Updated 2026-11-16.
What changed this week (checked 2026-11-16)
- Nothing new on tape. The two episodes since the last check, EP310 (feed upload 11 Nov 2026) and EP311 (13 Nov 2026), do not mention him. Both the whole-word grep on
Alvin|Grlin|Grayland|Graylinand the substring pass onalvin|grayl|greylcome back empty across transcripts and caption files. The whole-disk grep,backfill/included, still finds nine Moonshots transcripts that carry him. - Asia Society's own site uses the "Honorary" title. The Center for China Analysis newsletter, hosted on
centerforchinaanalysis.asiasociety.org(opened 2026-11-16), carries Reporting Back from WAIC 2026: How China's AI Ecosystem Is Adapting, dated 13 Aug 2026. The byline is "Center for China Analysis, Alvin W. Graylin, Paul Triolo, and Kristy Loke", and his section is headed "Non-Resident Honorary Senior Fellow on Technology, Asia Society Policy Institute's Center for China Analysis". Last week this qualifier was found only on the Chicago Council's page. It is now on an Asia Society–hosted page, so the full title is no longer a third-party reading. The piece predates this week but was not on this page before. Its subtitle reads: "Four experts assess how China is building around compute constraints across hardware, open models, safety, data, and governance." It was fetched and summarised, not read end to end. The CCA archive page shows only posts from 22 Sep onwards, and none of them carry his byline. - The dialogue timing, as of 16 Nov. A Brookings event page (opened 2026-11-16) for What is next for the US-China AI safety dialogue? (online, 23 Nov 2026) says: "Follow-up technical discussions are scheduled for November 2026, and will likely fall before the next meeting between Presidents Xi and Trump on November 18-19 in Shenzhen, China." He is not among the four speakers. Searches on 16 Nov found no confirmed date, venue or readout for the November technical round. Nothing found names him in connection with it.
- The earlier unconfirmed items stay unconfirmed. Searches on 16 Nov tied him to no session at Web Summit Lisbon (9–12 Nov) and found nothing on the Tallinn Digital Summit. A search summary again paraphrases a "recent interview" in which he hoped the late-November dialogue would produce something "more substantive". No source page was found, so it remains unverified and is not quoted as his words.
- Writing: nothing new. On 16 Nov the Substack archive still showed Coexistence Above All (12 Oct) as the newest of 11 posts. The Quincy author page still lists only The Wrong Race (18 Sep 2026), and its title line now reads "Senior Fellow, Technology, Asia Society". The CFR Myth of the AI Race episode is still dated 8 Jul 2026 (page JSON-LD).
- Record. 🔴 No White House advisory appointment or US delegation seat has appeared in the public record (searched 2026-11-16). The Asia Society bio returned 402 for the eighth week. The Stanford HAI people page again returned HTTP 200 with a body of only "404", for the third week. The Stanford Digital Economy Lab page is unchanged: "Digital Fellows • Seminar Speakers", title "Chairman, Virtual World Society | Lecturer, MIT", and the fellowship is undated. The Lawfare contributor page returned 403. The Cipher Brief URL tried this week (
/experts/alvin-wang-graylin) returned 404. This may be the wrong path rather than a removed page, because the 4 Nov page was opened from a different link. - No open programme, prize, grant or call turned up on any surface this week, so there are no pipeline rows.
Previous week (checked 2026-11-10)
- Nothing new on tape. The two episodes since the last check, EP308 (feed upload 4 Nov 2026) and #309 (7 Nov 2026), have no use of him. That holds for the whole-word grep on
Alvin|Grlin|Grayland|Graylinand for the substring pass onalvin|grayl|greyl, across transcripts and caption files. The whole-disk grep,backfill/included, still finds nine transcripts that carry him. - A Stanford page lists him for the first time, but with a different title. The Stanford Digital Economy Lab person page (
digitaleconomy.stanford.edu/person/alvin-graylin, opened 2026-11-10) lists him under "Digital Fellows • Seminar Speakers". His title line there is "Chairman, Virtual World Society | Lecturer, MIT", and it names one past seminar, "Alvin Graylin: The Realities of the U.S.-China AI Arms Race" (Monday, 21 Apr 2025). This is the first Stanford-hosted page found that puts him among the Digital Fellows. It is the Digital Economy Lab, not the HAI people page, which matches the "Stanford HAI/DEL" wording noted on 29 Oct. The page dates neither the fellowship nor its start. The only University of Washington mention on it is the BSEE degree. - The UW title: still nothing at the university. No UW page carrying the 22 Oct "Affiliate Professor" post turned up in this week's search either. The Lawfare contributor page is in the index with "professor of AI and technology policy at the University…", but the snippet is cut off and the page was not opened. The line stays as stated.
- The Asia Society title carries a qualifier on a third-party page. The Chicago Council on Global Affairs expert page (opened 2026-11-10) gives "Non-Resident Honorary Senior Fellow on Technology, Center for China Analysis, Asia Society Policy Institute". His other bios say "Senior Fellow". The same site lists one past event, The Global AI Race (28 Jul 2026), with Lindsay Gorman, moderated by V.S. Subrahmanian. That event predates this week but was not on this page before. ⚠️ This is the Council's wording, not Asia Society's, whose own bio returned 402 for the seventh week.
- Two appearances found only in a search summary, both unconfirmed. A search summary says a LinkedIn post has him speaking at the Tallinn Digital Summit (5–6 Nov 2026). The summit's own site (opened 2026-11-10) confirms those dates, but its seven "first speakers" do not include him and the speakers and agenda pages carry no names, so this is unverified. The same summary mentions a "Monitor interview" from about two weeks ago, in which he hoped the late-November dialogue would produce "something more substantive". No such piece was found. Unverified, and not quoted as his words.
- Web Summit is under way with no session found. The summit runs 9–12 Nov. The attendee page returned 404 to the fetcher for the second week, and no session, day or stage was found on any surface.
- Writing: nothing new. The Substack archive (opened 2026-11-10) still has Coexistence Above All (12 Oct) as the newest of 11 posts. The Quincy author page still lists only The Wrong Race (18 Sep 2026). The ChinaFile contributor page (bio "Last Updated: September 27, 2026") shows no piece bylined to him in the fetched text. The CFR Myth of the AI Race episode is still dated 8 Jul 2026 in its page metadata.
- Record. 🔴 Still no White House advisory appointment or US delegation seat in the public record (searched 2026-11-10). Stanford HAI people page: an HTTP 200 whose body is only "404", for the second week. On the dialogue, searches on 10 Nov found no confirmed date, venue or readout for the November round. A Shenzhen venue appears only in unverified summaries, and nothing names him.
- No open programme, prize, grant or call found on any surface this week. No pipeline rows.
Earlier (checked 2026-11-04)
- Nothing new on tape. The one episode since the last check, EP307 (feed upload 31 Oct 2026), has no use of him — neither the whole-word grep on
Alvin|Grlin|Grayland|Graylinnor the substring pass onalvin|grayl|greylacross the transcript and caption files. The whole-disk grep still finds nine transcripts that carry him (109 episodes on disk, watcher 4 Nov). One new substring hit in the #194 backfill is "Greylock" (Reid Hoffman's firm), not him. - The UW title: still nothing at the university. On 4 Nov the Evans School site search returns "No results for 'Graylin'" and the UW ECE site search "no results matched" — thirteen days after the 22 Oct self-announced "Affiliate Professor" X post, no UW page carries it. x.com still returns 402 to the fetcher and the read-only timeline pull could not be run this session, so the post itself remains known only from the search index. The line stays as stated.
- A new third-party surface carries the professor title — without "affiliate". The Cipher Brief expert page (opened 2026-11-04; page header dated 14 Oct 2026) gives "Professor of AI and Technology Policy at the University of Washington", alongside "Digital Fellow at Stanford's Human-Centered AI Institute" and the Asia Society senior fellowship. It lists two pieces by him, undated: The Biggest AI Models Are Not the Biggest Threats and America's AI Strategy Is Fighting the Last War — the second title was not on this page before (title only, not read). A Lawfare contributor page (
lawfaremedia.org/contributors/awgraylin) shows in search results but returned 403, unopened. ⚠️ Both are contributor bios — his own wording recycled, not UW's. - Web Summit, five days out: still listed, no session. The attendee page that opened on 29 Oct returned 404 to the fetcher on 4 Nov; the search index still lists it, now describing him as "Stanford HAI Digital Fellow, UW Professor, Asia Society Senior Fellow" (index wording, page unopened). No session, day or stage found on any surface. The summit runs 9–12 Nov.
- Writing: nothing new. The Substack archive (opened 2026-11-04) still has Coexistence Above All (12 Oct) as the newest of 11. The Quincy author page (opened 2026-11-04) still lists only The Wrong Race (18 Sep 2026).
- Record. 🔴 Still no White House advisory appointment or US delegation seat in the public record (searched 2026-11-04). Asia Society bio: 402 for the sixth week. Stanford HAI people page: now returns HTTP 200 whose entire body is the text "404" — effectively still gone. On the dialogue: searches on 4 Nov found no confirmed date, venue or participant list for the November round; a search summary of the SCMP commerce-ministry piece has the next dialogue convening "before the end of November" ⚠️ (page 403, headline only). Nothing names him.
- No open programme, prize, grant or call found on any surface this week. No pipeline rows.
Earlier (checked 2026-10-29)
- Nothing new on tape. The two episodes since the last check, EP305 (feed upload 23 Oct 2026) and EP306 (28 Oct 2026), have no use of him. That holds for the whole-word grep on
Alvin|Grlin|Grayland|Graylin, and also for a substring pass onalvin|grayl|greylacross the transcripts and the caption files. The whole-disk grep still finds nine transcripts that carry him (108 episodes on disk, watcher 29 Oct). - One new title, self-announced and not confirmed at the university. A search index returns an X post from his account dated 22 Oct 2026 (post). It says he is "starting a new position as Affiliate Professor at the University of Washington" to teach an "AI and Technology Policy" course, jointly offered by the UW College of Engineering and the Evans School. ⚠️ x.com returns 402 to the fetcher, so the post was not opened and the wording is the index's. On 29 Oct the Evans School and UW ECE site searches returned no results for "Graylin", and washington.edu search lists him only as an alumnus, guest speaker and advisory-board member. If it holds, this would be the first dated source for the "professor of AI policy at University of Washington" line on the #281 bio card. Until a UW page carries it, it stays as stated.
- Where the Stanford title sits. The UW ECE Advisory Board bio (opened 29 Oct) gives "Digital Fellow at Stanford's Digital Economy Lab (within HAI)", and his Web Summit card reads "Stanford HAI/DEL". That may explain why the HAI people page keeps returning 404. The title is still his own bio's wording, not Stanford's.
- Listed as a speaker at Web Summit Lisbon (9–12 Nov 2026, dates from websummit.com): "Author "Our Next Reality", Digital-Fellow", with no session listed (attendee page opened 2026-10-29). This is an appearance, not a programme.
- Writing: nothing new. The Substack archive and RSS feed (both opened 2026-10-29) still have Coexistence Above All (12 Oct) as the newest post, out of 11. The Quincy author page still lists only The Wrong Race (18 Sep 2026). The CFR Myth of the AI Race episode is dated 8 Jul 2026 by its page metadata. That date was not on this page before.
- Record. 🔴 Still no White House advisory appointment or US delegation seat in the public record (searched 2026-10-29). Asia Society bio: 402 for the fifth week. Stanford HAI people page: 404. On the dialogue, search results carry a CNBC headline, "China confirms first AI talks with U.S. have taken place" (URL dated 24 Sep 2026), and an SCMP report that the commerce ministry confirmed a first dialogue led by He Lifeng and Bessent. ⚠️ Both pages returned 403 and were read as headlines only. If they hold, a first round of talks happened in September, and the 30 Sep "before November" line refers to a later meeting. Neither headline names him.
- No open programme, prize, grant or call found on any surface this week. No pipeline rows.
Earlier (checked 2026-10-23)
- Nothing new on tape this week. The two episodes since the last check, EP303 (feed upload 17 Oct 2026) and EP304 (21 Oct 2026), have no use of him under
Alvin|Grlin|Grayland|Graylinas a whole word. (A substring grep hits "halving" in EP303's robot-cost segment; that is not him.) The 15 Oct file was re-checked: still none. - Two older transcripts were never logged, and the full-disk grep found them. The backfill files for #269 (feed upload 8 Jul 2026) and #282 (21 Aug 2026) carry four uses missing from the table. Speakers are not labelled in either file, so attribution is open:
- #269
0:45:23, six weeks before he was a guest: "I was speaking to uh Alvin Grlin who's a a friend of of ours about you know USChina relationships and the question is is there a structure in which you know we can see a US China alignment on AI" [attr?, the host's question to the panel]. This is the earliest use on disk, and he comes in as a contact, with no credential. - #282
0:16:17: "the September 25th visit 24th and 25th visit that Alvin was talking about on our… last podcast" [attr?]. The summit date is relayed from #281. - #282
0:26:08–0:26:30: "Alvin said on the last podcast, you know, at an internal anthropic meeting, not not validated. Alvin said it, but uh internal anthropic meeting, they said pretty soon there will only be one company and that will be anthropic" [attr?]. The speaker flags it as unvalidated on air. 🔴 A private account of another company's internal meeting, passed along second-hand. Do not repeat it as fact. - #282
1:42:39: "I didn't know until Alvin said it, but the US… had 50% of the world's manufacturing capacity… now it's one-third China, and he said it was about 15% US" [attr?]. ⚠️ These figures are relayed and were not checked at source this week. Now nine transcripts on disk carry him: #269, #281–#284, #289, #291, #294 and 7 Oct.
- #269
- Writing: nothing new. The Substack archive (opened 2026-10-23) shows Coexistence Above All (12 Oct) as the newest post. The archive lists 11 posts, and two of them were never named here: What Homer's Sirens and Nymphs Teaches Us About Thriving With AI and Post-Scarcity (6 Jan 2026) and Should We be Returning to Philosopher King Rule or Will AI Take the Throne? (5 Apr 2026). Titles and dates only. The Quincy author page still lists only The Wrong Race (18 Sep 2026).
- One public appearance, from a secondary source. IoT-Mesh, 19 Oct 2026 (report, opened 2026-10-23): a DISQ event at Soho House Berlin, event date not stated. It describes him only as an "American-Chinese technology expert" and paraphrases rather than quotes him. Advanced AI should be "widely accessible" rather than run as a private race, and it notes recent visits to China and South Korea. No government role, no dialogue, no call.
- Record. 🔴 Still no White House advisory appointment or US delegation seat in the public record (searched 2026-10-23). The searches this week turned up no date or readout for the first US–China AI dialogue meeting, so the "before November" wording on 30 Sep is still unconfirmed. Asia Society bio: 402 for the fourth week. Stanford HAI people page: 404 again. A DiarioBitcoin piece (20 Aug 2026) has his "helping prepare" the 24 Sep dialogues, but it is a recap of #281 and adds no independent source.
- No open programme, prize, grant or call found on any surface this week. No pipeline rows.
Earlier (checked 2026-10-17)
- One use on tape, and it was missed last week. The 7 Oct 2026 episode (unnumbered on the feed, EP301 by sequence; transcript on disk since 8 Oct) carries him once, in the Super Intelligence Force segment, by Dave Blundin [attr. from Peter's hand-off "Dave, what do you think about that the task force…"
0:27:01]: "Meanwhile, Alvin Graylin, our friend of the pod, is like, China is like, how is this a race? It goes to infinity for the next millennium." (0:27:18–0:27:25, Whisper). No credential attached this time, and for the first time since #281 the panel relays his argument rather than his access, in the same segment where Alex calls the task force a "1942 moment" (0:26:18). ⚠️ Correction to the 10-11 line below: that check called #298 the newest held transcript; the 7 Oct file was already on disk and was not grepped. Now seven held transcripts. The 15 Oct episode (on disk 15 Oct) has no use of him under any of the five spellings. - New writing, 12 Oct 2026. Coexistence Above All: Why We Should Raise AI, Not Race It (subtitle: An alternative approach to AI safety and alignment in a world constrained by increasing competitive dynamics), on his own Substack, Abundanist (post, opened 2026-10-17; the page shows "Oct 12", year inferred from the archive order). Read by fetch and summarised, not end to end: opens on "a week of Washington meetings" with labs, chip firms and national-security figures, states no official capacity; thesis "speed itself is now the risk"; proposes "a verifiable pacing agreement every nation can join". It names the US–China AI dialogue as the two presidents' agreement ("Last month Presidents Trump and Xi agreed to a standing US-China AI dialogue with a channel for reporting AI incidents"). It does not mention the White House task force, a delegation, the November meeting or any role of his own.
- A second surface the page did not have: the Substack.
abundanist.substack.com, "By Alvin W. Graylin", 11 posts since 13 May 2025 (archive opened 2026-10-17). It dates two essays that were only names on tape: Beyond Rivalry (14 Dec 2025) and The Great Reckoning Before the Reconnecting (14 Jul 2026). Also there: Abundanism (13 May 2025), A "Bill of Rights" for the Age of AI (14 Jul 2025), America Is Running the Wrong AI Race (29 Jul 2025), The AGI Windfall Mirage (1 Jun 2026), The Post-Labor Prophecy (24 Jun 2026), The Last Straw: How the AI Race May Deepen America's $40 Trillion Debt Crisis, Not Fix It (28 Aug 2026). Titles and dates only; none read end to end. - Record. 🔴 Still no White House advisory appointment or US delegation seat in the public record (searched 2026-10-17). Jay Clayton was named head of the White House task force on 4 Oct 2026 (Korea Times, Khaleej Times; secondary); a search summary says he criticised the appointment on X the same day, but
x.com/AGraylinreturned 402, so that post is unverified. Asia Society bio 402 for the third week; the Stanford HAI people page returned 404 and the directory search nothing, so the Digital Fellow title still rests on secondary speaker pages (Aspen Ideas, AWE). One row moved on a primary source: UW ECE's own page (29 Feb 2024) calls him "UW ECE alumnus Alvin Graylin (BSEE '93)"; the BSEE is now confirmed at the university, the "professor of AI policy at University of Washington" line is not on that page and stays as stated on tape. An Investopia speaker page lists him as "President, Virtual Reality Venture Capital Alliance" with no date; a search summary placed him at Investopia Bridge, Abu Dhabi, 7 Oct 2026, but the WAM coverage does not name him, so that is unverified and probably stale. - No open programme, prize, grant or call found on any surface this week; no pipeline rows.
Earlier (checked 2026-10-11)
- Nothing new on tape was the finding; see the correction above (the 7 Oct transcript was on disk and unchecked). 🔴 No White House advisory appointment or US delegation seat in the public record (search 2026-10-11 returned none).
- No new writing found dated 5–11 Oct; the Substack had not yet been located.
- The dialogue he is credited with, sourced to a report of the White House fact sheet. Enterprise AM, 30 Sep 2026: the US and China are launching an AI dialogue; "The first meeting will happen before November, according to a fact sheet," and the two sides "agreed to keep a communication line open to defuse serious incidents" (Enterprise AM, opened 2026-10-11). ⚠️ Secondary summary; the fact sheet itself was not opened, no meeting date is given, and "before November" differs from the earlier "by November" wording (Axios/TNW summaries, unverified). No source ties him to the dialogue or a delegation.
- Carried forward: the Axios and Asia Society pages stay unverified (403 / 402).
Earlier (checked 2026-10-05)
- The Wrong Race: The US, China, and AI Competition, Quincy Institute, 18 Sep 2026 (Quincy, opened 2026-10-05; still his only Quincy piece on 2026-10-17). Stated thesis: no decisive finish line; smaller specialised models are the sharper shared risk; a "dual-track engagement" of competition plus cooperation on safety standards, incident notification and evaluation.
- Al Jazeera, 23 Sep 2026: "The two countries host nearly all frontier labs and most of the models people download"; bilateral guardrails are "close to a precondition" (Al Jazeera).
- Search summaries of Axios / The Next Web (26 Sep) report a "Super Intelligence Dialogue" and an AI-incident channel, next exchange by November 2026. ⚠️ Axios 403, Asia Society 402: the readouts, the Shenzhen venue and his own commentary are unverified here.
What he is here for: the person the Moonshots with Peter Diamandis panel cites as its inside line on the US–China AI relationship, and whose published position is the opposite of the one the panel attaches his name to. A full guest on #281 (2026-08-18); since then a "friend of the pod" whose texts are read as the summit briefing. Cited in nine transcripts on disk (#269 the earliest), untracked by the regex index. The gap between what he argues and how he is used is the page.
How to read the tags. Public record = his own institutional pages or a named outlet. On
tape = said on Moonshots; auto-captions and Whisper carry no speaker labels, so attribution is
inferred from the hand-off and marked where uncertain. The captions spell him Grlin, Grayland
and Alvin; only the forms quoted here are used.
Who he is
Public record (checked 2026-10-02): Senior Fellow on Technology at the Asia Society Policy
Institute's Center for China Analysis; Digital Fellow at Stanford's Institute for Human-Centered
AI; a lecturer at MIT (Asia Society).
The Center for China Analysis's own newsletter (13 Aug 2026, opened 2026-11-16) gives the full form as
"Non-Resident Honorary Senior Fellow on Technology".
Co-author of Our Next Reality (2024), discussed on this channel at the time (#281 0:02:43).
The show's own bio card (Peter, #281 0:02:21–0:03:26, on tape): senior executive roles at
"HTC, Intel, IBM, Trend Micro, and Watchguard"; "built and founded over four startups"; "a
senior fellow at the Asia Society Policy Institute, professor of AI policy at University of
Washington"; dual master's from MIT and an electrical-engineering degree from the University of
Washington; "lived and operated extensively in the US, China, and Taiwan." ⚠️ The University of
Washington professorship and the startup count are the host's words, not checked at source.
The credential the panel keeps repeating, and where it comes from. Peter, #281 0:03:04:
"He's currently supporting the US government and for the coming US–China AI safety dialogues.
That's going to be happening on September the 24th in DC." That sentence is the only source for
every later use:
| on tape | said by | what the record shows |
|---|---|---|
"our friend of the pod Alvin Grlin declared himself the top expert on China AI race and now he got dragged in for the United Nations meetings on September 24th as a top adviser to the White House" (Moonshots #289 0:27:25) |
Dave Blundin | 🔴 No White House advisory appointment and no US delegation seat found in the public record (searched 2026-09-13 and again 2026-10-02). |
"also Alvin Grlin will be" in the room, alongside David Sacks (Moonshots #291 0:32:30) |
Dave Blundin [attr?] | as above |
"You were texting with him last night, and he's in the middle of it all. He's advising the White House, he's at the UN with the delegation" (Moonshots #294 0:19:02) |
Peter Diamandis | as above. That week he was quoted by Al Jazeera as an outside analyst on the summit and, on 29 Sep, on the industry's voluntary safety pact ("fell far short of what was needed"). |
Read: the credential is not in filings or the public record. Absence is not a contradiction — an informal advisory role would leave no trace — but it means every claim sourced to him through the panel is one private account, and must be attributed to the panel's report of him, never stated.
What he argues, in his own framing
From the #281 introduction and his published work. These are positions, labelled as such.
- The US–China AI relationship is a stag hunt, not a prisoner's dilemma — a coordination game
about trust, not a race. Peter's summary of his papers: "You've argued that the game isn't a
prisoners dilemma. It's a stag hunt, a game theory model about coordination and trust" (#281
0:03:49). Papers named on tape: Beyond Rivalry and Misdiagnosing the US–China AI Race (0:03:59); an essay, The Great Reckoning Before the Reconnecting (0:05:05); and "abundantism" as his term (0:05:16). On his Substack these are dated: Beyond Rivalry 14 Dec 2025, The Great Reckoning 14 Jul 2026, Abundanism 13 May 2025 (archive opened 2026-10-17). - "Raise it, don't race it." The 12 Oct 2026 essay Coexistence Above All moves from the game-theory frame to pacing: "speed itself is now the risk", and a "verifiable pacing agreement every nation can join" (fetched and summarised 2026-10-17, not read end to end).
- The biggest models are not the biggest threats. "he proposes that it's the smaller AI models
we need to be more worried about and cooperation is the only path toward safety" (#281
0:04:10, Peter's summary of the paper). - "The myth of the AI race." His CFR appearance carries that title (The President's Inbox, CFR); the Quincy Institute piece is The Wrong Race (18 Sep 2026, checked at source 2026-10-05).
- On what China's leadership believes: "They are not behaving like they believe ASI is around
the corner" (#281
0:00:12, cold open) and, as relayed by Dave Blundin on Moonshots #2840:50:30: "if you talk to Alvin Graylin, he's like they don't even think vaguely that way. They're just trying to have a…" [cut].
How the panel uses him
| episode | use | verbatim |
|---|---|---|
| #269 (feed upload 2026-07-08; backfill, logged 2026-10-23) | a contact, before he was a guest: the source for a question on whether US–China AI alignment has a structure; no credential | "I was speaking to uh Alvin Grlin who's a a friend of of ours about you know USChina relationships" 0:45:23 [attr?] |
| moonshots 281 (2026-08-18) | the guest, for the whole episode: "China's Endgame: ASI Timelines, US–China Relations, and the $1.7T AI Bubble" | "someone who's lived and operated inside both the Chinese and US technology ecosystems for over 35 years" 0:01:50 |
| #282 (feed upload 2026-08-21; backfill, logged 2026-10-23) | relayed from #281, three times: the summit date; an Anthropic internal-meeting claim, flagged "not validated" on air 🔴; a manufacturing-share figure ⚠️ | "Alvin said on the last podcast, you know, at an internal anthropic meeting, not not validated" 0:26:08; "the September 25th visit 24th and 25th visit that Alvin was talking about" 0:16:17; "he said it was about 15% US" 1:42:39 [attr?] |
| Moonshots #283 | the authority on Chinese press and perception | "Dave, this is what Al Alvin said on that, right?… China has seen a technology revolution" · "Alvin also said, 'We will habitually report if if one Waymo in one corner of San Francisco runs over…'" 1:48:30–1:49:10 [attr?] |
| Moonshots #284 | the authority on what the CCP intends | "if you talk to Alvin Graylin, he's like they don't even think vaguely that way" 0:50:30 |
| Moonshots #289 | the career template for getting into government on the China race — the segment in which Alex argues the race must be won | 0:27:25 (above) |
| Moonshots #291 | "in the room" for the summit | 0:32:30 (above) |
| Moonshots #294 | the summit briefing, by text: the US agenda is "stop US panic"; China's concern is "the CCP existing at all and the great firewall of China being penetrated by AI"; the parties "agreed to meet again in November" | Dave Blundin, 0:18:58–0:19:54 |
| Moonshots #301 (EP301 by sequence) | his argument, for once: the counter-voice to the task-force-as-race segment; no credential attached | "Meanwhile, Alvin Graylin, our friend of the pod, is like, China is like, how is this a race? It goes to infinity for the next millennium." Dave Blundin [attr. from hand-off], 0:27:18 |
The finding, stated once. On #289 the panel holds him up as the model for entering government on the China AI race, in a segment asserting that the race must be won; his published position is that the race framing is a myth. On #294 the one checkable thing relayed from his texts — a next meeting in November — is right (the bilateral AI dialogue announced by Bessent and He Lifeng is scheduled for November 2026); everything else relayed is a private account of two governments' intentions and cannot be checked. The panel borrows his access and discards his argument. The 7 Oct use is the one exception in nine transcripts: his "how is this a race?" is relayed as his view, without a credential, and then the segment moves on.
Disclosures on #281 — asked for, and given
To the show's credit, Peter asked for them: "one of the things we pride ourselves on the show is
disclosing all of our connections… your bio lists roles like vice chairman of [a] VR industry
alliance endorsed by the Chinese Ministry of Industry and Information Tech, and a three-year
professorship at Beihang, which is a defense link university, on the US entities list… Were you
compensated?" (0:05:26–0:06:30). His answer, on tape: a US citizen "for over 45 years", born
in China (0:06:30); the alliance was an industry association with "300 plus members and about a
third of them were international companies… Nvidia and Samsung and Qualcomm and AMD and Google",
which he led while running HTC's China business, "a Taiwanese company" (0:06:54–0:07:48);
Beihang he describes as the leading university for VR/AR teaching (0:07:48). ⚠️ Compensation and
current status were asked; the answer as transcribed does not state either. This is the one
episode where the panel ran a disclosure interview on a guest. It did not carry the practice
forward to its own use of him. → Moonshots: the disclosures ledger
What an operator can take from him
- Read the papers before the clip. The "stag hunt" framing is a usable test for any US–China claim on the show: does the story assume a race (prisoner's dilemma) or a coordination problem (stag hunt)? The panel assumes the first; its own expert argues the second.
- Treat relayed texts as one source. Anything introduced with "Alvin said" or "texting with him last night" is a private account; log it as such.
Questions for him
- Which parts of the US–China AI dialogue process has he supported, in what capacity, and is any of it public? The panel has credited him three times; he has not said it on this tape himself.
- If the biggest models are not the biggest threats, which small-model incident would move the dialogue first — and did the Medicare intrusion (an OpenAI agent, July 2026) count?
- What would falsify the stag-hunt reading? The panel's race framing has no stated falsifier; his framing should have one.
Where it came up
- #269
0:45:23· moonshots 281 · #2820:16:17,0:26:08,1:42:39· Moonshots #2831:48:30· Moonshots #2840:50:30· Moonshots #2890:27:25· Moonshots #2910:32:30· Moonshots #2940:18:58· Moonshots #3010:27:18
How this page stays current
Refreshed weekly by the Moonshots lane.
- Tape. Grep new transcripts for
Alvin|Grlin|Grayland|Graylin; add each use to the table with the verbatim and the timestamp, and mark whether a credential was attached. - Record. Re-check the Asia Society and Stanford HAI pages, and search for any announced government role. Move the credential row only on a primary source.
- His writing. Add new papers or essays by title and date. Do not summarise beyond the title
until read. Surfaces: Quincy author page, the Substack archive (
abundanist.substack.com/archive), CFR. X returns 402 to the fetcher; use the read-only timeline pull if a post is load-bearing. Grep every transcript on disk, not the newest numbered one (the 7 Oct unnumbered file was missed once). That includesbackfill/(#269 and #282 were on disk from 24 Sep and missed until 2026-10-23). Grep whole words only: a substring match onalvinalso hits "halving".
⚠️ Unverified
The bio card is the host's; the University of Washington professorship, the startup count, the HTC and other executive titles, and the MIT degrees are as stated on tape (the UW BSEE '93 is confirmed by UW ECE's own page, 29 Feb 2024). The 22 Oct "Affiliate Professor" X post is known only from a search index (x.com 402), and no UW page carried it on 2026-10-29 or 2026-11-04. The Cipher Brief bio's "Professor of AI and Technology Policy at the University of Washington" (opened 11-04) is a contributor bio, not a UW source; the Lawfare contributor page returned 403, unopened. The Quincy piece was opened 2026-10-05; the Asia Society bio returned 402 on 10-05, 10-11, 10-17, 10-23, 10-29, 11-04, 11-10 and 11-16; the Stanford HAI people page returned 404 on 10-17, 10-23 and 10-29, and on 11-04, 11-10 and 11-16 an HTTP 200 whose body is only "404". The Stanford Digital Economy Lab page (opened 11-10) lists him under "Digital Fellows" but gives his title as "Chairman, Virtual World Society | Lecturer, MIT" and does not date the fellowship. The Chicago Council's "Non-Resident Honorary Senior Fellow" wording is now matched by Asia Society's own Center for China Analysis newsletter (13 Aug 2026, opened 11-16); the 13 Aug WAIC piece was fetched and summarised, not read end to end. The Cipher Brief path tried on 11-16 returned 404 (possibly a wrong path). The 18–19 Nov Shenzhen leaders' meeting date is Brookings' wording. The Tallinn Digital Summit appearance (5–6 Nov) and a "Monitor interview" exist only in one search summary; neither is on a page that was opened. The September "first AI talks" reports (CNBC, SCMP) are headlines only, both pages 403; the "before the end of November" timing for the next dialogue round is a search summary of the same SCMP piece. The #269 and #282 uses are unattributed (no speaker labels), and the #282 Anthropic-meeting claim and manufacturing-share figures are relayed second-hand, not checked. The 19 Oct Berlin appearance rests on one trade-site report that paraphrases him. The 12 Oct Substack essay was fetched and summarised, not read end to end; the Substack dates for earlier essays are titles and dates only. The 4 Oct X post on the Clayton appointment exists only in a search summary. All quotes from tape are auto-captions or a Whisper pass; attribution is inferred from hand-offs.