TLDR;
This ARK Funds webinar provides a market update and answers questions about investment strategies in disruptive innovation. Key points include adapting to the rapid pace of AI, the potential commoditization of large language models, factors influencing the healthcare sector, updates on Tempus AI, the timeline for humanoid robots, a comparison of Joby and Archer, the implications of the DOJO shutdown for Tesla, the vision for Bitmine Immersion Technologies, and the differences between USDT and USDC.
- Economic outlook anticipates slowing growth and inflation, with productivity gains driven by AI.
- Investment strategy focuses on companies with visionary leadership, strong execution, and competitive moats.
- Healthcare sector poised for growth due to improving macro conditions, strategic M&A, and a more efficient FDA.
Intro [0:00]
Dan White introduces the August market update webinar, mentioning it was pre-recorded on August 15, 2025. He then hands it over to Kathy Wood, the CEO and CIO of ARK Funds, for the market update.
How do you structure your investment process to adapt quickly to the rapid pace of AI while maintaining a long-term portfolio focus? [6:38]
Kathy Wood explains that ARK Funds is designed to focus on technologically enabled disruptive innovation. The firm sizes new markets based on learning curves and estimates how quickly technologies will scale across sectors over the next five years. The investment process includes scoring companies based on visionary leadership, execution, R&D spending, competitive moats (speed of execution), product and service leadership, and valuation, assuming multiple compression to market averages over time. Thesis risk, especially government regulation, is also considered, with excitement about deregulation.
Is there the potential for large language models to be commoditized, and is this a winner takes most market? [11:12]
Frank discusses the competition among large language model (LLM) builders like OpenAI, Google, Anthropic, and XAI, noting that the pace of catching up and leapfrogging has increased. Open-source models from China are also becoming competitive. While the field is narrowing as some companies fail to compete at the frontier, LLMs are just one layer of the AI solution stack. The market is nuanced, with consumer-side leaders like ChatGPT and fast-growing enterprise applications like coding agents (Cursor, Replit). Specialization will lead to different companies winning at different tiers, using the best models available. Factors beyond raw performance, like personality and personalization, create stickiness. Horizontal plays like Palantir are expected to capture a majority of the value in the enterprise side.
How are broader government, macro, and FDA factors influencing the healthcare sector, and how might those dynamics shape ARKG’s path toward closing the performance gap with the rest of the ARK ETF suite? [18:34]
The multiomics team discusses factors influencing the healthcare sector, including macro conditions, strategic activity, and the FDA. High interest rates have pressured healthcare innovation, making improved capital costs important for early-stage innovators. Large pharma faces a patent cliff by 2030, driving acquisitions of innovative biotech companies like Eli Lilly's acquisition of VER therapeutics. The new FDA commissioner aims to modernize the agency, streamline regulatory pathways, and reduce uncertainty. The administration's focus on China as a competitor is also influencing biotech acquisitions.
Can you please provide an update on Tempus AI following the recent earnings report? [23:19]
Nemo provides an update on Tempus AI, highlighting a strong second quarter with revenue of $340 million. Genomics revenue was $240 million, and data and services grew 35% year-over-year to $72 million. The acquisition of embryomics is performing well, signaling Tempus's interest in expanding into rare disorders and newborn sequencing. Tempus aims to be a one-stop shop for diagnostics, using an AI-first approach to build a foundational model. Kathy Wood notes that Tempus AI could be the information backbone of the US economy and a potential "winner take most" company.
According to your modeling, what does timing look like for humanoid robots at mass scale and adoption? [27:31]
Sam discusses the timeline for humanoid robots, estimating mass-scale adoption in the five-to-ten-year timeframe. The focus is on improving AI capabilities to make robots useful. While some companies aim to produce thousands of robots in the next year, mass adoption implies billions. Near-term developments include useful applications in factories and demonstrations of tasks like folding laundry.
Please compare your 5-year outlooks for Joby and Archer. [29:42]
Sam compares Joby and Archer, two companies pushing the frontier of EV toll. He emphasizes that the industry is nascent and should have multiple winners. Joby is testing different approaches and hopes to own and operate, while Archer is moving quickly and looking to sell aircraft. Both aim to commercialize in the Middle East in the next year and gain approval in the United States. Archer aims to be the official EV toll provider for the LA Olympics.
What does the DOJO shutdown mean for Tesla and its shareholders? [31:57]
Frank discusses Tesla's decision to shut down the Dojo project, which was their custom chip for training in the data center. This decision reflects a consolidation of resources and a focus on their custom chip efforts in the car, where they have a competitive differentiation. Tesla will work with Samsung on the AI6 chip in their US-based fab in Texas, aiming to make it efficient for both inference in cars and training in the data center.
What is your vision for Bitmine Immersion Technologies (BMNR)? [35:45]
Lorenzo discusses Bitmine Immersion Technologies, led by Tom Lee, which is the biggest Ethereum DAT (digital asset treasury). The main focus is to accumulate Ethereum. Bitmine is seen as an efficient wrapper for Ethereum exposure, especially since ETFs do not allow staking. Bitmine's strategy of issuing shares at a premium to buy back Ethereum has been accretive to shareholders. The company holds 1.2 million ether, generating a yearly yield of $150 million.
How are USDT and USDC different? [39:44]
Lorenzo explains the differences between USDT (Tether) and USDC (Circle), which together hold over 90% of the stablecoin market share. Technologically, they are similar, using smart contracts on Ethereum. USDC is backed by short-term US Treasury bills, while USDT is backed by a mix of short-term securities, precious metals, corporate bonds, Bitcoin, and secured loans. USDT dominates emerging markets, while USDC focuses on developed markets. Frank adds that USDT aims for stable purchasing power with the inclusion of Bitcoin and gold.