| StatusGoStrategy in Motion |
AI in Healthcare · A Weekly Point of View |
|
| | The finding: AI’s billions, healthcare’s basis points. Across the ten largest AI and healthcare names reporting Q3 2026, the split shows up in the margins, not the mission. SpaceX posted a 30% net margin this quarter and Meta 15%, while most healthcare giants operate on low single digits, and one-time charges further dented GAAP profit at Johnson & Johnson, Abbott, Intuitive, and Molina. The companies with the fattest margins are financing the AI buildout, while the ones expected to deploy it in care must do so on far thinner cushions. |
| Company (mkt cap) | Type | Q3 Revenue | Net income | Net margin | | Artificial Intelligence | SpaceX
$2.10T | Aerospace & AI | $14.58B (+161%) | $4.40B | 30.2% | Meta
$1.80T | Big Tech | $64.6B (+26%) | $9.86B | 15.3% | IBM
$310B | Enterprise AI | $17.1B (+5%) | $2.19B | 12.8% | Tesla
$1.53T | EV & autonomy | $28.05B (flat) | $787M | 2.8% | | Healthcare | Intuitive Surgical
$151B | Surgical robotics | $3.05B (+22%) | $523M | 17.2% | Abbott
$174B | Medtech & dx | $13.0B (+7% org.) | $920M | 7.1% | UnitedHealth
$375B | Payer | $112.7B (-0.4%) | $4.11B | 3.6% | Elevance Health
$93B | Payer | $50.4B (+0.7%) | $1.37B | 2.7% | Johnson & Johnson
$625B | Pharma & medtech | $25.4B (+6%) | $210M | 0.8% | Molina
$11B | Payer (Medicaid) | $11.1B (-3.4%) | $38M | 0.3% |
Latest reported quarter (Q3 2026). Net margin = GAAP net income ÷ revenue (StatusGo calculation), shaded green (highest) to red (lowest). One-time items reduced GAAP profit at five names; adjusted net income was roughly $19B at Meta, $6.3B at Johnson & Johnson, $2.48B at Abbott, $1.05B at Intuitive, and $46M at Molina. Market caps approximate, as of the Oct 23 close. Source: company filings.
|
|
|
| 🎁 |
Giving It All Away Before the End of the World |
|
| | If AI creates enormous wealth, who shares in it, and at what risk? In the same stretch, Anthropic committed a fifth of its equity to a public benefit fund and warned in its IPO filing that its own technology could threaten humanity, while Elon Musk pledged a SpaceX share to every American child. Generosity and dread, side by side, even as fresh data suggests the broader payoff has yet to arrive. |
|
|
Anthropic pledges equity to a public benefit fund, just before its IPO AI Governance |
Ahead of its planned November IPO, Anthropic will commit equity to a new Anthropic Public Benefit Fund, starting at 5% of shares (pending stockholder approval) and rising 2% for every 50% gain in value above a baseline. With founder and employee pledges, roughly 20% of shares go to the public good, committed before public shareholders can buy in. Anthropic
|
| |
|
Anthropic’s IPO filing warns AI could pose existential risk to humanity AI Risk |
Anthropic’s IPO filing, made public Monday, warns in its risk factors that its technology could pose ‘catastrophic or existential risks to humanity,’ a concern it says exists within the company. Increasingly autonomous systems, it cautions, could take harmful actions independently and at scale, even as its ability to measure whether its models stay aligned may decline. The same week, it pledged equity to the public good. Fortune
|
| |
|
Musk pledges a SpaceX share to every American child AI Economics |
Elon Musk pledged on X to give one of his SpaceX shares to every child in America in 2027, a gift worth about $10.5 billion (roughly 73 million kids) at SpaceX’s $146 Friday close. He framed it as a direct answer to Anthropic’s public benefit fund a day earlier, part of what he calls building ‘universal high income’ through AI. CNBC
|
| |
|
Apollo: no sign of an AI payoff outside tech Markets |
Apollo chief economist Torsten Slok argues AI has delivered ‘zero financial payoff to corporate America outside of tech.’ Revenue per worker and profit margins are climbing for the Magnificent 7 but flat for the other S&P 493, and productivity data shows no lift yet. Outside tech, firms increasingly treat token spending as a margin cost. Apollo
|
| 🩺 |
AI in Healthcare, and a $5.7B Acquisition |
|
|
Microsoft’s diagnostic AI beats doctors in a Science study Clinical AI |
Microsoft’s AI Diagnostic Orchestrator (MAI-DxO), which debates a case like a panel of physicians, solved 83% of complex NEJM cases versus 36% for physicians using references, per a study in Science. Two caveats worth naming: the cases were historical and the doctors worked solo, without specialists; and most of the lift came from the model, not the harness (GPT-5 alone scored 78.9%; MAI-DxO added about four points). Beth Israel Deaconess trials are next. Science
|
| |
|
AI tool flags skin-cancer risk and shows its work Clinical AI |
Mayo Clinic researchers reported that DEXI, an explainable-AI dermoscopy tool trained on 66,000+ annotated images, scores skin lesions 0 to 10 for cancer risk and names the features driving each score. In a 261-lesion retrospective study, a 3.0 threshold correctly flagged 93.4% of cancers and 75% of benign lesions. Authors stress it supports, not replaces, clinical judgment. Mayo Clinic
|
| |
|
Bon Secours Mercy Health to buy Surgery Partners for $5.7B Healthcare M&A |
HealthSpan Partners, an affiliate of Bon Secours Mercy Health, will acquire Surgery Partners for about $5.7 billion including debt, or $23.50 a share in cash, taking the operator private. The deal adds more than 200 outpatient surgical sites across 30 states, a sign of how aggressively systems are buying into the shift of procedures out of the hospital. Expected to close in early 2027. Becker’s ASC Review
|
| 🕵️ |
AI Agent 007, Reporting for Duty |
|
| | The model race is quietly becoming an agent race. This week brought a proactive agent that messages you first, a cheap decision model built to replace LLMs at scale, and an open-weight challenger nearing the frontier. |
|
|
Anthropic’s Claude Duet makes the agent proactive AI Agents |
Anthropic opened a research preview of Claude Duet, a proactive agent on Claude Opus 5.5 that sends a morning brief, monitors your connected apps, and pings you only when something needs you. It runs on its own cloud machine and requires your approval before sending, paying, signing, or deleting; trading and moving money stay off-limits. US Pro and Max only. Anthropic
|
| |
|
OpenAI ships a decision model to take on Jev AI Infrastructure |
OpenAI made its Decisions API generally available, entering the fast-growing market for ‘decision models’: fast, cheap classifiers that pick among options with a confidence score and can replace LLMs in high-volume workflows. Built by reconfiguring ChatGPT Luna, it cuts responses from 1-3 seconds to under 200 milliseconds. On public benchmarks OpenAI ranks behind the leaders, Liquid’s D1 Ultra and Base Lab’s Elephant-1, as well as category creator Jev. TechCrunch
|
| |
|
Moonshot’s Kimi K3.1 Pro pushes open weights to the frontier Model Race |
Chinese lab Moonshot AI released Kimi K3.1 Pro, a 3-trillion-parameter open-weight model (896 experts, 16 active per query), with weights due Nov 4. On the lab’s own benchmarks it rivals frontier proprietary models, leading OfficeQA Pro v2 financial analysis and nearing Claude Opus 5.5 on coding and 3D, while trailing on cybersecurity by design. SiliconANGLE
|
|
