Six Steps for next US/China White House meeting Sep 24/26. ARD #77
Today’s frame: what to expect after the first US/China meeting of 2026.
Today, like yesterday, I’d like to focus on this one topic around an issue of critical importance to how this AI Tech Wave shapes up in its promise to the US. And the world.
With several points around my take on the issue — split into two buckets, Pragmatic Steps the political and tech ecosystem could realistically take in the near term, and Aspirational Steps that would make way for the real Win-Win-Win scenario longer term.
The first Xi/Trump meeting of this AI Tech Wave era ended this week with what looks like shallow wins for both sides. Trump on Air Force One to China, over 17 American CEOs in tow including Apple’s Tim Cook and Nvidia’s Jensen Huang (picked up by Air Force One on the way over in Alaska — “Home Alone briefly” as I put it on-air). A lot of drama, a lot of theater. Neither side got what they wanted in entirety, which is always the case with these things. Boeing did sell some jets, 200; the US said the expectation was 150; the actual expectation was 300. Bid-ask is wide.
On the things this show cares about — AI and technology — Nvidia’s Jensen got the US position: “hey China, you can have as many of the nerfed H-200 chips from Nvidia as you want for your AI endeavors.” But China didn’t jump in. China wants to encourage its own companies (Huawei in particular) to create chips that are further back in capability but catching up. Cold War framing, geopolitical tussles, binary us-vs-them — it suits the political realities of both sides. Bipartisan US anti-China posture on both sides of the aisle. And Xi gets political benefit locally because of his total control of Chinese media — the population of 1.4 billion really thinks the US are the bad guys, that we’ve done bad by them.
Xi at the White House September 24 this year — that’s the next set-piece, if all goes on schedule. Today I’d like to think through what could come next — what’s pragmatic in the current political environment, and what would be genuinely transformative if both sides could clear a higher bar.
(1) The First Xi/Trump Meeting Ends with Shallow Wins for Both Sides. The reporting: NYTimes — Scarce Details on the Trip. NYTimes — Nvidia’s Future Post Meeting. WSJ — “The Hard Part” Ahead: Tightly Choreographed Visit Masks Big Differences Between U.S. and China. Standing AI-RTZ context: AI-RTZ #1085 — No Nvidia with Jensen in China. And two earlier framings: AI-RTZ — US/China “Threading the Needle” and AI-RTZ #1075 — China’s Post-Meta/Manus Red Chip.
MP Take: Let me discuss My Take in two parts. First on the likely next steps on current political realities. Then on the key longer-term changes that would make way for the REAL “Win-Win-Win” scenario for the US, China, and the World around AI and technology. It’s more aspirational than pragmatic in the current environment. But nevertheless worth thinking through.
First — the Pragmatic Steps:
(P-1) The Tech companies (hardware and software) manage to join forces somewhat to encourage Washington and both parties to keep tech trade with China open. Even if it’s narrower than in the past. Even if it’s narrower than the current paths. This entails encouraging Chinese companies to invest in the US in Joint Ventures with US companies — not just in AI Infrastructure, but in Autos, Materials, and a range of non-tech industries like Solar, Energy, and Logistics. Specific near-term lift: taking a couple of steps back on some of the curtailments by the FCC, Brendan Carr in particular, on limits on drone imports from China or cellular modem imports from China. This is in the weeds, but this is core infrastructure that the US does not have the capability to make for the most part, and there aren’t non-China sources for it. We are literally cutting off our nose to spite our face by going after this for very, very shallow political reasons. Curtailing some of that would go a long way ahead of the next Xi-Trump meeting.
(P-2) Have a defined set of agreements for this Fall’s trip by President Xi to the US. Nvidia and Intel in particular could work on AI Infrastructure partnerships with SMIC, China’s version of TSMC. This way SMIC could have an opportunity similar to what TSMC has in the US. And it could even be structured as a JV between Intel and SMIC, with some help from its unusual shareholder, the US Government. Remember, Intel was basically left for dead a few months ago. Then post the US investment in the company + Elon Musk partnering with Intel + Apple now considering partnering with Intel — the stock is up over 200% because they’re committing to build their own fabs, their own chips against TSMC. A JV with SMIC could go a long way. Small item, but a big idea that is not being talked about.
(P-3) Big companies, Tech and otherwise — Ford, GM, Google, Meta, OpenAI, Apple and others with aspirations around AI devices — work on similar US-JV partnerships with Chinese AI hardware leaders like BYD (and Geely too on EVs), CATL (Contemporary Amperex Technology Co Limited — one of the world’s biggest battery providers), Unitree (which really makes robots at scale), DJI (drones + cameras, currently US-import-blocked) and others. A lot of our industries don’t have drones for regular things across so many vertical industries — from healthcare to education to building. Lobby for these partnerships ahead of the next series of US/China meetings. Small steps forward would go a long way.
Second — the Aspirational Steps:
(A-1) Members of Congress in the House and Senate take a deep breath, and hold off on the dystopian geopolitical pronouncements between the US and China. See if Congress can resume bipartisan trips to mainland China (not Taiwan as Nancy Pelosi did, but mainland China). They used to do this two or three times a year. It’s been years. Note that US business leaders + VCs + the CEO of Ford (look him up on YouTube) have already done this. They came back amazed at how far China is ahead on a lot of things. The story has flipped. China is no longer COPYING US tech in many domains. And is innovating way beyond it in several — BYD on EVs, CATL on batteries, Unitree on humanoid robotics, DJI on drones, and more. The frontier-model AI and most-advanced-process semiconductor gaps still favor the US, but the categorical “China is behind” frame from a few years ago is no longer empirically accurate. They can actually help us create a larger pie across things like EVs, drones, defense. China has built the manufacturing ecosystems for AI-related hardware and software at scale, with the necessary Power investments to make it all happen with GOBS of AI Compute. Even with inferior chips. Yes — they’re still behind on chips because of no access to ASML’s half-billion-dollar lithography machines. But on a lot of fronts they’re ahead.
(A-2) Both countries need to figure out a way to communicate RESPECT for each other on a bottom-up basis. R-E-S-P-E-C-T as Aretha used to sing (the Blues Brothers version, of course).
The media on both sides today portrays a zero-sum, geopolitical cold war state. One concrete mechanism on the US side: mainstream media outlets carry more first-person reporting from mainland China that’s not filtered through a Beijing-as-adversary frame. Mirror that on the China side. In reality, both populations have the same long-term goals — peace and prosperity for its populations longer term. Especially for the younger generations. Right now we think Chinese are the bad guys, Chinese think we’re the bad guys. Needs to be addressed with some proctive respect on US side.
(A-3) On the critical US/China issues on Taiwan, the US needs to figure out ways to foster more communication between China and Taiwan. Currently we just think of it as binary — China takes over Taiwan militarily, or not. The reality is the Taiwanese have 23 million people and over half a century of existing in a separate place governed separately from China. They have two parties that are themselves divided on the issue — the DPP (Democratic Progressive Party, currently in power, more anti-China) and the KMT (Kuomintang, more pro-economic-cooperation and peaceful engagement with mainland China). These are things the Chinese people across the strait — they were one at one time, they’re not currently, but they have been thinking about this without our help. We need to encourage that. Right now most of the discussion is just on war-or-no-war. Cooling it down a notch would be a step in the right direction.
Closing the MP Take — the missing ingredient
There are more possible steps in each bucket, but I’ll pause here for brevity.
In reality, nobody really wants War on any side, Cold or Hot. Especially after China has expanded more borrowed dollars to build up its mainland infrastructure, industries and economy, with the Herculean efforts with its generational population before they got any older.
But an easing of the broader US/China geopolitical tussles and kerfuffles needs a critical, missing ingredient.
As mentioned above, there is a dearth of exhibited RESPECT, Aretha style, for each side. From the highest levels on down. Not just heads of state, but in the political apparatus and mainstream media on both sides.
Respect is a fundamental ingredient in not just human relationships. Watch the opening sequence in 2001 A Space Odyssey. The monolith moment is about awe-before-conflict — that’s the order we’ve reversed in US-China.
Some respect could go a long way to some progress above the “pragmatic” steps above. It’s not a quixotic vision. Look at East and West Germany together now, despite ongoing issues that are being worked through still. Both sides are still trying to figure out how to respect the other, even as one country.
Catholics and Prostants the world over had a similar issue for millennia. As do other religions we all know around the world today.
And to be clear — the AI safety and AI values conversations stay on their own bilateral track. Different governance models, different alignment norms, different definitions of acceptable use. Those merit serious bilateral engagement on their own merits, not blocked by this trade reset, and not subsumed into it either. A parallel conversation, not a conflicting one.
Cooperation around Technology and AI supply chains between the East and West could still be a practical way to make these aspirations real. Prosperity-building steps that expand the global economic pie brought the world this far since WWII. With China bringing more people out of poverty than any other time in history.
And the US has a world-leading $120+ trillion total GDP, services-driven GDP of over $30 trillion, and over $70k+/capita — that’s an exceptional global achievement by any measure. It’s a quarter of the global GDP with less than 5% of the global 8 billion population. China is $18 trillion GDP and could be a larger US market than currently deemed pragmatic — especially for US Services. Think of it as Globalization 2.0.
AI especially can make the pie even bigger for both the US and China if some basic current frameworks are revised. And not viewed as a binary, zero-sum game.
NEXT: Gadget AI — The Real Interesting Meta Innovation Around AI Display Smart Glasses: The Wristband Controller. Meta announced an updated version of their wristband controller for their AI Display smart glasses — and while everyone else focuses on the smart glasses themselves, the wristband controller is the underrated piece: a neural-handwriting input device controlled via finger movements, no visible touch surfaces. From a Meta acquisition timely-one a couple of years ago. They’ve really innovated in a cool way. I think this is as interesting as the touch surfaces around screens. Meta has an opportunity here that I’ve written a lot about beyond just smart glasses: if they were to create this user interface as a platform for a whole host of AI wearables, I think even Google or Apple could be enticed to incorporate this way to interface with gadgets using a wrist controller. It’s a very low-tech but key tech that reasons the movements in our arms to control devices. Reporting: The Verge — Meta Ray-Ban Display Virtual Neural Handwriting Apps Developer. Standing backcat: AI-RTZ #849 — Meta Leans In on AI Smart Glasses.
MP Take: The wristband controller is a key Meta innovation via an acquisition, that could be the basis of an AI UI/UX controller platform INDEPENDENT of its smart glasses. Could give Meta a critical inroad into Apple iPhones and Google Android Smartphone control surfaces. If they execute it in that way. Preferably as a spun-off company that could be majority owned by Meta. But it’s not a likely outcome under the current blinders-on, big tech environment of “winner take all” silo mentality around AI and tech innovations. But it’s a key device to watch nevertheless.
Bonus — today’s AI-RTZ companion #1087 — The Lure of AI Companions — the morning written read on the consumer-AI mainstream-traction category I’ve been flagging in recent ARD episodes.
Closing Questions —
-
Why am I so excited about the Meta wrist controller? Personal story: I’ve been virtually touch-typing ever since I learned to do QWERTY typing in school, way back in my teens. For decades, through all the tech waves, I still do this. My wife still says, “hey, why are you moving your fingers around when I’m listening to something, a podcast or watching a movie?” That’s because I’m still typing with my fingers. So this business of controlling things with my fingers has been second nature for me for a long time. Apple Watch + Vision Pro had elements of this. The Meta display smart glasses with the wristband makes it a closer possibility.
MP Take: I’d love to see this Meta-acquired wristband controller technology used on other gadgets, AI or otherwise. The finger-movement interface is one of the most underrated AI UI/UX moves of 2026. Watch this space.
-
Question 2: What other control mechanism would I like to see in AI Wearables? Bendable smartphone-sized wearables on wrists. Much bigger than an Apple Watch — even the Apple Watch Ultra. I’ve experimented for years with bands and other mechanisms — there are third-party wrist brands you can get running on Amazon with MagSafe-magnetic connectors. With today’s foldable displays maturing, this notion of having bigger displays on your wrist is not a weird one. Remember — for thousands of years humans went sword-fighting with metal wrist straps to defend themselves. Watch Gladiator again. In today’s world where we spend seven to eight hours a day on the displays in our hands, the notion of having a bigger foldable display on our wrist is a cool one. AI Agents can have a tracking surface bigger than an Apple Watch but smaller than a phone-in-hand. Future ARD deep-dive topic.
MP Take: A topic worth discussing in more detail in a future podcast. The wrist as an AI tracking surface — Gladiator-era pattern updated with foldable display tech.
(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)
Segments + Clips from today’s episode
Today the ARD is structured with 2 deeper-dive Segments on US-China plus 4 Shorts on the surrounding context. (Magic Segment #2 on Meta’s Wristband — excluded today; only the 2 US-China Segments uploaded.)
Segments (~5 min each, 16:9 horizontal)
Segment — Revisiting China — Respect Aretha-Style, A Path to US-China Cooperation The Aspirational Steps bucket from today’s deep-dive — Congress resumes bipartisan trips to mainland China · RESPECT communicated bottom-up Aretha-style with concrete US-side mechanism (more first-person reporting from China not filtered through Beijing-as-adversary frame) · Taiwan diplomacy at least as intense as military posture (DPP vs KMT party dynamics).
MP Take: Cooperation around Technology + AI supply chains between East + West could still be a practical way to make these aspirations real. US $120T+ GDP / $30T services / $70k+/capita = a quarter of global GDP with less than 5% of 8 billion population. China $18T GDP could be a larger US market than currently deemed pragmatic, especially for US Services. Globalization 2.0. AI especially can make the pie even bigger for both the US and China if frameworks are revised. Not binary, not zero-sum. The missing ingredient is RESPECT — watch the opening of 2001: A Space Odyssey, where the monolith moment is about awe-before-conflict, the order we’ve reversed in US-China.
Segment — US-China Pragmatic Steps — Intel/SMIC JV, BYD/CATL/Unitree/DJI Partnerships The Pragmatic Steps bucket. (P-1) Tech companies join forces to keep US-China tech trade open + roll back FCC Brendan Carr drone/cellular-modem import curtailments. (P-2) Nvidia + Intel AI Infrastructure JV with SMIC — possibly structured as Intel-SMIC JV with USG-as-Intel-shareholder leverage. Intel up 200% post-USG investment + Elon partnership + Apple partnership consideration. (P-3) Big US co’s — Ford, GM, Google, Meta, OpenAI, Apple — work on US-JV partnerships with BYD + Geely (EVs), CATL (batteries), Unitree (robots), DJI (drones).
MP Take: These tactical moves go a long way to making the next US/China meeting in DC much less shallow. Even small steps forward count. We’ve been cutting off our nose to spite our face for shallow political reasons on the FCC drone + modem import curtailments alone. SMIC-Intel JVs are not being talked about and could be transformative. The picture is materially different from a few years ago — China has the manufacturing ecosystems + Power infrastructure + GOBS of AI Compute (even with inferior chips). US business leaders + VCs who’ve been to mainland China have already realized this.
Shorts (~1 min each, 9:16 vertical)
Short — Meta Wristband Controller — AI UI/UX Across Devices The real interesting Meta innovation in the new Ray-Ban Display smart glasses launch is the wristband controller — a neural-handwriting input device controlled via finger movements, no visible touch surfaces. From a 2-year-old Meta acquisition. Could become an AI UI/UX controller platform INDEPENDENT of the glasses.
MP Take: Could give Meta a critical inroad into Apple iPhones and Google Android Smartphone control surfaces. Preferably as a spun-off company majority-owned by Meta. But not likely under current blinders-on, big tech “winner take all” silo mentality. Key device to watch.
Short — Why MP Is Excited About the Meta Wrist Controller Personal story: I’ve been virtually touch-typing since learning QWERTY in my teens. Decades of finger-control instinct — my wife still asks why I’m moving my fingers around while listening to a podcast. Apple Watch + Vision Pro got close. Meta wristband + display glasses make it closer to mainstream.
MP Take: I’d love to see this Meta-acquired wristband controller technology used on other gadgets, AI or otherwise. The finger-movement interface is one of the most underrated AI UI/UX moves of 2026.
Short — Bendable Smartphone-Sized Wrist Wearables What other AI wearable control would MP like? Bendable smartphone-sized wearables on the wrist — bigger than Apple Watch Ultra. Years of experimentation with bands + MagSafe-magnetic third-party brands. With foldable displays maturing, bigger wrist displays could address current screen-time issues + give AI Agents a tracking surface beyond the phone-in-hand. Watch Gladiator — metal wrist straps for protection.
MP Take: We spend 7-8 hours daily on displays in our hands. Foldable wrist-sized display is not a weird notion. AI Agents need a tracking surface bigger than Apple Watch but smaller than phone-in-hand. Future ARD deep-dive topic.
Short — Congress Should Resume China Trips — See What’s Actually Changed Aspirational Step #1 from today’s deep-dive: Congress House + Senate take a deep breath. Hold off on the dystopian geopolitical pronouncements. Resume bipartisan trips to mainland China (not Taiwan as Pelosi did). They used to do this 2-3 times a year. US business + tech leaders + VCs already realized China is no longer COPYING US tech in many domains — and innovating beyond it in several (BYD/CATL/Unitree/DJI).
MP Take: The story has flipped. Yes China is still behind on chips (no ASML access). But on EVs, batteries, robots, drones — China is ahead. We can learn from them + benefit from JVs. Politicians visiting before the September Xi trip would go a long way.
Ep 77 scope: 1 Main + 4 Shorts + 2 Segments = 7 uploads. Second consecutive single-event Theme deep-dive (after Ep 76 Cerebras IPO).
About AI Ramblings Daily (ARD), and AI-RTZ
Both are daily. Both are free. Both are about AI. But they’re different mediums carrying different messages.
AI-RTZ is the morning text — a deeper written take on one idea, published by at least 5 AM EST. Today: post #1087 — The Lure of AI Companions.
AI Ramblings Daily is the afternoon video + podcast — my ad hoc takes and perspective on the day’s AI issues & news flow, around 18 minutes today (single-event deep-dive variant with two China segments + four shorter clips), with two embedded YouTube anchors on RESPECT and 2001 Space Odyssey. Today: episode #77.
Subscribe to either or both on michaelparekh.substack.com. They run as separate Sections you can opt into or out of.
Links used in today’s show (already embedded inline above; listed here for reference)
Theme framing — AI Tech Wave standing anchor:
Take 1 — First Xi/Trump Meeting of 2026:
Embedded YouTube videos (MP-canonical voice anchors in MP Take closing prose):
Gadget AI — Meta Ray-Ban Display + Wristband Controller:
Q1 + Q2 — AI Wearables (Meta wrist controller / bendable wrist wearables): (No external URLs — MP’s standing thesis. Both Qs extend yesterday’s Ep 76 AI Wearables thread.)
Companion text:
