Meta to Sell Excess Compute to Fund…

1. Key Themes

Theme 1: Big Tech Entering the AI Infrastructure Marketplace, Threatening Neoclouds

Meta's move to monetize excess compute signals that hyperscalers are now competing directly with pure-play AI cloud providers — compressing the market opportunity for independent players.

"Meta shares jumped 10% after reports that the company plans to sell excess AI compute to outside customers, a move that could help offset its planned $145 billion capex buildout while pressuring neocloud rivals CoreWeave and Nebius."

Theme 2: AI Infrastructure Capital Intensity Continues to Accelerate

Massive rounds across the infrastructure stack — compute, open-source model hosting, chip design — confirm that the infrastructure layer remains the dominant investment magnet in AI right now.

"Together AI...raised an $800 million round at an $8.3 billion post-money valuation...providing computing infrastructure for companies to train and run open-source AI models at scale."

"Oxmiq...develops chip-design architecture and software to lower the cost of building and running AI applications, raised a $35 million round."

Theme 3: AI-Accelerated Roll-Up as a Public Market Strategy

Bending Spoons' blockbuster debut validates a PE-style roll-up model for consumer internet brands — when turbo-charged with AI and taken public in a receptive market window.

"Bending Spoons applied some of the private equity playbook to a long series of acquisitions — Meetup, Eventbrite, Vimeo, WeTransfer, and many others. But it is not a flip-and-sell scheme: it wants to transform these companies with tech and then hold onto them."

"'In the past year and a half, we've witnessed an incredible acceleration in the pace at which we were able to ship new features and create value for users,' Danieli told TechCrunch."

Theme 4: Dot-Com-Era Warning Signs in AI Hardware vs. Software Divergence

The market structure increasingly mirrors the late 1990s tech bubble, with infrastructure suppliers outperforming the major AI spenders — a potential leading indicator of a correction.

"JPMorgan says the AI trade is flashing a dot-com-era warning, with chip and memory stocks soaring while major AI spenders like Microsoft and Meta lag, echoing the 1999 split between infrastructure suppliers and the companies pouring capital into the buildout."

Theme 5: AI Compute Geopolitics as a Corporate Competency

Securing chips and navigating export policy has evolved from a procurement function into a senior executive and diplomatic role at leading AI labs.

"Business Insider profiles Tom Brown, Anthropic's 39-year-old co-founder and chief compute officer, whose role securing chips and cloud capacity has expanded into diplomacy after he helped negotiate the Trump administration's rollback of export restrictions on Claude Fable 5 and Mythos 5."

2. Contrarian Perspectives

"Root for the AI Crash" — A Bubble Burst Could Be Net Positive

Former a16z Bio + Health head Vijay Pande argues that the rational investor move is to welcome an AI correction, not fear it — because the durable infrastructure and behavioral habits that survive a bust are more valuable than the inflated valuations that precede it.

"Vijay Pande says investors should 'root for the crash' in AI because a bubble burst would clear out inflated valuations, discipline feverish data-center spending, and leave behind the infrastructure and habits needed for a more durable AI economy."

Leading AI Model Companies Are Overcharging and Harvesting Customer Data

Palantir CEO Alex Karp offers a pointed insider critique of the AI model industry's business practices — suggesting that enterprise CEOs are privately far more skeptical of the major labs than their public commitments imply.

"He called the AI industry 'effing insane' in a CNBC interview, accusing leading model companies of overcharging customers, harvesting their data, and leaving CEOs privately 'livid.'"

Success Is About Minimizing Luck, Not Maximizing Risk

Bending Spoons' founding philosophy directly challenges the VC-culture narrative of bold, high-variance bets — positioning operational discipline and process as the real driver of compounding returns.

"After $18B IPO, Bending Spoons Founder Says Success Comes From Minimizing Luck."

3. Companies Identified

Meta

Description: Global social media and AI hyperscaler
Why mentioned: Plans to sell excess AI compute to third parties to help fund $145B capex buildout, pressuring neocloud competitors
Quote: "Meta shares jumped 10% after reports that the company plans to sell excess AI compute to outside customers, a move that could help offset its planned $145 billion capex buildout while pressuring neocloud rivals CoreWeave and Nebius."

Bending Spoons

Description: 13-year-old Milan-based consumer internet roll-up company; owns AOL, Vimeo, Eventbrite, WeTransfer, Brightcove, Meetup
Why mentioned: Nasdaq IPO debut with 40% first-day pop; $25.7B market cap; case study in AI-powered acquisition-and-hold model
Quote: "Bending Spoons shares closed nearly 40% higher in their Nasdaq debut, giving the Milan company behind AOL, Vimeo, Eventbrite, and Brightcove a $25.7 billion market value."

CoreWeave & Nebius

Description: Neocloud AI infrastructure providers
Why mentioned: Identified as direct competitive casualties of Meta's compute-selling move
Quote: "A move that could help offset its planned $145 billion capex buildout while pressuring neocloud rivals CoreWeave and Nebius."

Together AI

Description: Four-year-old San Francisco startup; computing infrastructure for open-source AI model training and inference at scale
Why mentioned: Raised $800M at $8.3B valuation — one of the largest AI infrastructure rounds in the newsletter
Quote: "Together AI...raised an $800 million round at an $8.3 billion post-money valuation...led by Aramco Ventures, with Vista Equity Partners, General Catalyst, Emergence Capital, Nvidia, MarchCapital, Pegatron, and S Ventures also investing."

SpaceX / xAI

Description: Elon Musk's aerospace/tech conglomerate and AI lab
Why mentioned: WSJ reported a prototype slim AI handset running its own OS with Qualcomm Snapdragon chip and xAI integration; Musk denies the report
Quote: "SpaceX showed investors a prototype of a slim, handset-like AI device designed to run on its own operating system, use a Qualcomm Snapdragon chip, and integrate xAI technology...Elon Musk calls the news 'utterly false.'"

Wayve

Description: Nine-year-old U.K. autonomous vehicle software company; uses data-trained models rather than prebuilt maps
Why mentioned: Conducted $85M employee tender offer at $8.5B valuation via the London Stock Exchange's new Private Securities Market — first major company to use it
Quote: "It is conducting the sale through the London Stock Exchange's new Private Securities Market, the first major company to test it."

TwelveLabs

Description: Five-year-old San Francisco startup; video intelligence platform for search, analysis, and structuring of large video archives
Why mentioned: Raised $100M Series B; notable cross-sector investor syndicate including Amazon, Index Ventures, and Red Bull Ventures
Quote: "TwelveLabs...helps media, government, and other organizations search, analyze, and turn large video archives into structured data."

Venice AI

Description: Two-year-old New York startup; privacy-focused access to 200+ AI models across modalities
Why mentioned: Raised $65M Series A at $1B valuation; backed by crypto-native investors (Dragonfly, Coinbase Ventures), suggesting crypto-AI convergence thesis
Quote: "Venice AI...offers privacy-focused access to more than 200 AI models for text, image, audio, and video generation, raised a $65 million Series A round at a $1 billion valuation."

Addi

Description: Eight-year-old Bogotá fintech; credit, payments, and financial services for Colombian consumers and merchants
Why mentioned: Raised $85M Series D; signals continued institutional appetite for LatAm fintech
Quote: "Addi...provides credit, payments, and financial services to Colombian consumers and merchants, raised an $85 million Series D round co-led by Citius and BTG Pactual."

Beeline Medicines

Description: One-year-old Boston biotech; precision therapies for autoimmune and inflammatory diseases including lupus
Why mentioned: Raised $126M Series A extension (total $426M) in just one year — notable capital efficiency signal for precision medicine
Quote: "Beeline Medicines...raised a $126 million Series A extension...The company has raised a total of $426.3 million."

InKind

Description: 12-year-old Austin company; provides restaurants upfront capital by purchasing future dining credits, sold to diners via rewards app
Why mentioned: Raised $320M from Liberty Mutual Investments — large non-traditional institutional round for a restaurant fintech
Quote: "InKind...provides restaurants with upfront capital by buying future dining credits and selling them to diners through a rewards app, raised a $320 million round."

Klarna / PriceRunner

Description: Klarna-owned price comparison platform
Why mentioned: Won $1.5B antitrust damages verdict against Google in Sweden — largest competition award ever issued by a Swedish court
Quote: "A Swedish court ordered Google to pay Klarna-owned PriceRunner about $1.5 billion in antitrust damages for favoring its own shopping service in search results."

Aligned

Description: Five-year-old New York/Tel Aviv startup; AI agents for B2B sales process coordination from first call to close
Why mentioned: Raised $60M Series B — notable signal for AI in enterprise sales automation
Quote: "Aligned...deploys AI agents to centralize deal materials and coordinate complex B2B sales processes from first call to close."

Oxmiq

Description: Three-year-old Campbell, CA startup; chip-design architecture and software to lower AI application costs
Why mentioned: Samsung Catalyst Fund and MediaTek backing signals strategic hardware-level interest in cost-reduction of AI inference
Quote: "Oxmiq...develops chip-design architecture and software to lower the cost of building and running AI applications, raised a $35 million round."

Passport

Description: Nine-year-old San Francisco cross-border e-commerce logistics startup
Why mentioned: Acquired by Global-e for $350M upfront + $75M earnout on $54M raised — strong return for investors including Kleiner Perkins and TCV
Quote: "Global-e is acquiring Passport...for $350 million upfront and up to $75 million more...The company raised a total of $54 million."

Raylu (Sponsor — included as notable company)

Description: AI platform for VC deal sourcing and founder outreach
Why mentioned: Claims 4x reply rates on automated founder outbound; trusted by 50+ leading investment funds
Quote: "Raylu's AI agents find companies matching your thesis, score them against your firm's investment criteria, sync bi-directionally with your CRM, and run automated founder outbound that hits 4x reply rates."

4. People Identified

Matteo Danieli

Description: Co-founder and Chief Product Officer, Bending Spoons
Why mentioned: Central voice explaining the company's acquisition and AI-transformation philosophy at IPO
Quote: "We want to place ourselves as an operator that takes beloved brands and makes them much better."

Alex Karp

Description: CEO, Palantir
Why mentioned: Delivered a sharp public critique of the AI model industry's pricing and data practices; positioned Palantir/Nvidia partnership as a national-security alternative
Quote: "He called the AI industry 'effing insane'...accusing leading model companies of overcharging customers, harvesting their data, and leaving CEOs privately 'livid.'"

Tom Brown

Description: Co-founder and Chief Compute Officer, Anthropic; age 39
Why mentioned: Profiled for expanding his role from chip procurement into geopolitical diplomacy, including negotiating Trump administration export restriction rollbacks
Quote: "His role securing chips and cloud capacity has expanded into diplomacy after he helped negotiate the Trump administration's rollback of export restrictions on Claude Fable 5 and Mythos 5."

Vijay Pande

Description: Former head of a16z Bio + Health
Why mentioned: Made a contrarian public call to "root for the AI crash," arguing a correction would produce more durable AI infrastructure outcomes
Quote: "Investors should 'root for the crash' in AI because a bubble burst would clear out inflated valuations, discipline feverish data-center spending, and leave behind the infrastructure and habits needed for a more durable AI economy."

Ashton Kutcher

Description: Co-founder, Sound Ventures (departing); launching new VC firm focused on early-stage AI infrastructure, energy, and deep tech
Why mentioned: High-profile fund transition; partnering with former NFX partner Morgan Beller signals a thesis shift toward hard infrastructure
Quote: "Ashton Kutcher is leaving Sound Ventures...to launch a new VC firm with former NFX partner Morgan Beller focused on early-stage AI infrastructure, energy, and deep tech startups."

Morgan Beller

Description: Former partner at NFX; joining Ashton Kutcher's new fund
Why mentioned: Operationally credible co-founder for the new deep tech/AI infrastructure fund
Quote: "Ashton Kutcher is leaving Sound Ventures...to launch a new VC firm with former NFX partner Morgan Beller focused on early-stage AI infrastructure, energy, and deep tech startups."

Luca Ferrari, Francesco Patarnello, Luca Querella

Description: Co-founders, Bending Spoons
Why mentioned: Named as the founding team who built the company from the lessons of a failed AI startup (Evertale) into an $18B+ public company
Quote: "It taught lessons to the cofounders and team members who now lead Bending Spoons — Luca Ferrari, Francesco Patarnello, Luca Querella, and Danieli himself."

5. Operating Insights

AI as a Feature-Shipping Accelerant — Not Just a Product Category

For operators running mature or acquired software businesses, AI's most immediate ROI may be internal development velocity rather than new product lines. Bending Spoons' story suggests that AI-powered feature shipping is what justifies premium valuations on "aging" assets.

"'In the past year and a half, we've witnessed an incredible acceleration in the pace at which we were able to ship new features and create value for users,' Danieli told TechCrunch."

Failure as Prerequisite — Pre-Validate Your Operating Model Before Scaling

Bending Spoons' success is rooted in a prior failed startup (Evertale) that gave the founding team hard-won lessons before they deployed capital at scale. Operators should treat early-stage failures as curriculum, not disqualifiers.

"That startup failed, but it taught lessons to the cofounders and team members who now lead Bending Spoons."

Monetization Paywalls on Accessibility Features Carry Brand Risk

Meta's decision to limit its voice-amplification accessibility feature to six minutes per day on the free tier — and charge $19.99/month for full access — is a cautionary example of how monetization decisions on features that users perceive as essential (especially accessibility tools) can generate outsized reputational blowback.

"Meta is capping its on-device voice-amplification feature, Conversation Focus, at three hours a month (about six minutes a day) — unless you pay $19.99/month for Meta One Premium."

6. Overlooked Insights

The London Stock Exchange's Private Securities Market as a New Liquidity Mechanism

Wayve's use of the LSE's newly launched Private Securities Market for its employee tender offer is a meaningful structural development in private market liquidity — particularly relevant for late-stage private companies and their employees seeking exits outside of traditional IPO or M&A paths. If it gains traction, it could reshape how European (and eventually global) private companies manage employee liquidity.

"It is conducting the sale through the London Stock Exchange's new Private Securities Market, the first major company to test it."

The AI Consciousness Research Arms Race Among Top Labs

Anthropic, Google, and Meta are quietly building internal research teams — hiring neuroscientists and philosophers — to study AI consciousness and moral status. This is largely unreported relative to its potential regulatory, ethical, and liability implications for the industry.

"Anthropic, Google, and Meta are hiring computer scientists, neuroscientists, and philosophers to study whether advanced AI models could become conscious or deserve moral consideration, even as many brain researchers remain skeptical that today's chatbots experience anything at all."