3% over the past 125 years to 7% over the next seven years.

Five Innovative Technology Platforms

Technological convergence may bring about macroeconomic structural changes more impactful than the first and second industrial revolutions. Globally, the emergence of robots is revitalizing manufacturing, robot taxis are changing transportation, and AI is enhancing the productivity of knowledge workers, leading to accelerated real economic growth.

Driven by breakthroughs in AI technology, by 2030, the global stock market value related to disruptive innovation could increase from 16% of total market capitalization to over 60%. The annualized stock return related to disruptive innovation could exceed 40% by 2030. Over the next seven years, its market value is expected to grow from approximately $19 trillion to about $220 trillion by 2030.

The report states that the impact of the convergence of disruptive technologies like AI and robotics on the economy will surpass that of previous general technologies such as the steam engine, railroads, electricity, and the telephone. The degree of convergence and influence among these disruptive technologies varies; some have a very high degree of integration (like AI), while others are lower (like precision therapies). The degree of AI integration can demonstrate the position and role of core technology catalysts.

The pace of AI development is also faster than market expectations. In 2019, the market generally expected the emergence of general AI to take 80 years; this was reduced to 50 years in 2020 and to 34 years in 2021. After the emergence of GPT-4, the expected timeline was even shortened to 8 years. Ark anticipates that general AI could appear as early as 2026 and no later than 2030.

Ark believes that the advancement of individual disruptive technologies can create significant new market opportunities when they converge, such as in the general robotics market and the autonomous taxi market.

Neural networks combined with battery technology could enable the scaling of autonomous mobile devices (like autonomous taxis). In addition to requiring batteries and AI support, general robots also need components like motors and sensors. As autonomous taxis scale, the costs of these technologies will decrease, leading to rapid development in the general robotics market.

Compared to industrial robots, internet information technology, and steam engines, disruptive technologies, especially artificial intelligence, will have a tremendous impact on the economy. The report predicts that disruptive technological innovation will dominate the global stock market capitalization. By 2030, disruptive technologies will not only expand total market capitalization by about three times but will also become a dominant force in market value.

Artificial Intelligence

As one of the core driving forces of today's technological revolution, artificial intelligence is changing human lifestyles, work habits, and social structures at an unprecedented speed. Its rapid development has not only spawned new industries and business models but has also triggered profound changes in traditional industries. ARK points out in the report that thanks to the rapid decline in the cost of training AI and the open-source efforts of tech giants, AI will bring far more than just efficiency improvements; it will also drive rapid global economic growth.

The core technologies of AI include machine learning, deep learning, natural language processing, and computer vision. Among them, deep learning has made groundbreaking progress in recent years, significantly enhancing AI's performance in areas such as image recognition, speech recognition, and natural language understanding. These technological advancements lay a solid foundation for the widespread application of AI.

ARK notes that the emergence of ChatGPT has amazed businesses and satisfied users, significantly increasing productivity: programming assistants like GitHub, Copilot, and ReplitAI have achieved certain results, improving the efficiency and work state of software developers.

The rapid development of large models for text-to-image generation is also reshaping graphic design, with the output quality of image models now comparable to that of professional graphic designers. The cost of creating written content is also plummeting; over the past century, the cost of writing has remained relatively stable in real terms. However, in the past two years, as the quality of large language model writing has improved, costs have also decreased. Employees who previously performed poorly have benefited more from the emergence of AI compared to high-performing employees.

ARK points out that as the application areas of AI expand, researchers are innovating in AI training and reasoning, hardware, and model design to improve performance and reduce costs. The cost of reasoning seems to be declining at an annual rate of about 86%. By 2030, the integration of hardware and software could reduce AI training costs by 75% annually.

According to use cases based on company size, the cost of reasoning appears to be declining at an annual rate of about 86%, even faster than training costs. Currently, the reasoning costs associated with GPT-4 Turbo are already lower than those of GPT-3 a year ago.

As Wright's Law states, improvements in accelerated computing hardware should reduce the production cost of AI-related computing units (RCUs) by 53% each year, while improvements in algorithm models further reduce training costs by 47% annually. In other words, the integration of hardware and software could lead to a 75% annual decrease in AI training costs by 2030. (Note: The core content of Wright's Law is that for every doubling of cumulative production of a certain product, costs will decrease by a constant percentage. For example, in the automotive sector, every doubling of production leads to a 15% drop in cost.)

The report indicates that open-source models are competing with closed-source models, and overall, open-source models are improving their performance faster than closed-source models: the open-source model field, led by leading companies like Meta, is gradually challenging the closed-source models of OpenAI and Google.

In 2023, open-source models made rapid progress in performance benchmarking, gaining continuous support from developers in large enterprises, startups, and academic institutions. Ark's report expresses great anticipation for the achievements of the open-source community in 2024.

Regarding current investor concerns about whether the training data for large language models will be exhausted, thus limiting their performance, Ark points out in the report that model optimization requires more training data. EpochAI estimates that high-quality language/data sources like books and scientific papers may be exhausted by 2024, but there is still a vast amount of untapped visual data.

Microsoft CEO Satya Nadella mentioned in Microsoft's earnings report for the first time that Microsoft is preparing for AI monetization. Ark mentioned in this year's report that customized AI products should enjoy more pricing power.

With the emergence of open-source alternatives and declining costs, vendors developing and customizing AI software for businesses should find it easier to monetize. In contrast, AI applications with simple functions will quickly commoditize, leading to reduced profitability in intense competition.

Ark believes in the report that from the perspective of continuously improving the productivity of knowledge workers, the potential opportunity for AI software vendors reaches trillions of dollars, and the global software market could grow tenfold: by 2030, AI has the potential to automate most tasks in knowledge-driven professions, significantly enhancing employee productivity. Software solution providers for automating and accelerating knowledge work tasks should be the main beneficiaries. If the new wave of AI application innovators retains similar pricing power as today, and the productivity improvements from AI are as significant as imagined, then by 2030, the global software market could grow tenfold.

Autonomous Driving

Breakthroughs in artificial intelligence will drive autonomous taxis to fundamentally change urban mobility and significantly alter or reduce the demand for personal car ownership, impacting the auto loan market that relies on personal car sales. According to ARK's research, robot taxi platforms will redefine personal mobility and create $28 trillion in enterprise value over the next five to ten years:

It is estimated that the cost of scaled autonomous taxis could be as low as $0.25 per mile, and such low costs may drive widespread adoption of autonomous taxis.

The report points out that autonomous vehicles are safer than human-driven cars, and the application of large language models and generative AI can accelerate the development of autonomous driving technology:

The accident rate of autonomous vehicles will be 80% lower than that of human drivers, potentially reducing approximately 40,000 car-related deaths in the U.S. each year and about 1.35 million globally.

In full self-driving (FSD) mode, Tesla's safety on the ground is five times that of manual mode and 16 times the national average. Waymo's autonomous vehicles are about 2-3 times safer than the national average.

Neural networks trained by GPT-4 to perform robotic tasks outperform human expert programmers in 83% of tasks, with an improvement rate of 52%. Large language models support text-based training, validation, and self-explanation, which should help facilitate regulatory approval.

Multimodal models can train autonomous vehicles through images and text, potentially enhancing system performance. Generative AI can simulate training and validate the safety of autonomous vehicles.

Ark's report emphasizes that the increase in market share of autonomous taxis will disrupt the U.S. auto loan industry. By 2030, the enterprise value of autonomous platform providers could reach $28 trillion, equivalent to nine times the market value of all automakers in 2023. Over the past three years, rising interest rates have increased monthly auto loan payments for new cars by about 27%, from $581 to $739. This has also led to a recent record high in the number of auto loans overdue by more than 60 days. As electric vehicle prices continue to decline, more users are beginning to adopt autonomous taxi technology, reducing the value of gasoline vehicles.

Robotics Technology

Ark believes that the integration of AI and hardware could drive the application of robotics in a broader range of fields, with general robotics expected to welcome new market opportunities, generating annual revenues exceeding $24 trillion.

Ark points out that the rapid advancement of robot performance and the significant decrease in costs are stimulating factories to increase their adoption of robots: improvements in robot performance are further driving demand for industrial robots. Advances in computer vision and deep learning have increased robot performance by 33 times over seven years, with robots now outperforming humans by more than two times, and the upper limit remains unclear.

With the aid of AI and computer vision, robots should be able to operate economically and efficiently in unstructured environments. Lower prices have stimulated demand for industrial robots; when robot production doubles, the cost of industrial robots decreases by 50%.

Ark emphasizes that robots working collaboratively with humans are reaching a critical development stage known as the "S-curve inflection point," about to enter a rapid growth phase:

The S-curve is a commonly used graphical representation that describes the market adoption rate of new technologies or products over time. It starts with slow growth, then rapidly increases, and finally slows down again, forming an S shape. When the market share of a new technology approaches 10% to 20%, it typically indicates that it is about to enter a phase of rapid growth.

Taking Amazon's deployment of robots as an example, it is evident that Amazon significantly increased its use of robots in 2023, reaching a historical high, comparable to the number of human employees.

The use of robots has also had a tremendous impact on productivity; in Amazon warehouses, the time from customer order to product shipment has been reduced by 78%, measured in minutes.

Therefore, Ark believes that in the future, general robotics will include not only household robots but also manufacturing robots, with global manufacturing GDP expected to benefit from the application of robotics, soaring to $28.5 trillion by 2030.

Digital Consumers

According to ARK's research, digital leisure spending is expected to gain a larger market share from the physical economy, growing at a compound annual growth rate of 19% over the next seven years, from $7 trillion in 2023 to $23 trillion by 2030. The report identifies five trends driving this growth:

1. Advertising spending in smart TVs (CTV) is projected to grow at a compound annual growth rate of 17%, increasing from $25 billion in 2023 to $73 billion by 2030.

2. E-commerce revenue on social platforms is expected to grow at a compound annual growth rate of 32%, rising from $730 billion in 2023 to over $5 trillion by 2030.

3. Consumer demand for sports betting remains strong and will continue to grow rapidly.

4. AI-assisted game creation is set to become a new wave in the gaming industry, with user-generated content (UGC) platforms like Roblox potentially leading to explosive growth in game content. Roblox has provided over 470 million experiences globally, which is 52 times the total number of PC, console, and mobile games.

5. The era of AI + hardware is opening up, potentially redefining wearable devices. If virtual reality (VR) devices continue to face adjustments, new AI hardware devices are bound to emerge.

Ark's report indicates that the emergence of AI will further reduce average working hours and stimulate digital entertainment consumption. Generative AI could reduce average labor time by 1.3%, from 5.0 hours per day in 2022 to 4.5 hours by 2030. As a result, consumers may have more time for online entertainment, with online time in daily life increasing from 40% in 2023 to 49% by 2030.

Digital Wallets

Ark's report points out that leading vertical software platforms are creating a closed-loop consumption system through bilateral markets, facilitating closed-loop transactions from consumers to merchants, merchants to employees, and employees back to merchants. Digital wallets on these platforms will achieve a fully closed payment ecosystem, with C2B digital wallet payment totals expected to grow at a rate of 20% annually, reaching approximately $7 trillion by 2030:

In addition to supporting core business operations, vertical software providers like Block, Shopify, and Toast are also integrating financial services for merchants. Centered around digital wallets, they collaborate with banks and fintech companies (or have their own banking licenses) to eliminate inefficient interactions between merchants and traditional financial institutions.

Over the next seven years, C2B digital wallet payment totals are expected to grow at a rate of 20% annually, increasing from approximately $2 trillion in 2023 to about $7 trillion by 2030. The share of closed-loop payments will rise from 4% to 25%, with payment revenues for BlockSquare, Shopify, and Toast projected to increase from $3.5 billion to $21 billion, representing an annual growth rate of 29%.

Ark believes that bilateral markets can close the financial loop between consumers and merchants, and the closed-loop payment ecosystem achieves internal transfers in three ways:

From consumers to merchants, from merchants to employees, and from employees (who are also consumers) back to merchants. To build these payment ecosystems, platforms must possess: 1) a large and highly engaged bilateral network, 2) end-to-end visibility into merchant operations and finances, and 3) vertical industry expertise.

Digital wallets have the potential to replace consumers' positions in the enterprise (C2B) payment ecosystem, allowing transactions to bypass banks and card networks, saving exchange fees for payment institutions, merchants, and consumers. Vertical software platforms with scaled consumer and merchant ecosystems will leverage digital wallets to facilitate closed-loop transactions.

Vertical software platforms can provide financial services to merchants. Through digital wallets, these platforms not only enhance convenience but also monetize deposits, reducing the steps from payment authorization to merchant settlement from 16 to 5, more than doubling platform profitability.

Precision Medicine

Ark points out that over the past two decades, there has been a surge in new models of precision medicine, CRISPR gene editing, RNA therapies, and targeted protein degradation. Driven by AI, CRISPR gene editing, and new sequencing technologies, innovative therapies have increased the returns on R&D, enabling some diseases previously considered untreatable with targeted drugs to now be treated with newly developed medications, offering new possibilities for certain conditions:

Companies in the field of precision medicine are expected to experience significant growth. Precision medicine is a medical approach that customizes treatment plans based on patients' specific genetic information, involving in-depth research and application across multiple biomolecular levels, including DNA, RNA, and proteins.

According to ARK Investment Research's forecasts, from 2023 to 2030, the enterprise value of companies focused on precision medicine will grow at an annual rate of 28%, increasing from approximately $820 billion to about $4.5 trillion:

In the past thirty years, new therapeutic approaches with entirely new mechanisms of action have emerged. They not only expand the number of treatable diseases but also improve efficacy and safety. In 2023, over 25% of clinical trials are utilizing new therapeutic models.

According to Ark's research, the combination of new therapeutic models and R&D methods, along with regulatory approvals for "precision" therapies, is expected to reverse the trend of declining investment returns in the pharmaceutical industry.

An increasing number of precision therapies are becoming multiomic and curative, with mechanisms of action spanning DNA, RNA, and proteins. According to ARK's research, the enterprise value of companies focused on precision medicine is expected to grow at an annual rate of 28% over the next seven years, rising from $820 billion in 2023 to $4.5 trillion by 2030.

Precision medicine, including RNA-based drugs and "targeted protein degraders" (TPDs), not only expands the number of proteins in the human genome that can be treated with drugs but also increases the number of tissue types that can be treated.

Ark notes that in the past decade, the continuous development and refinement of biological tools and technologies, particularly the advancements in high-throughput proteomics, artificial intelligence (AI), and single-cell sequencing, have become key forces driving biological research and medical technology development. It is expected that drug development expenditures could decrease by over 25%, with the enterprise value in the precision medicine field increasing at a compound annual growth rate of 26%, from approximately $820 billion in 2023 to about $4.5 trillion by 2030:

The combined use of these technologies enhances the productivity and efficiency of research and development work, as well as the precision of medical applications, such as disease diagnosis, personalized treatment, and new drug development.

According to ARK's research, artificial intelligence and automation are providing stronger support for drug development, and technological advancements should significantly reduce the costs associated with developing each drug:

In the past decade, advancements in mass spectrometry and bioinformatics have greatly improved proteomic analysis, enhancing resolution, accuracy, and the ability to analyze multiple samples simultaneously.

Wright's Law predicts the decline in proteomics costs, enabling detailed exploration of proteomes in health and disease, and accelerating the discovery of cancer biomarkers and the development of targeted therapies. Single-cell RNA sequencing is revolutionizing our understanding of cancer.

Ark's report believes that the development of artificial intelligence and automation will lower drug costs and streamline approval processes. Meanwhile, advancements in foundational biology, artificial intelligence, automation, and experimental design should significantly reduce preclinical drug development costs by eliminating unpromising drug candidates early in the drug development process, preventing misallocation of downstream R&D funding, and creating greater space in the discovery phase.

In the next decade, companies that fully leverage these technologies could reduce the cost of each approved drug by nearly 50%, partly due to the success rate of candidate drugs entering clinical trials increasing more than twofold.

Electric Vehicles

Ark's report indicates that after battery costs rose due to supply chain disruptions, battery costs are now declining in accordance with Wright's Law, which will drive down electric vehicle (EV) prices. It is expected that by 2030, electric vehicles will account for 95-100% of total vehicle sales, with EV sales projected to grow at an annual rate of 33%, increasing from 10 million units in 2023 to 74 million units by 2030.

Electric vehicles continue to capture market share from internal combustion engine vehicles. If electric vehicles continue to take market share from gasoline vehicles, gasoline vehicle manufacturers may be forced to restructure and consolidate.

According to Wright's Law, for every doubling of kilowatt-hour production, battery costs will decrease by 28%. Lithium iron phosphate batteries are capturing market share from nickel-rich batteries, indicating that predicting commodity prices is very challenging due to the continuous changes in battery chemistry.

Wright's Law also points to faster charging speeds for electric vehicles, which seem to represent overall performance well, including efficiency, range, and power.

In the past five years, the charging speed for vehicles with a 200-mile range has improved nearly threefold, from 40 minutes to 12 minutes, and may decrease another threefold in the next five years to reach 4 minutes. As electric vehicle charging speeds reach acceptable levels, manufacturers may optimize other features, including autonomous driving, safety, and entertainment.

Below are the main highlights of the "Big Ideas 2024" report:

1. The five major technology platforms of artificial intelligence, public blockchain, multiomic sequencing, energy storage, and robotics are merging to create synergy, potentially accelerating global economic growth from an average of 3% over the past 125 years to over 7% in the next seven years.

2. The market value of stocks related to disruptive innovation is expected to grow at an annual rate of 40%, skyrocketing from the current 16% of the global stock market total to over 60% by 2030, with market value increasing from approximately $19 trillion to about $220 trillion by 2030.

3. By 2030, the integration of hardware and software could reduce AI training costs by 75% annually. The global software market could grow tenfold by 2030.

4. By 2040, investments in AI hardware are expected to reach $1.3 trillion, which will drive AI software sales to $13 trillion, maintaining a gross margin of 75% for the software industry.

5. Robot taxi platforms will redefine personal mobility and create $28 trillion in enterprise value over the next 5-10 years, with approximately 74 million robot taxis sold annually, accounting for a significant portion of the automotive market.

6. As manufacturing integrates, battery costs are declining, lowering vehicle prices. Batteries account for 20% of the value of electric vehicles, with battery manufacturers generating $30 billion in revenue for electric vehicle original equipment manufacturers annually.

7. Thanks to 3D printing technology, automotive production is entering an unprecedented realm, expected to reduce car development time by 50% and mold design validation costs by 97%.

8. Precision therapies account for 25% of newly launched drugs, and by 2030, drug revenues will increase by 15%, approximately $300 billion.

9. Under the comprehensive penetration of AI-enhanced multiomic technologies, R&D efficiency related to drug development will double. By 2035, the actual return on investment in R&D will improve by 10%.

10. Blood tests for early detection of multiple cancers have become the standard of care, reducing cancer mortality rates by 25% in certain age groups. In developed markets, 30% of patients benefit from the new diagnostic regime.

11. Digital leisure spending is expected to gain a larger market share from the physical economy, growing at a compound annual growth rate of 19% over the next seven years, from $7 trillion in 2023 to $23 trillion by 2030.

12. By 2030, revenues from smart devices, entertainment, and social platforms are expected to reach $5.4 trillion, with advertising and commerce revenues accounting for 80%.