Anthropic's Quiet 'wins' in Elon Musk/OpenAI Legal Hullabaloo. ARD #79
Today’s Theme is Anthropic — the AI company making quieter but more consistent progress on the big AI must-dos, while its two main peers — xAI and OpenAI — soak up the bulk of the media attention around peripheral issues.
Three Takes today, all on Anthropic. Then my Overall Take that ties them together. Links below.
(1) Anthropic — The Quiet Winner in the Musk vs. OpenAI Verdict
The big AI legal headline yesterday was the verdict in Elon Musk versus Sam Altman / OpenAI. OpenAI won the case on the contested counts. Musk’s legal effort to claw back control of OpenAI’s for-profit structure — and the broader narrative damages — was largely rejected.
Two pieces frame the moment well.
Axios ran a piece titled “In Musk v Altman trial, the entire AI Industry lost” — a sober reading that argues the trial’s discovery-and-publicity costs hurt the whole industry’s reputation regardless of who prevailed on the legal counts.
The New York Times ran “AI’s Long Hot AI Summer post Elon Court Loss” — which makes the broader claim that the verdict opens up a more aggressive competitive summer across the AI labs.
For full context — I covered the verdict itself in today’s AI-RTZ #1091, “OpenAI Wins Elon Musk Lawsuit” — and on AI Ramblings Daily a few episodes back I framed the underlying competitive dynamic in ARD #75, “Anthropic Tortoise Laps OpenAI Hare.” That Ep 75 framing is doing real work today — Anthropic’s quieter, more consistent execution is what’s actually compounding while OpenAI and xAI are in the courtroom and on the front pages.
MP Take: The Anthropic angle nobody else is naming is that while OpenAI was defending its legal foundations and Musk was running through years of internal discovery, Anthropic spent that same time shipping product, hiring talent, and signing enterprise deals. The verdict is structurally good for OpenAI in narrow legal terms — but Anthropic gets to skip the entire reputation-damage discount that both Musk and OpenAI now carry into the rest of 2026.
(2) Anthropic Inches Further Toward a Broader Mythos Release
Quiet but meaningful product progress this week.
The Wall Street Journal reported that Anthropic is changing its stance on the Mythos cybersecurity process — specifically, allowing Mythos users to share cyber threat data with other users. That’s a small policy shift on paper but a big one structurally — it positions Mythos as a more collaborative, enterprise-grade security tool rather than a sealed-off single-tenant capability.
For longtime listeners — I covered Anthropic’s broader Mythos rollout strategy in AI-RTZ #1051, “Anthropic Claude Peek-a-boo” — which captured the staged-release pattern Anthropic has been running. And in ARD #61, the “Mythos AI Model Grind” episode, I framed the underlying execution discipline that makes Anthropic’s release cadence work.
MP Take: This week’s WSJ piece is a small but legitimate signal that Mythos is moving from sealed-preview into wider commercial availability — less front page drama, but compounds the revenue ramp. OpenAI is trying the same with GPT 5.5/’Spud’, but everyone’s eyes are on Mythos because it takes Claude Code and Claude Cowork beyond 4.7 — the current version. We’ll see how that goes.
(3) Uber-AI coder Andrej Karpathy Joins Anthropic in AI Research
The big AI talent story broke in favor of Anthropic
Andrej Karpathy — co-founder of OpenAI, former Tesla AI Director, creator of the most-watched AI educational content on the internet, including the legendary “Let’s Build GPT From Scratch” lecture series — has joined Anthropic in an AI Research role.
Axios broke the story under the headline “Andrej Karpathy Joins Anthropic.” Karpathy himself confirmed and elaborated in his own words on X. Listen to the full clip and read his X post — the framing he gives is worth reading directly, not paraphrased.
For longtime listeners — Karpathy’s work on “vibe coding” has been a recurring topic on AI-RTZ. I captured it in AI-RTZ #661, “AI Vibe Coding.” And his broader “waiting for AI god / AGI trek” thinking — about the long arc from current models to AGI — I covered in AI-RTZ “Waiting for AI God-Like AGI.” Those two pieces give you the philosophical lens Karpathy brings to whatever AI lab he joins.
MP Take: The structural read is that Karpathy joining Anthropic, specifically NOT joining xAI or rejoining OpenAI, is a major directional signal. Talent flow is the leading indicator in this race. Karpathy is one of the few AI researchers whose presence alone moves a lab’s research velocity meaningfully forward. He’s chosen Anthropic. That is a strong signal.
Overall MP Take
Anthropic continues to be the least heralded of the three mega-AI IPO companies.
While Elon, SpaceX, xAI, and OpenAI get the most media and investor attention — for lawsuits, for IPO valuations, for product launches — Anthropic is steadily doing what needs to be done to move the ball forward. Across most critical areas of execution: AI Research, People, Product, and Markets.
The ramping numbers are what follows. Relatively flawless execution.
Three independent data points just this week make the case:
-
A legal-environment shift that’s structurally favorable to Anthropic without any Anthropic effort.
-
A product-policy move on Mythos that broadens commercial-customer accessibility.
-
A top-tier talent hire — Karpathy — that competitors couldn’t get.
My core issues with Anthropic remain the same:
First — Dario Amodei’s anti-China stance. I wrote about this directly when Anthropic published its 2028 AI Leadership research piece. The research itself is reasonable. But the lobbying posture that Anthropic has been running in Washington — aggressively pushing to curtail China AI advances through US export controls and policy actions — is, in my view, the wrong frame for this moment. China is going to build AI regardless. The constructive question is how the US engages, not how it walls off. (I went deep on this in ARD #77 on US/China — the Pragmatic Steps and Aspirational Steps framing.)
Second — Dario’s strident Job Loss Fear discussion. I captured this in AI-RTZ #789, Jensen vs Dario Debate on AI Jobs — where Jensen Huang publicly pushed back on Dario’s repeated framing of AI as a mass-unemployment threat. Dario’s framing keeps adding to the AI doomer noise that the broader culture is already over-indexed on. Anthropic has the credibility to set a calmer tone here. It hasn’t.
These are real issues. They are not, in my view, blockers on the company’s execution. Anthropic the company keeps doing the right things on Research, People, Product, and Markets. Anthropic the public-facing voice — especially the CEO’s public-facing voice — keeps getting in the way of its own story.
Net-net: Anthropic remains the strongest of the three on the operating layer. Watch the talent flow. Watch the customer ramp. Don’t let the headline coverage of its peers distract from what’s actually happening here.
Gadget AI — Apple’s Product-Design Changes Under New Chief Hardware Head Srouji
Bloomberg reported yesterday that Johny Srouji — Apple’s new Chief Hardware Head — is shaking up oversight of product design across the company. Specifically, the changes consolidate hardware-engineering authority around Srouji’s organization in ways that point to deeper hardware-software integration on AI-relevant product surfaces.
For longtime listeners — I covered the Cook-to-Ternus succession context and the broader Apple hardware-leadership map in AI-RTZ #1063, “The Long-Expected Apple Cook-to-Ternus Hand-Off.” Srouji’s expanded scope under that map is the operational shoe dropping.
MP Take: This is the beginning of many detailed changes to come as Apple zones in on AI opportunities around its unique global hardware/software/services ecosystem. The changes today start with product design — and they point to close integration of hardware and software. Especially with open-source small language models (SLMs) running on Apple Silicon at the edge. That’s the technical wedge Apple has been quietly preparing for years. Apple remains uniquely positioned with its own Apple Silicon platform and its unparalleled global supply chain for hardware built under the 15-year Tim Cook CEO run. No other AI player has both pieces. I walked through the broader open-source-AI thesis for Apple — and specifically Apple paired with Nvidia as the two-company answer to the US open-source AI question through 2030 — in AI-RTZ #1089, How Nvidia & Apple Can Be the Global, US Open Source AI Champions vs China. Today’s Srouji reorg is directly downstream of that thesis. Worth reading in full if you haven’t.
Questions
Q1 — What is the core MP Apple product beyond the iPhone?
Answer: The AirPods Pro — for their deep software and hardware versatility.
I use them daily across phone calls, podcasts, music, transit, transcription pickup, and increasingly as a voice-interface conduit for AI assistants. They’re an under-appreciated AI form-factor sitting in roughly a billion-plus ears globally. The deep integration between AirPods Pro hardware (active noise canceling, transparency mode, spatial audio) and Apple-ecosystem software is the kind of vertically-integrated product experience Apple does better than anyone.
Q2 — What is MP’s key daily gripe with an Apple product?
Answer: The Apple Watch constantly asking for the security PIN daily.
It’s a real friction-pain on an otherwise great product. Every time you take it off — for showering, for bed, for an exercise change — you have to re-enter the PIN on a tiny screen with a tiny keypad. Apple has solved this elegantly elsewhere — Face ID on iPhone, Touch ID + watch-unlock on Mac, AirPods auto-pairing. The Apple Watch security UX feels like the last unsolved daily-friction point in the Apple ecosystem for me.
Clips
Why Apple + Nvidia Lead the US Open Source AI Race
A short distillation of MP’s broader Apple+Nvidia thesis — both companies control hardware silicon, and everything else (open source software in particular) can be built on top. Apple Silicon + 15-year Cook supply chain run = uniquely positioned.
MP Take: Nvidia + Apple is the two-company answer to the US open-source AI question through 2030. Today’s Srouji reorg is directly downstream of that thesis.
https://youtube.com/shorts/TT8M7CJc9O8
Apple’s Srouji Reorg — Bloomberg Breaks the Hardware Shakeup
A short distillation of the Gadget AI news lead-in — Bloomberg breaks Apple’s hardware reorg under Johny Srouji, formerly head of Apple Silicon, now redesigning Apple design across all products. Personnel changes, integrates Apple seriousness about the Apple+Nvidia hardware-software AI thesis.
MP Take: Field post this Sunday — Apple along with Nvidia is very well positioned. The Srouji reorg confirms the thesis at Apple’s operational layer.
https://youtube.com/shorts/dcRSoSOR9SU
Apple Watch Security PIN — MP’s Daily Gripe
A short distillation of the Q2 daily gripe — Apple Watch security PIN re-entry friction. The watch has umpteen sensors that sense it’s been on the wrist all day, but still asks for the four-digit PIN every time you raise your hand to look at the time.
MP Take: There should be a change here. Will bring it to John Turnis’ attention and Srouji. Just change that piece because the watch is the core daily-use product.
https://youtube.com/shorts/EKmUxJ6pCTw
Anthropic Mythos — Quiet Cyber Threat-Sharing Release
A short distillation of Take 2 — Anthropic’s quiet cybersecurity progress with Mythos. Anthropic has been just doing things steadily. The next-generation Mythos language model is anticipated by governmental and corporate levels because of its potential threat on the cybersecurity side. Anthropic has been careful working with all the big entities to make sure Mythos is within safety guidelines.
MP Take: They made a stance change that allowed cybersecurity reviews a little more leeway in accelerating ramp. OpenAI trying the same with GPT 5.5 — but everyone’s eyes are on Mythos.
https://youtube.com/shorts/SudC__uxx6g
Today’s AI-RTZ — #1091 — is “OpenAI Wins Elon Musk Lawsuit” — which gives you the full verdict-and-aftermath read in writing if you want it.
Tomorrow — ARD 80 and AI-RTZ 1092.
Thanks for tuning in, AI Curious Folk. Stay tuned.
(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)
Links
Take 1 — Anthropic the Quiet Winner in the Musk vs. OpenAI Verdict:
-
Axios — “In Musk v Altman Trial, the Entire AI Industry Lost” — https://www.axios.com/2026/05/19/musk-altman-openai-trial
-
NY Times — “AI’s Long Hot AI Summer Post Elon Court Loss” — https://www.nytimes.com/2026/05/18/technology/elon-musk-openai-trial.html
-
AI-RTZ #1091 — OpenAI Wins Elon Musk Lawsuit (today’s companion) —
-
ARD #75 — Anthropic Tortoise Laps OpenAI Hare (backcat) —
Take 2 — Anthropic Mythos Cybersecurity Threat-Sharing Release:
-
WSJ — Anthropic Lets Mythos Users Share Cyber Threats With Others — https://www.wsj.com/tech/ai/anthropic-lets-mythos-users-share-cyber-threats-with-others-26d17bc6
-
AI-RTZ #1051 — Anthropic Claude “Peek-a-boo” (backcat) —
-
ARD #61 — Mythos AI Model Grind (backcat) —
Take 3 — Andrej Karpathy Joins Anthropic in AI Research:
-
Axios — Andrej Karpathy Joins Anthropic — https://www.axios.com/2026/05/19/anthropic-openai-karpathy-andrej-claude
-
Karpathy’s own announcement (X) —
-
AI-RTZ #661 — AI Vibe Coding (backcat) —
-
AI-RTZ — Waiting for AI God-Like AGI Trek (backcat) —
Overall MP Take — Embedded Reference Links:
-
Anthropic Research — 2028 AI Leadership (Dario anti-China posture context) — https://www.anthropic.com/research/2028-ai-leadership
-
AI-RTZ #789 — Jensen vs Dario Debate on AI Jobs (Job Loss Fear context) —
Gadget AI — Apple Product-Design Changes Under Srouji:
-
Bloomberg — Apple’s New Hardware Chief Shakes Up Oversight of Product Design — https://www.bloomberg.com/news/articles/2026-05-19/apple-s-new-hardware-chief-shakes-up-oversight-of-product-design
-
AI-RTZ #1063 — Long-Expected Apple Cook to Ternus (backcat) —
-
AI-RTZ #1089 — How Nvidia & Apple Can Be the Global, US Open Source AI Champions vs China (Sunday Bigger Picture) —
Companion text:
-
Subscribe for daily ARD episodes + AI-RTZ posts — michaelparekh.substack.com
-
Follow on X — @MParekh
-
Follow on TikTok — @mparekh1
-
ARD on YouTube — @MParekh
