Cathie Wood: “AI and Bitcoin Are Moving Faster Than We Expected” | EP #226
Cathie Wood of ARK Invest walks the Moonshots panel through highlights of ARK's 2026 Big Ideas report, arguing that AI, robotics, energy storage, blockchain and multiomic sequencing are converging to push global real GDP growth to 7%+ (a 2.5x step change she calls conservative), while inflation heads below 2% and negative on falling AI, energy and manufacturing costs. She reiterates ARK's bull-case Bitcoin target of $1.5 million by 2030, discusses gold's re-emerging correlation with Bitcoin, and argues GDP as a metric is fundamentally broken in an AI economy because AI-driven abundance and delegation to agents distort what is measured. The group also covers the viral open-source agent 'Claudebot' (renamed Moltbot), the US-China open-source AI race following DeepSeek, orbital data centers and Wright's Law cost curves for rockets and robots, nuclear energy's regulatory-driven cost overhang, and ARK's thesis that Tesla will beat Waymo in robotaxis on vertical-integration cost advantages.
The Great Acceleration: AI-driven GDP step changeEconomy▶ 0:07:01
Cathie Wood argues every technology revolution produces a step-function jump in GDP growth (0.6% pre-1900, 3% for 125 years after), and the current convergence of five platforms (robotics, energy storage, AI, blockchain, multiomic sequencing) should push growth to 7%+, which she calls conservative.
Orbital data centers and Wright's Law cost curvesSpace▶ 0:13:15
ARK's updated SpaceX model (with Mach33) now includes orbital data centers; the panel debates whether launch-cost declines alone capture the real cost curve, since Musk is reportedly also building in-house GPU fabs to bypass TSMC/Nvidia margins and leveraging space's 6x more efficient solar power.
AI inference cost collapse and Jevons paradoxCompute▶ 0:19:12
Token/inference costs are collapsing at an extraordinary rate, but Alex Wissner-Gross argues demand for intelligence is essentially infinite, so falling unit costs drive explosive unit growth (Jevons paradox) rather than shrinking total AI spend.
GDP mismeasurement in the AI age ('Kathy Wood's law')Economy▶ 0:21:51
The panel proposes that AI curing disease or automating unpaid labor (childcare, cooking) can show up as negative GDP even though it adds real value, and conversely that delegating human tasks to paid AI agents inflates GDP without adding real wealth — making GDP a broken proxy for progress.
Moltbot (formerly Claudebot): viral open-source personal AI agentAI▶ 0:31:09
An open-source personal-assistant agent that took the internet by storm over a weekend, renamed from 'Claudebot' to 'Moltbot' (mascot: a lobster) after trademark pushback from Anthropic; it connects to email, social accounts and a whole computer, which makes it powerful but risky.
Wood argues that US software firms cutting off China (over IP theft) inadvertently pushed China into open-source AI leadership (DeepSeek, plus Meta's Llama 4 stumbling), while China's economy keeps ~40% of GDP in investment versus ~20% in the US.
China's AI-driven biotech and healthcare boomBiotech▶ 0:51:29
Insilico Medicine's Hong Kong IPO was 1,200x oversubscribed; China is running far more clinical trials than the West, aided by lighter regulation and heavy government investment, though the new FDA commissioner is easing US barriers.
Bitcoin outlook, gold correlation, and stablecoinsCrypto/Web3▶ 0:57:24
ARK holds its bull-case $1.5M Bitcoin target for 2030 despite stablecoins displacing part of Bitcoin's inflation-hedge role in emerging markets; gold's recent surge and historically leading correlation with Bitcoin (near-zero the last five years) suggest Bitcoin is due for a run, reinforced by real-world cases like Iran's bitcoin-based bazaar economy.
Digital assets, tokenization, and the ICO vs IPO questionCrypto/Web3▶ 1:07:06
ARK projects digital assets could reach $28 trillion in market value; the panel discusses whether tokenized/decentralized ownership (via platforms like Robinhood) could replace or supplement traditional IPOs and provide liquidity for late-stage private companies.
Energy: nuclear regulation, Wright's Law, and the 1971 divergenceEnergy▶ 1:20:40
Wood argues 1970s nuclear regulation broke the technology's Wright's Law cost-decline curve; had it continued, US electricity would be 40% cheaper today. China is building ~28 large reactors versus none under construction in the US, though US nuclear still supplies 20% of electricity.
Robotaxis: Tesla vs Waymo cost structureTransportation▶ 1:35:14
ARK expects Tesla to be the biggest robotaxi platform winner over Waymo due to vertical integration (versus Waymo's dependence on suppliers like Zeekr/Hyundai), projecting Tesla's costs 50% below Waymo's and pricing near 20 cents/mile at scale versus Uber's rising per-mile fares.
Autonomous delivery: drones and ground robotsRobotics▶ 1:52:15
Fully autonomous delivery is scaling fast — Zipline (started delivering medical supplies in Rwanda, cutting maternal bleedout deaths by over 50%), Wing, Matternet, and ground robots like Coco Robotics — with air delivery seen as less congested than ground/road delivery for now.
Predictions made
openCathie Wood: Bitcoin reaches ARK's bull-case price target of $1.5 million per coin.
openCathie Wood: Global real GDP growth accelerates to a step-function increase of roughly 2.5x historical trend, implying 7%+ annual growth, driven by convergence of robotics, energy storage, AI, blockchain and multiomic sequencing.
EP #? · · due: unspecified (ARK's 5-year investment horizon, through 2030) · ▶ watch
“I think the 7% plus is conservative.”
Your call:
openCathie Wood: US inflation falls below 2% and turns negative.
EP #? · · due: within the next year (~2027) · ▶ watch
“I think that uh within the next year we'll see inflation below 2% and heading negative.”
Your call:
openCathie Wood: ARK's disruptive-innovation investment strategies (including flagship ARKK) compound at a 35% annualized rate.
“disruptive innovation we believe is going to compound uh in terms of returns in the market at a 35% annualized rate for the next five years.”
Your call:
openCathie Wood: Digital assets' total market value reaches $28 trillion, roughly the size of the entire 2010 US stock market and comparable to the current US economy.
EP #? · · due: unspecified (ARK Big Ideas 2026 report horizon) · ▶ watch
“digital assets could reach 28 trillion in market value”
Your call:
openCathie Wood: A $100 trillion-valuation company emerges, most likely Tesla (possibly via a merger involving SpaceX), enabled by proprietary-data convergence across Musk's companies.
“I could see, yes, 100 trillion. I think it's going to happen because of convergence.”
Your call:
openCathie Wood: A Robinhood-style tokenized/decentralized private-company ownership model emerges as an alternative to traditional IPOs for late-stage private companies.
EP #? · · due: within the next three years (~2029) · ▶ watch
“I I think it's very possible in the next in the next three years.”
Your call:
openPeter Diamandis: Autonomous (robotaxi) vehicles will make up about 80% of vehicles on the road in his area (Santa Monica).
EP #? · · due: in about four or five years (~2030-2031) · ▶ watch
“I'm imagining in about four or five years it's going to be 80% autonomous vehicles.”
Your call:
openCathie Wood: Tesla's robotaxi cost structure will run about 50% below Waymo's, allowing Tesla to price around 20 cents per mile at scale.
EP #? · · due: unspecified ('between now and then') · ▶ watch
“we think that Tesla's solution from a cost point of view will be 50% lower than Whimo's... Tesla will be able to price at 20 cents per mile at scale”
Your call:
openCathie Wood: Cumulative global investment in power infrastructure needs to rise to $10 trillion.
“Cumulative investment in global power needs to increase to 10 trillion by 2030.”
Your call:
Numbers that matter
ARK's global GDP growth forecast: 7%+ per yearCathie Wood calls this conservative; roughly double the IMF's outlook and about 2.5x the ~3% growth rate that has held since the last technology revolution.
Historical GDP growth: ~0.6% (1500-1900) stepping up 5x to ~3% for the next 125 yearsBasis for ARK's claim that each technology revolution produces a step-function GDP increase; sourced to chief futurist Brett Winton's academic research.
Industrial robot costs decline roughly 20%+ for every cumulative doubling of units producedWright's Law learning-curve example cited alongside rocket cost declines.
TSMC captures ~50% margin and Nvidia ~80% margin on GPU chipsDave Blundin describing cost layers Musk is reportedly trying to bypass with in-house chip fabs for orbital data centers.
NASA space shuttle launch cost ~$600 million vs. SpaceX launch ~$60 million, with another 10x reduction expectedCited as an example of deflationary technology cost curves feeding into the Jevons-paradox GDP debate.
Trueflation (real-time index of 10,000 items) shows inflation already down to 1.2-2%Contrasted with the Fed's official 2-3% inflation framing.
Cognition/intelligence is being commoditized at roughly 99% per yearReferring to the rate at which AI cost-per-unit-of-intelligence is falling, per the AI infrastructure slide discussed.
Uber = ~1% of urban miles traveled in the US; scaling to that share needs only ~140,000 cars, while 100% of urban miles would need ~24 million carsCompared against ~400 million cars owned and ~15 million new cars sold per year in the US, underpinning ARK's robotaxi disruption thesis.
OpenAI reportedly planning to charge ~$60 per thousand ad views vs. Facebook's ~$20 per thousandDescribed as 'Super Bowl kinds of pricing' as OpenAI moves to monetize via advertising.
OpenAI has about 900 million usersCited as OpenAI's competitive head start despite monetization/margin pressure risk from Gemini.
OpenAI's reported internal mandate: $75 billion in ad revenue within ~18 months to two years, from zeroCompared to Amazon's advertising business reaching ~$50 billion after about 7 years of building it.
AI agents show an 80% success rate on long-duration tasksPeter Diamandis notes a human employee at 80% success would get fired; Alex Wissner-Gross counters that running 100 parallel agents pushes effective success toward 90%+.
ARKK has returned roughly 31-33% annualized over the last two yearsCited ahead of ARK's stated goal of 35% annualized returns for its disruptive-innovation strategies over the next five years.
Investment as a share of GDP: ~20%+ in the US vs. ~40% in ChinaChina's investment share has stayed elevated since joining the WTO, now redirected from property toward technology under Xi Jinping's 'new productive forces' push.
China is building roughly 28 large nuclear reactors versus none currently under construction in the USCited as part of China's energy buildout for AI/data-center dominance; US nuclear still supplies about 20% of US electricity generation.
Worth digging into
🕳️ Elon Musk's rumored in-house GPU fab and orbital data center economics
If Musk is bypassing TSMC's ~50% and Nvidia's ~80% margins with in-house chip fabs, plus using 6x-more-efficient space solar power, the true cost curve for orbital compute could be far steeper than ARK's current model shows — a potentially huge, underpriced disruption.
🕳️ Moltbot (formerly Claudebot): viral open-source personal AI agent
An agent that can read a user's email, social accounts, and full computer access, exploded in adoption over a single weekend and is explicitly flagged as capable of 'scrambling your entire computer in two seconds' — a live case study in agentic AI's power/safety tradeoff.
🕳️ GDP vs GNI mismeasurement in the AI economy ('Kathy Wood's law')
Wood argues productivity growth is understated by ~2%/year, meaning both real GDP growth is undercounted and inflation is overcounted — a claim that, if right, would materially change how policymakers and the Fed should be reading the economy.
🕳️ China's AI-and-biotech capital markets boom (Insilico Medicine's 1,200x oversubscribed Hong Kong IPO)
A 1,200x oversubscription rate signals extraordinary capital hunger for Chinese AI-driven biotech just as US-China competition in health/biotech intensifies and US FDA regulation eases.
🕳️ Tesla vs Waymo robotaxi cost-structure bet
ARK's specific, falsifiable claim — Tesla 50% cheaper than Waymo, 20 cents/mile at scale — is a concrete prediction that can be tracked against real pricing and market-share data as robotaxis scale.
🕳️ Bitcoin vs. stablecoins in currency-collapse states (Iran case study)
Peter Diamandis's anecdote about Bitcoin being the only functional transaction medium at Iranian bazaars amid a currency collapse and revolution offers a real-time natural experiment on ARK's thesis that emerging-market wealth eventually shifts from transactional stablecoin use to Bitcoin as a store of value.