Thursday, July 30, 2026HotTea archive editionVerified 12:06 AM PDT

9 minutes. Facts before narrative.

AI's bill came due in public.

Microsoft showed how cloud demand can carry the spend, Meta showed how fast costs can outrun earnings, central bankers put AI inside the policy problem, OpenAI widened researcher access without opening its systems, defense autonomy moved into a swarm test, and record memory profits still failed to calm the market.

Published daily by 6:45 AM Pacific. No forced optimism. No manufactured panic.

Listen to today’s briefing

AI's bill came due in public.

The sourced HotTea edition, condensed into a chaptered morning podcast with verified audio and a full transcript.

Microsoft showed the revenue. Meta showed the burn.

The two hyperscalers reported on the same afternoon, but their numbers told different stories about how quickly AI infrastructure can turn into durable economics.

What happened

Microsoft reported $90.0 billion in quarterly revenue, up 18% year over year, and $35.8 billion in net income, up 31%. Azure and other cloud-services revenue grew 43%, and Microsoft said annual Azure revenue exceeded $100 billion for the first time. Axios reported that Microsoft's quarterly capital expenditures rose 70% to $41 billion. Meta reported $60.8 billion in revenue, up 28%, while costs and expenses rose 55% to $42.0 billion. Meta's net income fell 14% to $15.8 billion, free cash flow was $784 million, and quarterly capital expenditures including finance-lease principal were $31.1 billion. Meta narrowed its full-year capital-spending range to $130 billion to $145 billion by raising the low end.

Why it matters

The spending race is no longer one undifferentiated AI-capex story. Microsoft paired a large infrastructure bill with accelerating cloud revenue and higher profit; Meta paired strong revenue growth with faster expense growth, lower profit, and little free cash flow. The pressure point is conversion: investors now have public evidence that scale alone does not determine how quickly AI investment reaches revenue, margin, or cash generation.

What to watch

Microsoft's forecast for more than $50 billion of current-quarter capital spending, the share of that spend tied to short-lived GPUs and CPUs, Meta's third-quarter revenue range of $61 billion to $64 billion, Meta's post-layoff cost base, and whether either company discloses product-level returns rather than broad AI attribution.

The caveat

This is not a clean head-to-head comparison. Microsoft and Meta have different businesses, accounting mixes, customer bases, and investment cycles. Meta's quarter also included $2.4 billion in legal charges and $1.18 billion in severance expenses, while Microsoft's GAAP income included investment effects. One quarter cannot establish the lifetime return on either company's infrastructure.

Read this story on its own →

Worth knowing

The rest of the morning

Facts, pressure point, next evidence.

02

OpenAI offered 100,000 researchers the tool, not the system.

Axios reported that OpenAI is launching a program to give 100,000 academic researchers free access to its most advanced hosted models through 2027. The first 10,000 participants are expected to begin receiving access this summer, and OpenAI said the broader effort is part of more than $250 million it plans to invest in external scientific research and discovery through 2027.

Pressure point The program can widen access to expensive frontier inference, but the reported terms do not provide model weights or training data. OpenAI also told Axios that participant data would not be used for training and that business-grade privacy protections would apply; those remain company commitments whose implementation and research value need independent evidence.

Watch Selection criteria, geographic and institutional distribution, published work from the first cohort, whether negative or safety findings can be released without restriction, and whether access produces reproducible results that researchers outside the program can audit.

Axios
Read article →
03

Central banks now have to price an AI boom they cannot yet measure.

A new Bank for International Settlements bulletin said the AI boom is driving a large, increasingly debt-financed investment surge while its productivity payoff remains uncertain and uneven. The authors said AI affects demand and supply at the same time, blurring cyclical signals and increasing the risk of monetary-policy miscalibration. On July 29, the Federal Reserve held its target range at 3.5% to 3.75% by a 9-3 vote; the three dissents favored a quarter-point increase as inflation remained above the 2% goal.

Pressure point AI can create near-term demand through data-center investment and market wealth while producing longer-term supply gains through productivity. Policymakers cannot safely assume the timing or size of either effect. The BIS bulletin is an analytical assessment, not an observed productivity verdict, and the Fed statement did not attribute its decision specifically to AI.

Watch The Fed's year-end task-force findings on AI, productivity, labor, inflation measurement, and data collection; revisions to business-investment data; debt financing for AI infrastructure; labor-market separations in exposed occupations; and evidence that productivity gains are spreading beyond a small group of firms.

Bank for International SettlementsFederal Reserve BoardAxios
Read article →
04

A defense partnership put swarm coordination into the test lane.

Axios reported that AeroVironment and Applied Intuition are collaborating on the Mayhem 10 launched-effect system and Applied's Acuity ISR/Strike software, and that they recently tested the drone's ability to synchronize and swarm. AeroVironment describes Mayhem 10 as an autonomous, multi-role system derived from its Switchblade family, while the reported software layer supports data sharing and coordinated behavior.

Pressure point This is a company-reported test and partnership, not independent evidence of reliable battlefield performance. Public material does not establish failure rates, communications resilience, target-identification accuracy, rules of engagement, or the exact human-authorization boundary. No live operational, casualty, or targeting claim is included here.

Watch Independent test results, the operator-approval model for any strike function, performance under jamming and degraded communications, audit logs for machine recommendations, acquisition contracts, and whether the Pentagon publishes test and evaluation criteria before fielding coordinated systems.

AxiosAeroVironment
Read article →
05

Memory makers printed records. Their stocks still lost the argument.

SK hynix reported record second-quarter results on July 29, and AP reported the following day that Samsung Electronics posted a record 89.5 trillion won operating profit for the April-June period, with nearly all of it coming from semiconductors. Yet AP said SK hynix shares fell more than 9% on Wednesday after its result missed higher market expectations, while Axios documented a broader pullback across semiconductor and memory names.

Pressure point The contradiction is not weak current demand. It is that investors are testing how long elevated memory pricing, AI capital spending, and incumbent market power can survive new capacity and Chinese competition. Record profit is backward-looking evidence; planned fabs, customer concentration, and the cost of the next supply cycle determine whether it persists.

Watch HBM contract pricing, SK hynix and Samsung capital-spending guidance, CXMT capacity and yields, customer concentration among hyperscalers, the pace of new fab construction, and whether rising supply closes the gap before AI-server demand slows.

SK hynix NewsroomAssociated PressAxios
Read article →

The whole AI power map

AI is no longer a tech beat.

HotTea follows where AI moves power, money, labor, security, and state capacity—not only where a new model scores higher.

01

Politics & regulation

Elections, procurement, courts, surveillance, lobbying, and state power.

02

Economics & labor

Productivity, wages, employment, capital spending, concentration, and who captures the gains.

03

War & security

Autonomy, cyber operations, intelligence, targeting, export controls, and escalation risk.

04

AI geopolitics

Chips, energy, alliances, sovereign capability, supply chains, and strategic competition.

05

Markets & companies

Funding, revenue, margins, model economics, enterprise adoption, and infrastructure bets.

06

Science & society

Medicine, education, climate, culture, research, rights, and measurable public outcomes.

Economics

The common thread is conversion: access, compute, and autonomy all need proof beyond scale.

July 30's evidence does not say the AI boom is ending or that it has paid off. It says the next phase will be judged by conversion: capital into cash flow, model access into reproducible science, investment into measurable productivity, and autonomous capability into accountable performance.

1

2

3

The watchlist

Signals that could change the read

How HotTea works

No optimism quota. No negativity quota. Just the honest read.

Every reported item links to its source. Company claims remain company claims. High-risk stories require stronger corroboration. Material caveats, conflicts, and unknowns stay in the story. HotTea’s interpretation is visibly separated so readers can disagree without losing the facts.

Edition validated · 5 stories · 12 unique sources

Audit today’s sources →

Tomorrow’s signal, before tomorrow’s noise

Open HotTea. Know what changed.

A new verified edition every morning. If the evidence or release gate fails, the last verified briefing stays live.

Back to today’s top ↑