2026-06-11

Brian Armstrong on Bitcoin, Anthropic Drops Fable 5 & Mythos 5, NewLimit's $435M Age-Reversal | 264

Coinbase CEO and New Limit co-founder Brian Armstrong joins the Moonshots panel for a wide-ranging conversation spanning Bitcoin's institutional adoption and quantum-computing risk, Coinbase's rapid build-out of AI-agent payment rails, and the escalating debate over the US government taking equity stakes in frontier AI labs and potentially standing up a sovereign wealth fund. The panel also digs into the trillion-dollar IPO wave (OpenAI's S-1 filing, SpaceX, Anthropic), Google's $11B/year GPU deal with SpaceX and Elon Musk's AI1 satellite and Gigasat factory buildout toward a Dyson swarm, and closes with Armstrong detailing New Limit's $435 million raise for epigenetic age-reversal drugs headed to clinical trials in 2027. Along the way the group covers Anthropic's Fable 5 / Mythos 5 launch reclaiming the AI frontier, Apple's move to rebuild Siri on Google's Gemini, and Columbia researchers' embryo gene-editing of PCSK9 and HBG genes.

▶ Watch on YouTube

Topics

Bitcoin outlook: institutional adoption, digital gold & the countercyclical thesis Crypto/Web3 ▶ 3:45

Sparked by Citibank's $189,000 end-of-2026 Bitcoin price projection, Armstrong argues Bitcoin has become digital gold and institutional sentiment has flipped from 'rat poison' to mainstream price targets; he and Peter debate whether Bitcoin's promised countercyclical (safe-haven) behavior still holds.

AI vs crypto competition for energy, chips & capital Crypto/Web3 ▶ 9:57

The panel discusses whether Bitcoin's price weakness partly reflects miners and investors reallocating GPUs, energy, and risk capital toward AI training/inference and upcoming mega-IPOs, and whether Bitcoin and AI compute genuinely compete for the same scarce chips and power.

Quantum computing risk to Bitcoin & post-quantum migration debate Crypto/Web3 ▶ 14:39

Armstrong describes Coinbase's quantum advisory council, the Bitcoin BIP-360 post-quantum proposal, and the unresolved three-way debate over whether to freeze, preserve, or create an appeals process for pre-quantum-safe coins (including Satoshi's) once a powerful quantum computer emerges.

AI agent economy & Coinbase's agentic wallets Crypto/Web3 ▶ 20:56

Armstrong lays out Coinbase's three-layer strategy for AI-agent finance (LLM-to-account MCP access, in-app Coinbase Advisor, and self-custodial no-KYC wallets for autonomous agents), framing stablecoin rails as the likely default payment layer for a coming 'agentic economy.'

US government equity stakes in AI firms & sovereign wealth fund debate Economy ▶ 31:33

Following Trump's comments on government stakes in AI giants and existing 10% government stakes in companies like Intel and MP Materials, the panel debates quasi-nationalization of Anthropic/OpenAI, comparisons to the Atomic Energy Commission, and whether a US sovereign wealth fund or Bernie Sanders' proposed 50% equity transfer could happen.

OpenAI files S-1 amid the trillion-dollar IPO wave Economy ▶ 50:20

OpenAI's S-1 filing (following SpaceX and Anthropic) prompts discussion of CFO Sarah Friar's earlier public-readiness concerns, OpenAI's pivot from a consumer/Sora-heavy forecast to a Codex/enterprise-heavy one, and the risk of three back-to-back trillion-dollar IPOs straining available liquidity.

SpaceX becomes an AI hyperscaler: Google's $11B/yr GPU deal & the compute shortage Compute ▶ 59:24

Google is paying SpaceX $11 billion per year through 2029 for access to 110,000 Nvidia GPUs inside xAI's data center, a deal the panel reads as proof of an acute AI compute shortage and of Elon Musk repositioning SpaceX/xAI as a hyperscaler rather than a pure frontier-model lab.

SpaceX's orbital Dyson swarm buildout: AI1 satellites & Gigasat factory Space ▶ 1:05:10

The panel reacts to SpaceX's newly unveiled AI1 satellite (150kW peak compute, 70m wingspan, radiative cooling, micrometeorite shielding) and its new Texas Gigasat factory, framing them as the first mainframe-era node toward a full space-based Dyson swarm and a shift of the AI infrastructure bottleneck from real estate to launch capacity.

SpaceX valuation trajectory toward a $100 trillion company Economy ▶ 1:18:53

Morgan Stanley's projection that SpaceX revenue could grow from $18.7B (2025) to $3.4 trillion by 2040 implies a $50-100 trillion valuation; the panel weighs this against the fact no company has ever hit $1 trillion in revenue and questions whether Morgan Stanley's numbers are pre-IPO fundraising hype.

NewLimit's $435M raise & epigenetic reprogramming for age reversal Longevity ▶ 1:23:29

Armstrong explains New Limit's origin story and its approach of reprogramming a cell's age (not its type), using AI-guided high-throughput screening across 10 quadrillion protein combinations, with the first human-cell-reprogramming drug candidates entering clinical trials in 2027.

Longevity escape velocity (LEV) timing debate Longevity ▶ 1:27:07

Prompted by Kurzweil's 2033 LEV prediction, Alex argues LEV may be 'spiky' like AGI and could already be quietly underway in subpopulations, citing new epigenetic-clock evidence of age reversal from GLP-1 drugs; Peter and Armstrong discuss functional-reversal benchmarks like the $101M XPRIZE Healthspan competition.

Embryo gene editing (PCSK9/HBG) & the ethics of designer babies Biotech ▶ 1:36:02

Columbia researchers' base-editing of the PCSK9 (cholesterol) and HBG (hemoglobin) genes in embryos in vitro reopens the debate on disease-prevention vs. enhancement editing, jurisdictional loopholes (offshore clinics), and public support levels per a cited Pew Research survey.

Anthropic launches Fable 5 & Mythos 5 AI ▶ 1:42:47

Anthropic's new models (Mythos as the less-restricted version, Fable as the safety-scaffolded version) reclaim the AI frontier from GPT-5.5 per Alex, showing strong long-range visual reasoning (e.g., playing Pokémon from screen alone) alongside heavier safety guardrails and doubled pricing.

Apple rebuilds Siri on Google's Gemini AI ▶ 1:49:07

Apple's multi-year partnership to power a rebuilt, agentic Siri with Google Gemini is read as both a concession that Apple lost the foundation-model race and a bet that personal context (not the underlying model) is the real moat; the panel debates whether this represents lost tech sovereignty or a smart localization strategy.

Predictions made

open Brian Armstrong: Bitcoin will reach a much higher price than today.

EP #? · · due: 2030 · ▶ watch

“I think by 2030 we're going to have a much higher higher price and I'm long Bitcoin just like always.”

Your call:

open Brian Armstrong: Bitcoin will likely land somewhere in the $100,000-$200,000 range.

EP #? · · due: end of 2026 · ▶ watch

“something like the you know maybe the 100 to 200k range seems plausible to me by end of year.”

Your call:

open Peter Diamandis: The push for government equity stakes in frontier AI companies will continue and result in at least a 10% ownership arrangement.

EP #? · · due: unspecified · ▶ watch

“I think we're going to see a push towards at least a 10% ownership.”

Your call:

open Dave Blundin: Bernie Sanders' proposal to transfer 50% of top AI companies' equity into a public fund will not happen.

EP #? · · due: unspecified · ▶ watch

“No, I don't think it'll happen. I think Bernie's just trying to attract news.”

Your call:

open Alex Wissner-Gross: The US Treasury will take golden shares (roughly 5-10%) in OpenAI and Anthropic, and will later gradually rebalance those holdings into a broad total-market/S&P 500 index fund, effectively creating a form of universal basic income or equity.

EP #? · · due: within the next few years · ▶ watch

“it's not difficult for me to imagine a scenario where the US Treasury does take golden shares call it 5 to 10% in open AI and anthropic to complement its existing portfolio and then some amount of time from now maybe sometime in the next few years there's a great rebalancing we get our sovereign wealth fund.”

Your call:

open Peter Diamandis: The pace of AI/space change will be so extreme that the question of whether Elon Musk and Dario Amodei remain aligned partners will become moot and look quaint in hindsight.

EP #? · · due: next few years · ▶ watch

“I think there are going to be such crazy things happening in the next few years that we we'll look back and laugh at ourselves for even asking the question of whether Elon will be shaking hands with Dar.”

Your call:

open Alex Wissner-Gross: Elon Musk/SpaceX AI will pursue a strategy of becoming the dominant hyperscaler first (using Google and Anthropic compute-rental revenue), then use that compute lead to leapfrog back to frontier AI model status.

EP #? · · due: unspecified · ▶ watch

“his thinking maybe speculatively he can buy enough compute to buy his way back to the frontier.”

Your call:

open Brian Armstrong: New Limit's first reprogramming drug candidates will enter clinical trials, with more to follow.

EP #? · · due: 2027 · ▶ watch

“we've actually been able to demonstrate successful reprogramming of human cells now in our first drug candidates going into the clinic next year, hopefully followed by a bunch more.”

Your call:

open Brian Armstrong: Longevity escape velocity could plausibly be reached, echoing Kurzweil's estimate.

EP #? · · due: 2033 · ▶ watch

“I don't think it'll happen this year or next year, but um 2033 could could potentially happen.”

Your call:

open Alex Wissner-Gross: Longevity escape velocity may already be achieved in a spiky, subpopulation-specific way that goes largely unnoticed, similar to how AGI quietly passed the Turing test.

EP #? · · due: 2026 (possibly already, or next year) · ▶ watch

“on certain spikes it seems possible it might be achieved later this year... as soon as this year we're going to in a spiky way in a subpopulation fly past LEV without broader notice that it just happened.”

Your call:

open Alex Wissner-Gross: Apple will strike a Gemini-analogous AI partnership with one of the major Chinese AI labs or tech companies to power Siri in China.

EP #? · · due: unspecified · ▶ watch

“focusing just on China, I would expect Apple to strike some partnership that's roughly analogous to the partnership with Google with one of the three big Chinese AI labs or tech companies.”

Your call:

open Dave Blundin: If Apple does not act aggressively, it will lose the Siri/interface layer to Android and Google, while TSMC shifts chip manufacturing capacity away from Apple toward higher-paying AI data-center customers, pushing phone hardware manufacturing toward Samsung.

EP #? · · due: unspecified · ▶ watch

“The way it would play out if Apple doesn't do something aggressive uh is you lose the interface... you've lost your hardware manufacturing under the covers and you've lost the interface on the other side and then everybody just moves to Android.”

Your call:

Numbers that matter

Citibank projects Bitcoin could reach up to $189,000 by end of 2026 Cited as the story hook and a sign of institutional bullishness on Bitcoin.
~30% of Bitcoin capital treats it as an inflation hedge/digital gold vs. ~70% as a volatile risk asset Armstrong's estimate of how the market currently trades Bitcoin versus the intended countercyclical thesis.
Polymarket prediction market puts Bitcoin at $84,000 by end of 2026 Contrasted against Citibank's much higher $189,000 projection, prompting discussion of how wide forecasts diverge.
Coinbase network agents had done about 3.1 million transactions and a little over $1 million in value transfer (initial figure cited) Early data point on AI agents acting as autonomous customers via crypto wallets.
Updated: roughly 100 million agent transactions and about $50 million in value transferred Armstrong corrects the initial figures upward, showing rapid growth in agent-driven crypto payments.
US government holds equity stakes including 10% of Intel, 15% of MP Materials, 10% of Lithium Americas, 10% of Trilogy Metals, 10% of USA Rare Earth, and 10% of Korea Zinc Cited as precedent for a potential 10% government stake in frontier AI labs.
Bernie Sanders proposed transferring 50% of equity from top AI companies into a public fund Sam Altman reportedly rejected this as 'not going to happen, that's way too much,' but the panel sees it as opening a negotiation.
Polymarket odds on OpenAI's IPO: 46% say it comes out at $1.5 trillion or greater valuation, 26% say the IPO won't happen this year Following OpenAI's S-1 filing, framed as the third trillion-dollar IPO after Anthropic and SpaceX.
Anthropic reportedly growing revenue at 640% per year Cited as a reason coming out third in the IPO sequence (after SpaceX and Anthropic) is risky for OpenAI.
Google is paying SpaceX $11 billion per year through 2029 for access to 110,000 Nvidia GPUs inside xAI's data center Framed as evidence SpaceX has become a hyperscaler and that AI compute is in acute shortage.
SpaceX's IPO valuation multiple reportedly moved from 100x revenue to 50x revenue on the strength of this one Google GPU deal Quoted from Austin Reef, illustrating how much a single infrastructure contract can move a valuation multiple.
AI1 satellite specs: 150 kW peak compute, 70 kW per ton, ~2-ton satellite, 70-meter wingspan, 110 square meters of deployable radiative cooling Specs of SpaceX/xAI's newly unveiled AI1 satellite, the first node toward a planned Dyson swarm.
Target launch cost of $100 per kilogram; original filing for 500,000 AI1 satellites later increased to 1 million satellites Discussion of the scale of SpaceX's orbital compute ambitions and Starship's role in enabling it.
SpaceX's new Gigasat factory spans 1,000 acres with capacity for 11 million square feet of facilities, targeting AI1 satellite production in late 2027 Compared to SpaceX's Gigafactory playbook for Optimus production; emphasizes full vertical integration from solar ingots to finished satellites.
Polymarket expects SpaceX's valuation to hit $2.13 trillion at the close of its first trading day Sets up the broader discussion of SpaceX's long-term valuation trajectory.

Worth digging into

🕳️ BIP-360 and the post-quantum Bitcoin migration fight

Roughly 5-10% of all Bitcoin supply (including Satoshi's) is vulnerable to a future quantum computer, and the community is split three ways on whether to freeze, preserve, or create an appeal process for unmigrated coins -- a live governance battle with billions at stake.

🕳️ On-chain reputation ('FICO score') system for AI agents

Armstrong floats a PageRank-style trust graph to reduce fraud in agent-to-agent commerce, but no concrete spec or product exists yet -- a foundational building block for the agentic economy he describes.

🕳️ New Limit's epigenetic reprogramming clinical pipeline

The claim of successful human-cell reprogramming with drug candidates entering clinic in 2027 is specific and checkable, and directly bears on the longevity-escape-velocity debate later in the episode.

🕳️ US Treasury golden shares / sovereign wealth fund for AI

Alex's specific scenario (5-10% Treasury stakes in OpenAI/Anthropic later rebalanced into a total-market index) is a concrete policy thread tied to an actual executive order exploring a sovereign wealth fund.

🕳️ Morgan Stanley's $3.4 trillion SpaceX revenue-by-2040 forecast and the Google-SpaceX GPU deal terms

Both are extraordinary, high-stakes financial claims released just ahead of SpaceX's IPO, and the panel itself flags the Morgan Stanley note as a likely fundraising tool.

🕳️ Columbia's PCSK9/HBG embryo base-editing study

This is reportedly a first-of-its-kind editing of these specific genes in human embryos in vitro, directly reopening 'designer baby' governance questions the panel says will require serious guardrails.