Bloomberg Tech
2026-07-02 · Hosted by Caroline Hyde, Ed Ludlow · Bloomberg / iHeartMedia
Executive Summary
Bloomberg’s scoop that Meta is developing a two-part cloud business — selling API access to AI models on its infrastructure and renting out raw AI compute — drove the stock to its biggest jump in a year (up ~11%) while sending neo-cloud CoreWeave down ~12-15% as investors weighed a powerful new compute competitor. In a separate policy shift, the Trump administration lifted foreign access restrictions on Anthropic’s Fable 5 model after the company agreed to stronger safeguards, ending a weeks-long standoff. The episode also covered Lime’s NASDAQ debut (raised $174M, ~6x oversubscribed), a wave of capital flooding AI infrastructure (MGX’s ~$50B fund, Starwood’s $10.2B), and fresh data showing AI is beginning to drag down hiring in tech and financials.
Key Stories & Changes
1. Meta Launches “Meta Compute” Cloud Business
Bloomberg reports Meta is developing plans for a cloud business in two parts: (1) selling API access to AI models on Meta’s infrastructure (akin to AWS Bedrock) and (2) renting out raw AI computing power from its data center footprint
Effort is called “Meta Compute,” led by Meta’s head of infrastructure, Daniel Gross on the superintelligence team, and Meta president Dina Palma
Meta stock up ~11%, its best day in a year; CoreWeave down more than Meta was up (down ~12-15%, worst day since February)
Amazon and Microsoft traded higher, suggesting entrenched hyperscalers not seen as threatened
CEO Mark Zuckerberg telegraphed this on prior earnings calls — as far back as January saying “if we don’t use it, somebody else will and we’ll find a way to resell it”
BI analyst Mandeep Singh: signals Meta feels pressure to show ROI on capex; calls Meta “late to the inferencing party”; warns compute rental is crowded and low-margin for a company with ~50% operating margins
2. Trump Administration Lifts Restrictions on Anthropic’s Fable 5
The Trump administration lifted foreign access restrictions on Anthropic’s Fable 5 model after Anthropic addressed government safety concerns
Anthropic to begin re-enabling access as soon as today; Fable is similar to its powerful “mythos” model but with additional safeguards against jailbreaks (e.g., hacking, bank access)
Anthropic worked with government on stronger safeguards and is building an industry framework (with Amazon and others) to address AI security holes systematically
Anthropic also released an in-house pre-clinical drug research model, entering pharma discovery
Analyst Ali Mehlum (author of Code War): a “big win” for Anthropic and the U.S.; warns organizations are now diversifying models to avoid a repeat; balance of controls needed rather than outright blocking
3. Lime IPO Debuts on NASDAQ
Lime raised $174M, priced at the midpoint ($25), shares indicated to open ~$27; reportedly ~6x oversubscribed
CEO Wayne Ting: operates in 230 cities, 29 countries, 5 continents; free cash flow positive with 50%+ cash margins per trip and <1-year payback on bikes/scooters
Biggest growth is deepening existing markets (London grew faster than the company overall); M&A now possible with public equity
4. Capital Floods AI Infrastructure & Other Headlines
Abu Dhabi’s MGX secured nearly $50B for one of the largest dedicated tech funds ever; Starwood Capital raised $10.2B for data center property
ByteDance to build its largest data center outside China in Brazil — a $39B project, first facility by late 2027
Marlin Spike Partners closed an oversubscribed ~$127-170M defense-tech fund (Fund II) to “Rearm America”
U.S. declined to renew the USMCA trade pact, opting for annual reviews instead (pact remains in force ~10 more years)
Supreme Court agreed to hear Apple’s appeal of a contempt ruling in the Epic Games app-store payments case (October)
Kalshi became a World Cup co-sponsor alongside ADI at a steep discount (~$20M vs. ADI’s ~$150M)
5. AI’s Imprint on the Jobs Market
U.S. added 98,000 private sector jobs in June (ADP), capping the best three-month hiring run in over a year
Bloomberg economics reporter Matthew Boesler: AI appears to drag down two sectors — financials and tech; finance is 3x the size of the information sector in employment, with many routine-processing roles vulnerable
Trends Identified
1. The AI Trade Broadens Beyond Hyperscalers to a Global Ecosystem
Prince of Wealth’s See Michelle framed a “new exceptionalism” still centered on the U.S. but increasingly global, with South Korea and Taiwan intrinsically tied to U.S. AI via the memory and chip trade. Meta’s move and the record quarter for the SOX index reinforce that AI capex is now embedded in nearly every investment conversation — across equities and fixed income — rather than being confined to a single tech sector.
2. Return on Investment Is Becoming the AI Trade’s Central Question
Multiple guests converged on ROI as the key variable that could either sustain or topple the tech trade. Meta’s pivot to monetizing excess compute is read as a direct response to persistent investor questions about how the hyperscalers will earn returns on enormous capex — a reassuring signal to some, but to skeptics a sign that Meta’s own AI products (agents, standalone apps) haven’t scaled.
3. AI’s Labor Impact Is Now Visible and Sector-Specific
The data increasingly shows a split between the broad economy and two sectors — tech and finance — where AI is beginning to suppress hiring. Because finance employs three times as many people as the information sector, with large pools of routine clerical and customer-service roles, it may bear the larger AI-driven workforce imprint over the coming decade.
4. Frontier AI Regulation Is Moving From Ad Hoc to Systematic
The Anthropic-Fable 5 episode is being treated as a case study in how different arms of the U.S. government exercise oversight of frontier models. The resolution — with agreed safeguards and an industry framework — hints at a more predictable regime, though the initial restriction damaged trust and pushed firms toward model diversification. —-
Sentiment Analysis
Overall Market Sentiment: Cautiously Optimistic
The dominant mood is constructive on AI’s long-term trajectory and enormous ongoing spend, tempered by real-time repricing of the “picks and shovels” names and open questions about ROI and regulation.
Risk Factors Highlighted
ROI shortfall on AI capex: If enormous hyperscaler spending fails to generate returns, it is the one thing guests keep citing that could topple the tech trade.
Compute rental commoditization: A crowded, low-margin business that could dilute Meta’s ~50% operating margins over time.
AI-driven job losses in finance and tech: A sector three times larger than tech faces automation of routine roles.
Frontier model security holes: Jailbreaks and vulnerabilities in powerful models (cyber, bio) cannot be fully eliminated, only limited.
Regulatory unpredictability: Ad hoc, arbitrary export/access restrictions damage industry trust and push model diversification.
Chip stock volatility: The SOX index and semi names swung sharply as investors reassessed the AI build-out.
Trade policy uncertainty: USMCA moving to annual reviews introduces recurring negotiation risk over continent-wide supply chains.
Apple platform monetization risk: Supreme Court review, EU DMA, and DOJ actions all threaten how Apple earns on its app store.
This episode was covered in today’s The Market Signal — 2026-07-02, a cross-source synthesis of multiple podcast reports.