Coach Kyle

Anthropic: The Rise, The Reckoning, and a Priceless Future

Newsletter LibraryCompany researchSpecial Edition

G.L.A. INVEST RESEARCH · Issue dated May 11, 2026 Private Markets · AI Infrastructure · Pre-IPO Watchlist

View original edition ↗Reader disclosures
Anthropic: Business Growth, Compute Partnerships and Research Risks newsletter cover
Original artwork published with this archived edition.

Anthropic: The Rise, The Reckoning, and a Priceless Future

Original illustration from this archived newsletter

G.L.A. INVEST RESEARCH · Issue dated May 11, 2026
Private Markets · AI Infrastructure · Pre-IPO Watchlist

The Headline

In the span of three years, Anthropic has gone from a safety-focused OpenAI spinout to one of the most strategically important — and arguably priceless — private companies on Earth. As of today, secondary shares are bid around $1,400 per share on Hiive (the live tape on Hiive is showing $1,446.25 with 38 live orders as of May 11), implying a roughly $1 trillion+ valuation on the private market (Hiive; Yahoo Finance).

Last week’s SpaceX partnership changed the conversation entirely. An IPO is expected later this year. And on the All-In Podcast over the weekend, the besties effectively called this the “Mag-1” moment (Shortform – All-In May 8, 2026).

Here is what every serious investor needs to know.

Original illustration from this archived newsletter

Part 1 · The History: From OpenAI Defectors to a Trillion-Dollar Force

Anthropic was founded in early 2021 by Dario Amodei (former VP of Research at OpenAI), his sister Daniela Amodei, and seven other senior ex-OpenAI researchers who left over directional and safety disagreements (Business Insider; Wikipedia). The name itself — Anthropic — was chosen to signal the firm’s human-centered mission.

The company’s early bet was contrarian: build the most capable model in the world, but anchor it to a safety framework called Constitutional AI — a method of training models using a written constitution rather than purely human feedback (Anthropic Research; Claude’s Constitution).

By summer 2022, Claude was already trained — but Dario delayed launch out of concern over real-world impact. That restraint became a brand. When Claude was released, enterprises ate it up.

Key funding milestones:

Round

Investors

Note

2023 – 2024

Google ($3B), Amazon ($4B initial)

Cloud-aligned, capped below 33% to preserve independence (TechFundingNews)

2025

Amazon top-up (~$8B total)

Stake later marked at $70B+ on Amazon’s books (Fortune)

Late 2025

Microsoft + Nvidia (~$15B); $30B Azure commitment

Multi-cloud strategy fully unlocked (Yahoo Finance)

Feb 2026

Series G – ~$30.6B raised of $61.15B total

Per Hiive’s tape (Hiive)

Current talks

$50B at $850B–$1T pre-money

Dragoneer, General Catalyst, Lightspeed reportedly at the table (TechCrunch; The AI Corner)

Original illustration from this archived newsletter

Part 2 · The Uprising: A Revenue Curve Salesforce Took 20 Years to Hit

This is the single most important chart in AI right now.

Anthropic ARR trajectory (VentureBeat):

  • Jan 2024: $87M run rate

  • Dec 2024: $1B

  • Year-end 2025: $9B

  • Feb 2026: $14B

  • Mar 2026: $19B

  • Apr 2026: $30B run rate

For context: Salesforce took ~20 years to reach $30B in annual revenue. Anthropic did it in under three years.

Original illustration from this archived newsletter

This is the curve that broke the secondary market. Anthropic’s implied valuation has tripled in three months on private exchanges (Yahoo Finance; Forge / Jupiter Prestocks).

Part 3 · The SpaceX Partnership — Why Anthropic Is Now Priceless

On May 6, 2026, Anthropic announced it has secured access to the full capacity of SpaceX’s Colossus 1 data center in Memphis — over 300 megawatts of additional compute and 220,000+ Nvidia H100/H200/GB200 GPUs (Axios; NBC News).

The bigger story — and the reason we are calling Anthropic effectively priceless — is what comes next:

“As part of this agreement, Anthropic has also shown interest in collaborating to create multiple gigawatts of orbital AI computing capacity.” — SpaceX blog, May 6, 2026 (Axios)

Original illustration from this archived newsletter

That is space-based data centers. Anthropic now has a contractual on-ramp to the only compute frontier that scales beyond Earth’s grid limits — and it’s locked in with the one company on Earth (and off it) that can deliver mass to orbit at cost.

Recall the backdrop: SpaceX acquired xAI in January 2026 and has an option on Cursor for $60B. The combined “Elon Web Services” stack — launch, satellites, compute, AI labs — now has Anthropic as its flagship enterprise tenant (Axios).

Why this changes the valuation math:

  1. The compute constraint is solved structurally, not just contractually

  2. Two pre-IPO behemoths are now economically interlinked — buying Anthropic is partial exposure to SpaceX’s AI ambitions

  3. Orbital compute optionality is uncapped — there is no public comparable for this asset

  4. It defangs the OpenAI/Stargate narrative — Anthropic is no longer the “compute-starved” lab

Part 4 · All-In Podcast — Weekend Commentary (May 8, 2026 Episode)

In “Elon’s Anthropic Deal, The Next AI Monopoly?, ‘FDA for AI’ Panic, Trading the AI Boom” — released Friday — the besties devoted the first ~25 minutes to this exact story (Apple Podcasts; Shortform recap).

Key takeaways from the hosts:

  • Chamath Palihapitiya and David Sacks framed Anthropic’s growth as unprecedented — noting that while Google grows ~20% YoY, Anthropic’s acceleration is in a category that has never existed before in software.

  • The panel floated that we may be transitioning from a “Mag-7” world to a “Mag-1” — with Anthropic positioned as the singular dominant AI company if current curves hold.

  • The SpaceX deal was characterized as a turning point — Sacks called it the moment “Elon Web Services” became real, with Anthropic as its anchor customer.

  • Jason Calacanis raised the wealth-distribution question: with AI companies compounding this fast, he proposed tech leaders pledge 1% of holdings annually for 20 years (healthcare/education/housing), and floated AI IPOs allocating 1–5% of shares to ordinary Americans via “Invest America” accounts.

  • The hosts also debated whether Anthropic’s Mythos model — and the White House’s reported “FDA for AI” framework — represents a real risk or political theater.

G.L.A. read: when Chamath and Sacks — historically skeptical of valuation excess — start using the phrase “next great monopoly” unironically, that is a signal worth marking.

Part 5 · The IPO Is Coming — This Year

Original illustration from this archived newsletter

Multiple credible reports now converge on a 2026 IPO:

  • The Information first reported a target of a ~$60B raise via IPO, potentially as early as October 2026, against a private mark already at $380B at the time of reporting (MarketWise)

  • Anthropic engaged Wilson Sonsini for IPO prep and has held discussions with major investment banks (Yahoo Finance / FT)

  • LinkedIn News (May 8) reported the implied valuation on private exchanges has touched $1.2 trillion (LinkedIn News)

If pricing holds, this will rank among the largest IPOs in history — and unlike most mega-IPOs, the demand side is reinforced by every public AI proxy trade investors have been forced into for two years.

Part 6 · The G.L.A. Take — How We’re Thinking About Position Sizing

Bull case (base):

  • $30B ARR growing 80x in 28 months — even a 50% deceleration leaves a ~$60B+ ARR run rate at IPO

  • SpaceX compute lock removes the single biggest bear thesis (capacity constraint)

  • Multi-cloud distribution via Amazon, Google, Microsoft, Azure → enterprise sales flywheel

  • Constitutional AI / safety brand = the preferred vendor for regulated industries

Risks we are watching:

  • Regulatory: “FDA for AI” / Mythos oversight could compress release cadence

  • Margins: Inference costs at scale remain opaque; gross margin disclosure at S-1 will reset expectations either direction

  • Concentration: Heavy revenue from a small number of enterprise + API customers

  • Political crosswinds: David Sacks (now AI Czar) has publicly criticized Anthropic’s hiring of former Biden officials (KSAT / All-In context)

  • Secondary entry price: At ~$1,400/share, you are paying ~$1T for a company that may price its IPO meaningfully below current secondary marks — historical pattern for hot pre-IPOs

What we are doing at G.L.A.:

  • Maintaining existing private allocations

  • Not chasing the Hiive bid above $1,500/share absent clearer IPO terms

  • Tracking the Series H announcement (expected within 60 days) as the next pricing anchor

  • Building a comp sheet against OpenAI, xAI, and Cursor for relative-value framework

Bottom Line

Three years ago, Anthropic was a safety lab.
Today, it is on the doorstep of being the first trillion-dollar IPO of the AI era, with SpaceX as its compute partner and a revenue curve that has no precedent in enterprise software history.

Is it priceless? On the orbital-compute optionality alone — very arguably yes.
Is it a buy at $1,400 on Hiive? Depends on your hold horizon and your view on IPO discount mechanics.

” I view Anthropic as a clear longterm exponential growth investment with great mgmt, great backers, great partners, and the minute it comes public it will jump to our top ten “Apex 10” portfolio.” – Coach Kyle

Original illustration from this archived newsletter

G.L.A. INVEST RESEARCH · Public + Private Equity Research
Not investment advice. Private securities are illiquid and accredited-investor only. Do your own due diligence.