Jim Rutt

Who he was

Affiliations at death: SFI Trustee Emeritus and Distinguished Fellow — researcher in residence 2002–2004; "In 2003, he joined the SFI Board of Trustees, served as Board Chair from 2009–2012, and remained an active Board member until 2022" (SFI, 2026-07-06; the Fralin bio's "elected chairman in 2010" conflicts — SFI's own memoriam wins). Chairman of the Board, California Institute for Machine Consciousness (CIMC, Joscha Bach's nonprofit; launched 2025-05-30; "Chairman of CIMC's Board," CIMC, 2025-06-24; he repeated the title himself on 2025-10-09; an undated Deep Tech Week listing now says "Board Chairman Emeritus" — the date of that change is not established). Host, The Jim Rutt Show — "Since the show's first episode in July, 2019, with SFI External Professor Simon DeDeo, Rutt hosted 346 episodes" (SFI); 461 total posts including Currents shorts, weekly cadence, 4.8/5 from 603 ratings (Rephonic). Co-founder of the MIT Free Speech Alliance after MIT's September-2021 Dorian Abbot disinvitation — "served as its first board chair," "some 1,200 members" (MFSA, 2026-09-05; his own Works page says "president"; the obituary says "founding board member" — three sources, three titles, one role). Also: MIT Brain & Cognitive Sciences visiting committee; advisory board, Fralin Biomedical Research Institute at Virginia Tech; co-founder, Staunton (VA) Makerspace (2014, Wikipedia); member and moderator of The WELL since 1989 and part of the user group that bought it from Salon in 2012 (his Works page); executive producer of the Game~B film (An Initiation to Game~B, no release date on its site) and Zombiewood; "JPR Ventures" is named on his SFI profile as his current affiliation — no public portfolio found under that name.
Earlier career (obituary + SFI profile): The Source (early online service); co-founded Business Research Corporation with William A. Benjamin, which built Investext; First Call (online equity research); first CTO of Thomson Corporation; CEO, Network Solutions — "In May 1999, Jim Rutt was named chief executive officer of the company; he stepped down in February 2001" (Wikipedia, citing Washington Post 2001-02-27). The VeriSign deal: announced 2000-03-07 "for $21 billion in stock" (ZDNet via Wikipedia); "On June 8, 2000, VeriSign, Inc. completed the acquisition of Network Solutions, Inc." for "approximately 78 million shares of VeriSign common stock" at 1.075 shares each, no dollar value stated (VeriSign 8-K filed 2000-06-19). SFI and Fralin both say "$15 billion" — the same deal valued at VeriSign's lower June share price; SFI's "In 2001" is the year he left, not the closing year. Chairman of Analog Design Automation, whose assets Synopsys agreed to buy on 2004-01-28 for an undisclosed sum (Synopsys / EE Times). Appointed to the New Mexico State Investment Council by Gov. Richardson in 2006 (obituary). MIT Sloan B.S. Management 1975. Wikipedia credits him with first using "snail mail" in 1981. Survived by Celia (45 years married), daughter Kat Walker, two granddaughters.
One-line position (v1, carried): The current civilizational operating system ("Game A" — rivalrous, extractive, growth-on-a-finite-planet) is hitting hard limits; AI is an accelerant that could tip it; the response is not reform but designing a successor ("Game B") robust enough to handle complexity without authoritarian collapse. Final form (2025-10 to 2026-05): Game A runs on one signal — "the money-on-money return loop, a single-dimensional signal" — and exponential growth on a finite planet makes it "self-terminating" (his own post, 2025-10-09); Game B is "membranes (semi-permeable boundaries)," "coherent pluralism," and "multi-dimensional economics." His last essay on the subject reads as a verdict: "Intelligence is easier to scale than wisdom, and power is easier to distribute than responsibility … The hour is late. The test is underway" (The Silent Sky and the Test Ahead, 2026-05-04).

Discipline & technical bet

Complexity science applied to social systems, by an operator rather than an academic: he ran the .com/.net/.org registry through the bubble, chaired the Santa Fe Institute, then spent seven years interviewing the field. His method in the last year was explicitly quantitative when it could be — a "violence reproduction number" Rᵥ for political assassinations ("If the geometric mean exceeds 1 (i.e., N × N′ > 1 on average), incidents grow superlinearly; if ≤ 1, they stabilize or decay," 2025-09-11) and a ratchet model of mid-decade redistricting (2026-04-21). Concrete wagers on the record:

• Machine consciousness is a testable research program. As CIMC chair he backed a computationalist institute whose mission is "to formulate testable theories of machine consciousness" (CIMC, 2025-06-24); his own gloss: machine consciousness would be "analogous to human consciousness, not identical" (2025-10-09). This was his only institutional AI bet.
• Personal information agents as sense-making infrastructure. He built a prototype that "processes 117 Substacks daily" with AI summarization, pitched as "collective sense-making" with "no fucking advertising" (2025-10-09); in EP 284 he described agents that align with each other — "my info agent aligns itself with other info agents … almost identical to your idea of the Civitas" (2025-02-13).
• The "priesthood of the power supply." His alternative to a superintelligence treaty: "all facilities able to generate a megawatt or more of power ought to be under the supervision of the priesthood … authorized … to first shut down the interconnects into data centers … second, shut down all the power supplies" — "more practical and probably almost as effective" than research bans (EP 327, 2025-10-15).
• Liquid democracy — "a technologically enabled hybrid of direct and representative democracy" where "Every voter can vote directly on any issue, any time" or delegate a revocable proxy across roughly 25 issue areas; funded by a "Democracy Allowance" of $5 per registered voter per month (~$12.7 billion a year) paid to delegates; entry via pledge-candidates for city councils and school boards (2026-05-14). No timeline given.
• LLM skepticism, held to the end. "I remain an LLM skeptic … I wouldn't bet the house, right? Though I may be betting the world"; transformers are "one of the less dangerous architectures"; the risk "keyhole" he watched was reasoning models becoming "able to actually reason about scientific theory" (EP 327, 2025-10-15).

Key claims (Says)

• Game A is structurally limited (empirical + normative): zero-sum competition, extraction, and GDP-as-scoreboard work at small scale and break at planetary scale (ecological overshoot, institutional decay, meaning crisis) — v1: contested; "limits" thesis is decades old and repeatedly deferred. Held on endorsement — restated as "self-terminating" (2025-10-09) and as "exponential growth on a finite planet driven by an economic system" in his own worldview interview (EP 332, 2026-02-17); the signature framing never softened. Outcome half: unfalsifiable — rewrite or retire. He never attached a date or an observable; the closest proxy is that AI could compress "60 years" of Game A runway into "30" (EP 231, 2024-03-27), which still has no anchor year.
• Multipolar traps / "Moloch" are the core mechanism (empirical/interpretive): rational local actors locked in competition produce collectively catastrophic outcomes nobody chose; AI arms-race dynamics are the live instance. Held — and extended in-window from AI to two other domains with the same math: political violence (Rᵥ, 2025-09-11) and redistricting — "Every locally rational move adds up to a House of Representatives drawn by partisan operatives in state capitals rather than by anything resembling the population" (2026-04-21). In his last essay competitive dynamics are named as potentially more dangerous than any single technology (2026-05-04).
• AI is an accelerant of Game A, not a neutral tool (empirical bet): it deepens concentration, speed, and race pressure faster than institutions can absorb — the danger is systemic destabilization, not (only) rogue ASI. Held — "AI in general, even not just talking about singularity, is clearly accelerating Game A" (EP 284, 2025-02-13); "The pervasive speeding up of everything is its own kind of risk" was risk #9 in his own taxonomy (EP 231, 2024-03-27). Check: by 2030, has any frontier developer paused a flagship training run for more than 90 days on safety grounds? A yes weakens the accelerant claim; none found as of 2026-09.
• AI could also be a Game B enabler (normative/speculative): better collective sense-making, coordination, and individual capability could lower the cost of cooperative structures — v1: weak evidence; aspirational. Held (still untested) — he narrowed it to a specific architecture, "intimate AI" agents networked into a Civitas-like structure (EP 284), and built a one-user prototype (2025-10-09). No adoption evidence; the claim now has an artifact but no result.
• Game B is a direction, not a plan (normative): a set of design principles (anti-rivalry, coherence without central control, individual sovereignty inside collective intelligence) for what grows inside Game A and carries the load when it breaks. Drifted — the direction got more specified while the movement got less active. He had already written a stepwise path in 2020 ("From today to a civilization level social operating system by a series of doable steps," announcing A Journey to GameB, 2020-01-13) and by 2025–26 named mechanisms (membranes, coherent pluralism, multi-dimensional economics, liquid democracy). Meanwhile the community's own infrastructure went quiet: gameb.wiki's main page was "last edited on 29 August 2021"; game-b.org is a "Request to Join" page with no dated projects; the film he executive-produced has no release date (all read 2026-09-21). Lydia Laurenson's tribute (2026-06-03) describes Game~B as "a loose subcultural community" and "a memetic tag."
• Sovereignty as prerequisite (normative): individual capacity for sense-making and choice without dependence on institutional authority is the unit Game B is built from. Drifted — the primitive changed. By 2025 his organizing idea was "coherent pluralism," which he called "essentially a technique," not a foundation: "we acknowledge there are no foundations … allow intelligent collections of humans … to experiment in the high-dimensional design space" (EP 284, 2025-02-13); values became "emergent properties of a culture's exploration of its adjacent possible" (2026-04-01); and liquid democracy (2026-05-14) lets a low-capacity voter delegate rather than requiring sovereignty of everyone — a partial answer to the v1 "average is over" objection.
• Collective intelligence over consensus (normative/empirical): groups that think better than any member via emergent coherence — not voting, not central planning. Held — liquid democracy's revocable, chainable proxies and "fractional voting" (2026-05-14) are exactly this; so is the networked-agent design.
• Complexity science is the right lens (epistemic): civilization is a complex adaptive system; agent-based / evolutionary dynamics, not equilibrium economics, predict its behavior. Held — the Rᵥ branching model and the redistricting ratchet are the lens in use; and the epistemic humility that goes with it was his stated creed: "we are grossly overconfident in how much we know and how much we can control the unfolding of the universe" (EP 332, 2026-02-17).
• The metacrisis is coordination, not any single threat (interpretive): AI, climate, institutional decay, and meaning collapse are facets of one generator-function failure — v1: overlaps Schmachtenberger; TBD: distinguish Rutt's version. Held; TBD resolved: Rutt's version is game-theoretic and, where he could, numerical (reproduction numbers, ratchets), and his last essay recast it as a Great Filter argument — "technological civilizations emerge often and destroy themselves quickly" is one of four cases, and "the hardest step may not be behind us. It may be ahead" (2026-05-04). Schmachtenberger's is a generator-function diagnosis; Rutt's is a hazard-rate argument.
• NEW — "overindex on the opacity issue" (empirical): Yudkowsky/Soares overweight model opacity because interpretability is progressing (Anthropic features, activation vectors); "maybe something way less than an ASI can solve the alignment problem" (EP 327, 2025-10-15). Drifted — the bench's own Olah file records independent sanity checks in 2026-02 that recovered ~9% of ground-truth features, with random baselines matching; his hedge ("may") keeps this off Wrong, but the progress he leaned on was overstated at the time he cited it.
• NEW — transformers are "one of the less dangerous architectures" (empirical): "I ain't too scared of one of them things" (EP 327). Held (so far) — check: by end-2028, does any LLM-family system produce, end to end, a novel scientific theory accepted in peer review (his own "keyhole")? If yes, the claim fails on his own terms.
• NEW — humans may be alone, and the universe is "fine tuned not only for stars, but also for emergence": "humans may be the only general intelligence in the universe"; AGI is "a potential third" major transition after language and science; three core values — "human well-being, ecological richness, and preserving humanity's path to bring the universe to life" (EP 332, 2026-02-17). Unfalsifiable — noted as the value frame under everything else, not graded.
• NEW — no supernatural, no simulation: "I am a strong disbeliever in supernatural anything" (EP 332); on simulation theory, "The concept has no value, and therefore I act on the basis of the actual reality that I perceive" (EP 342, 2026-05-06). Position, not a forecast; it is the live split with Hall (below).

Notable predictions — with falsifiable checks

• (~2013–2016, ongoing) Game A is approaching systemic limits and will face escalating crises absent a transition — v1: not cleanly falsifiable; "limits" framing risks moving goalposts (CONVENTIONS rule 18). Unfalsifiable — rewrite or retire. Rewrite offered: "by 2036 a named planetary-scale metric (choose one: real GDP per capita in the OECD, global cereal yield per hectare, or the Ecological Footprint overshoot day) reverses trend for three consecutive years." He never chose one; the bench should, or retire it.
• (Ongoing, ~2020–) AI will intensify multipolar-trap / arms-race dynamics faster than institutions adapt — v1: directionally consistent; TBD verify dated claims. Held; dated claims found and attached: (a) "autonomous tanks … within 10 years, we could almost certainly have" (EP 181, 2023-03-23) — check 2033-03: any state fielding armed ground vehicles that select and engage targets without a human in the loop; (b) "we'll get to … humans using strong, narrow AI for bad purposes before we have to deal with AGI" (same episode) — consistent with 2023–26 misuse patterns, no reversal found.
• (Ongoing) No single actor "wins" AI cleanly; coordination failure is the dominant risk channel — v1: TBD; contested vs fast-takeoff single-lab framings. Held (live) — but note his own admission that the single-actor branch has no answer: "I've never seen a convincing story on how to preserve humanity in the case of a FOOM" (EP 231, 2024-03-27). Check: if by 2030 one developer holds a capability lead of two or more years by consensus evals, the "no clean winner" branch fails.
• (Speculative) Game B structures will emerge inside Game A before Game A fully breaks — v1: weak evidence; aspirational. Drifted — fourteen years after he "co-founded the Game-B Network" (2012, Fralin bio) the visible structures are a wiki dormant since 2021, a join-request page, an unreleased film, and "an annual budget for 'microgrants' to help lubricate worthy endeavors in the liminal web world" (Laurenson, 2026-06-03). Check: by 2030, name one organization with a budget of $1M/yr or more that self-describes as Game B; none found 2026-09.
• NEW (2024-03-27) AGI timing — "Some people are arguing that artificial general intelligence … could be as close as three to five years away. I have my doubts about that. If I had to hang a number, it's sort of in the 10 to 20 range" (EP 231). Pending — window 2034–2044. The near half of his doubt is already gradeable: the 3–5-year camp's window closes 2027–2029; log the outcome then.
• NEW (2025-09-11) After the Kirk assassination: "Most shocks of this kind do not sustain an Rᵥ above 1; they flare, then settle," unless "early rounds show multi‑for‑one retaliation, shortening lags, and widening target sets." Pending — not graded this cycle: no dataset of politically motivated killings for 2025-09 to 2026-09 was checked in this run; grade at next review against a count, not a vibe.
• NEW (2026-03-23) AI coding and developers: "AI does not merely substitute for developers. It lowers the cost of turning an idea into working software"; the outcome over "five to ten years" is "multiplication with redistribution … Whether that is good news depends on which developer you are." Pending — check 2031–2036: BLS software-developer employment above its 2025 level (multiplication) with entry-level share down (redistribution). The bench's Canaries payroll evidence (entry-level −19% relative in exposed occupations, see Brynjolfsson) already matches the redistribution clause at the hiring margin.
• NEW (2026-04-21) Redistricting: absent a federal rule that "congressional maps may be redrawn only after a decennial census, absent court order," "the trap will therefore continue, and the equilibrium will worsen." Pending — check 2028-11: number of states that redrew congressional maps mid-decade since Texas in 2025, and whether any such federal statute passed. His Virginia datum: a proposed shift from 6–5 to "ten Democrats and a single Republican" in a "53-47" state.
• NEW (2026-05-14) Liquid democracy will grow from pledge-candidates in local offices toward "fractional voting" in legislatures. Unfalsifiable as stated (no date). Proxy check offered: any US candidate for a local office pledging to a liquid-democracy platform by 2028-11.

Revealed behavior (Does)

• Time went to the microphone, not to Game B institution-building. 346 numbered episodes in under seven years at weekly cadence; in his last five months, after the tumor, he re-pointed the show inward into a "Worldviews" series (EP 329 Krakauer 2026-01-15 through EP 345 Yunkaporta 2026-05-29, including his own EP 332, 2026-02-17) and wrote six essays in his final eight weeks. The U.S. Transhumanist Party's notice (2026-06-08) says he recorded an episode the day he died and considered it "one of his best."
• Money went to small grants and other people's institutions. A microgrants budget for the "liminal web," anonymous help to individuals (Laurenson, 2026-06-03); executive-producer money into the Game~B film; a share of The WELL (2012); a makerspace (2014); the chair of a machine-consciousness nonprofit (2025). No public venture portfolio; "JPR Ventures" appears only as a label on his SFI profile.
• The institution he actually founded in the window was for free speech, not Game B — MFSA (2021–), which he chaired to ~1,200 members. Graded by where time went, his priorities were podcast, free speech, machine consciousness, then Game B.
• Signed the 2023 CAIS statement that "Mitigating the risk of extinction from AI should be a global priority" and said so on air (EP 327) — then declined to endorse the Soares/Yudkowsky treaty and offered the power-supply proposal instead.
• Practised what he prescribed on attention: "every year from July 1st to January 2nd, I'm completely off Facebook and Twitter," six years running (2025-10-09).
• Turned his own lens on his own side. The Thiel essay (2025-10-15) files the Antichrist lectures as "adaptive self-deception in a system experiencing phase transition" and "second-order warfare on the epistemic commons"; EP 264 (2024-10-13) is titled "Bret Weinstein and Jim Argue Politics."

Feels

Wanted it to be true that humanity is the universe's one shot and can pass the test — "bring the universe to life" — and feared the Great Filter is ahead of us, not behind. Facing death: "All things considered, I've had a fine life," and he would "play this game another time" (EP 332, 2026-02-17); of the dead, "He still has impact in the universe downstream from his biological life" (EP 342, 2026-05-06).

Hears

SFI's complexity scientists (DeDeo, Krakauer, Farmer, Lloyd); the Game B circle (Hall, Schmachtenberger, Bonnitta Roy, Zak Stein); Alexander Bard, Goertzel, Bach, Boeree, Forrest Landry; a Yudkowsky–Hanson takeoff debate he sat in on in 2004; The WELL since 1989; and, at the end, 117 Substacks a day through his own agent.

Sees

What running the internet's naming registry through a mania looks like from the CEO chair; what a complexity institute's board sees before it publishes; and, as the liminal web's quiet funder, who in that scene was doing real work and who was only talking. He described his own place plainly: "humans are the stupidest possible general intelligence" (EP 181, 2023) — the vantage of someone who expected to be surpassed.

Incentive map

Re-derived this cycle. No employer, no lab money, no fund to market — his wealth came from exits two decades old, which made him one of the few AI voices on the bench with no frontier-lab conflict. Three positions still bent the record: (1) CIMC chairmanship — a stake in "consciousness is computational" and in machine-consciousness research being fundable, which sits behind his "analogous, not identical" line and his AGI-as-third-transition framing; discount those moderately. (2) Game B reputation — fifteen years as the framework's public face made "it failed as a movement" unsayable; the reframe to "a search," "a memetic tag," and "coherent pluralism as a technique" is what that constraint looks like from outside. (3) MFSA — a free-speech political commitment that colors his platform-moderation writing (Quillette, 2022). Podcast economics were small (Substack "over 1,000 subscribers"; paid tier terms not retrieved). Spouse check: no public role found for Celia Rutt. Rule-25 discount: low on AI-lab questions, moderate on machine-consciousness and Game-B-success claims. Now fixed: the record can no longer be bent by anything after 2026-05-27.

Theories aligned with

• Game A vs. Game B — his framework; the spine of this persona (theory file is a June-2026 stub; it still lists him as living)
• Postmodern Agency — Game B "sovereignty" maps directly onto individual self-directed agency without institutional scaffolding; his late move to delegable proxies and networked agents is the same idea with a lower entry bar
• The Man Handed Over the Keys — AI built by Game A institutions may be the tool that breaks Game A; "the AI is already out of control of our genome via the dating apps" (EP 284) is his version
• AI Alignment / X-Risk — coordination-failure variant; shares the race-dynamics diagnosis, locates threat in multipolar traps more than single-agent misalignment; the nine-risk taxonomy (EP 231) is the fullest statement
• Conditional/critical relationship to Techno-Optimism — optimistic about a designed successor, not about Game A growth continuing
• NEW: Jevons Paradox — his 2026-03-23 essay is a direct application to software labor ("multiplication with redistribution")
• NEW: Post-Political Progressive — Liquid Democracy: Governance After Red vs. Blue (2026-05-14) is a governance design that routes around the party binary

What he was reacting against

• Naive techno-optimism (Andreessen/Diamandis register) — rejects "growth + markets will sort it out"; argues exponentials on a finite substrate are precisely the failure mode
• Naive doomerism — agrees civilizational risk is real, but rejects fatalism and single-agent "treacherous turn" as the whole story; locates risk in coordination failure. In-window: took Soares seriously ("I take this issue with the utmost seriousness, and we all should") and still declined the treaty
• Revolution-as-solution — argues toppling-and-replacing is still Game A logic (win-lose); the transition has to be an emergence, not a coup
• Reform-from-within optimism — skeptical that captured Game A institutions can self-correct at the needed speed
• Reductive AI debates that ignore the systems layer — capability benchmarks miss the multipolar-trap dynamics that actually drive outcomes
• NEW (2024–26): theological and supernatural framings entering the tech-politics debate — Hall's turn to Christianity (EP 255 "Is God Real?", 2024-08-20; EP 284; EP 342) and Thiel's Antichrist lectures, which he called "non-falsifiable, binary narratives" that "obliterate the middle ground where governance of complex systems must occur" (2025-10-15)
• NEW (2025-10): the "Foom" story told without a survival plan — his critique cut both ways: doomers had no plan for FOOM, and neither did he
• NEW (2026-04): mid-decade partisan redistricting as a live Moloch case in his own state

Where he overlaps / splits (with Rich)

• Overlaps with Jordan Hall (co-originator) on the whole Game A/B frame; splits on register — Rutt the complexity-science / empirical-systems voice, Hall the deep-sovereignty / collective-sense-making philosopher. v1 TBD (Hall persona in backlog) resolved — file exists. UPDATE: the split is now metaphysical, not just tonal — Hall's Christian turn versus Rutt's "strong disbeliever in supernatural anything"; when Hall said Rutt had described "a foundation," Rutt answered that his mechanisms are "provisional and subject to empirical verification" (EP 284, 2025-02-13). Their last conversation was EP 342 (2026-05-06), three weeks before his death.
• Overlaps with Aza Raskin on Moloch / race-to-the-bottom incentives; splits on scope — Raskin targets attention-economy + tech-deployment incentives, Rutt the whole civilizational operating system
• Overlaps with Vitalik Buterin on distributed coordination as a safety property; splits on mechanism — Buterin reaches for crypto/mechanism-design, Rutt for cultural / collective-intelligence structures (liquid democracy is his closest approach to mechanism design)
• Overlaps with Eliezer Yudkowsky / Bengio on race dynamics being dangerous; splits on threat model — coordination failure / gradual systemic collapse over treacherous-turn ASI. UPDATE: hosted MIRI's Soares on the book (EP 327, 2025-10-15), conceded the seriousness, rejected the opacity emphasis and the treaty, and offered power-supply governance as the practical substitute — the bench's one named alternative to the pause/treaty program
• Overlaps with Yanis Varoufakis that Game A could capture AI (techno-feudalism); splits on exit — Rutt sees AI as a possible escape route from Game A, Varoufakis sees deeper capture; Rutt's own EP 284 list of AI paths includes "neo-feudalism" and "AI singletons" as the bad branches
• Splits with Andreessen / Diamandis on whether more growth + markets is the answer (Rutt: that's the disease, not the cure)
• Overlaps with Thiel / Cowen on the diagnosis that Game A institutions are stagnating/decaying; splits on remedy (founder-monopoly vs. civilizational redesign). UPDATE: he wrote the bench's most direct outside reading of the Antichrist lectures — Thiel "likely believes his own narrative while remaining systematically biased by material incentives" (2025-10-15)
• NEW — Overlaps with Cory Doctorow on building "sovereign alternatives to Big Tech" (EP 346, 2026-06-03, his posthumous final episode) — Doctorow's exit-and-build is Rutt's "membranes" in different vocabulary
• NEW — Between Brynjolfsson and Frey on AI and labor: "multiplication with redistribution" (2026-03-23) predicts more developers overall and a worse deal for entry-level ones — a third position the two bench files do not yet name
• NEW — With Rich: the theory file asks whether the cos- bench / Agency Lab is a Game B structure. Rutt's last build — a personal agent digesting 117 Substacks, meant to network with other people's agents — is the closest thing on the bench to Rich's own stack, and he reached it from the sovereignty side rather than the tooling side. The liquid-democracy essay is also the only Game B document with a dollar figure and an entry path, which makes it the one Rich could actually test against §-level principles.

Track record

• Built and ran a real complex system at scale (Network Solutions during the dot-com domain explosion) — operator credibility, not just theory. Dates resolved: named CEO May 1999, deal announced 2000-03-07 at $21B in stock, closed 2000-06-08, stepped down February 2001 (Wikipedia/WaPo; 8-K). The Whirl tribute recalls the NYT calling him "the Internet's bad boy."
• Chaired the Santa Fe Institute board — genuine standing in complexity science, the field he applies. v1 TBD (exact tenure) resolved: trustee 2003–2022, chair 2009–2012, researcher in residence 2002–2004, Trustee Emeritus and Distinguished Fellow at death (SFI)
• Co-originated and popularized "Game B" as a named framework now used across an online intellectual community — memetic influence real, institutional adoption near-zero. Held, now with evidence rather than a question: wiki dormant since 2021-08-29, no funded institution, film undated. Origin date corrected: "co-founded the Game-B Network" in 2012 (Fralin bio), first dated talk "Minimum Necessary Radicalism" 2013-04-27, In Search of the 5th Attractor at The Feast 2014-11-26 (his Works page) — v1's "~2013–2016" was a year late
• The Jim Rutt Show has run a long, substantive interview series across complexity, AI, and governance — high-bandwidth synthesizer, low direct policy footprint. v1 TBD (launch, count) resolved: first episode July 2019 with Simon DeDeo; 346 numbered episodes plus Currents shorts (461 posts); the final episode was published a week after his death
• NEW: earlier ventures — Investext (Business Research Corp., co-founded with William A. Benjamin) and First Call, two financial-information products that outlived him; first CTO of Thomson; chairman of Analog Design Automation through its 2004 sale to Synopsys; creator of the Network Wars mobile game (SFI profile)
• NEW: institution-builder for other people's fields — SFI (complexity), MFSA (free speech, 2021), CIMC (machine consciousness, 2025) — and never for his own framework; that asymmetry is the most informative fact in the file
• NEW: public-money steward — New Mexico State Investment Council, appointed 2006, a body managing a "$20.5 billion permanent endowment" (Fralin bio)
• Pattern: exceptional operator and convener; rigorous where a model was available; honest about the gap where it was not ("I've never seen a convincing story on how to preserve humanity in the case of a FOOM"); the one thing he never produced was a falsifiable Game B

Empirical vs normative

• Empirical: multipolar traps produce collectively bad outcomes; civilization behaves as a complex adaptive system; AI intensifies race dynamics; transformers are a low-danger architecture; AI coding multiplies developers while redistributing the gains; mid-decade redistricting ratchets without a federal rule
• Normative: Game A should be succeeded; coherent pluralism and collective intelligence are the right design primitives (sovereignty demoted from prerequisite to one input); revolution is the wrong path; civilization-scale technologies need an "existential safety case" with "public adversarial review" before deployment (2026-05-04)
• Speculative: Game B can emerge and outcompete Game A; AI can be turned to cooperative coordination; humans may be the universe's only general intelligence
• Note (v1, carried): "Game A is failing" blends a falsifiable claim with a values stance — keep them separate per CONVENTIONS rule 10; the framework is a direction with little operational specification, so weight as research lens more than forecast. UPDATE: the operational specification improved in 2025–26 (membranes, pluralism, liquid democracy with a budget) — weight the late essays higher than the 2019 podcast material
• Conflict note (CONVENTIONS rule 25): no lab or fund conflict; moderate discount on machine-consciousness claims (CIMC chair) and on Game-B-success claims (reputation)

Weak spots / open questions

• Game B is under-specified — "a direction, not a plan" is honest but makes it nearly unfalsifiable; what concrete institution counts as Game B succeeding? (CONVENTIONS rule 23) Re-graded: partially addressed — mechanisms now named (membranes, coherent pluralism, multi-dimensional economics, liquid democracy) — but he never stated a success criterion; the $1M/yr-organization check above is the bench's substitute. Open, and now permanently his
• The "Game A is hitting its limits" thesis has been asserted for decades; what specifically makes the AI moment different from prior deferred-collapse predictions? (rule 23) Re-graded: he answered the mechanism (AI compresses the runway, "60 years" into "30") and the framing (a test ahead, not behind) but never the date. Open
• Sovereignty-as-prerequisite has an "average is over" problem — it requires high individual capacity not everyone has; risks an elite-exit framing dressed as universal. Re-graded: partially answered by liquid democracy's delegation — a voter with no capacity delegates a revocable proxy — and by his own demotion of sovereignty to one input among many. Narrowed
• The optimistic claim (AI enables Game B coordination) and pessimistic claim (AI accelerates Game A) coexist — is the dividing line stable, or does it dissolve under capability + concentration scaling? Re-graded: he resolved it by scale — frontier-lab AI "is clearly accelerating Game A," personal "intimate AI" agents are the escape — and the line is a design choice, not a property of the model. Whether it survives concentration is untested. Stable as stated, unproven
• Heavy reliance on complexity-science vocabulary can substitute metaphor for mechanism; where are the testable models vs. the evocative analogies? (rule 21) Re-graded: the Rᵥ branching model and the redistricting ratchet are real, checkable mechanisms; Game B itself still is not. Half answered
• Memetic reach is real but the framework has near-zero institutional uptake — is that "too early" or evidence the design doesn't survive contact with Game A incentives? Re-graded: fourteen years, a dormant wiki, and a founder who built institutions for three other fields but not this one — the "too early" reading has run out of road; log as evidence, not question. Answered against the framework
• NEW: the CIMC "Chairman Emeritus" wording is undated — did he step back after the November-2025 diagnosis, or is it posthumous? The institute's own posts (white paper 2026-01-30; MC0001 assembly 2026-02/05) name no board
• NEW: the obituary says an essay collection was planned — the record may still grow; a posthumous volume is the one wake trigger that could add positions
• NEW: "JPR Ventures" and his investing are opaque; the name "Jim Rutt" collides in databases with a Dana Foundation CIO (New York) and a Palomar Ventures GP, and "Rutt Bridges" is a different person — PitchBook's "Jim Rutt" profile (VideoStorm, Solido Design Automation) could not be attributed and is not carried in this file
• NEW: the last essay's "existential safety case" with "public adversarial review" has no proposed reviewer, venue, or trigger — a proposal without a mechanism from the man who insisted on mechanisms

Rich's take

• (your synthesis here)

Delta log

2026-09-21 — v1→v2 migration + validation (bulk run)

• Grades: 10 Held / 4 Drifted / 0 Wrong (3 unfalsifiable items marked "rewrite or retire" or noted; 5 new dated checks pending, not yet due). Nothing graded Wrong on his side; the Wrong grade belongs to the file itself — see status line.
• Status change — the single largest delta: died 2026-05-27 (Obaugh Funeral Home obituary; SFI In Memoriam, 2026-07-06; MFSA tribute, 2026-09-05, which dates the tumor disclosure to November 2025). The v1 page's 2026-06-09 review post-dates the death and the obituary and changed nothing — the May/June-2026 pass is now confirmed defective on a persona who was no longer alive.
• Held: multipolar traps as mechanism (Moloch Comes to Virginia, 2026-04-21; Kirk essay, 2025-09-11); AI accelerates Game A (EP 284, 2025-02-13; EP 231, 2024-03-27); AI as Game B enabler (untested; his post, 2025-10-09); collective intelligence over consensus (Liquid Democracy, 2026-05-14); complexity lens and epistemic humility (EP 332, 2026-02-17); metacrisis-as-coordination, recast as Great Filter (The Silent Sky, 2026-05-04); Game-A-limits on endorsement only; transformers low-danger (so far) and no-clean-winner (live), both from EP 327, 2025-10-15; "memetic reach, zero institutions" track-record line, now evidenced (gameb.wiki last edit 2021-08-29; game-b.org; Laurenson).
• Drifted: "direction not a plan" (specified further, community dormant); sovereignty-as-prerequisite (replaced by coherent pluralism and delegation — EP 284; Values essay, 2026-04-01; Liquid Democracy); Game-B-emerges-inside-Game-A (no structure after 14 years); "overindex on opacity" (bench Olah file, 2026-02 ground-truth checks).
• Unfalsifiable, marked: Game-A-limits outcome (rewrite offered); only-general-intelligence / fine-tuned-for-emergence (value frame); liquid-democracy adoption (no date; proxy offered).
• New dated checks attached: autonomous tanks by 2033-03 (EP 181, 2023-03-23); AGI 2034–2044 and the 3–5-year camp by 2027–29 (EP 231); political-violence Rᵥ, grade at next review with a count; developer employment 2031–36 (Jevons essay, 2026-03-23); mid-decade redistricting count and federal statute by 2028-11.
• Employer/role check (mandatory), run as a dated news search: three affiliations v1 never had — CIMC chair (CIMC, 2025-06-24; emeritus per Deep Tech Week, undated), MFSA co-founder/first chair (2021–), Fralin advisory board; SFI tenure, NSI dates, VeriSign figures (8-K, 2000-06-19) and podcast launch/count all resolved from v1 TBDs. Game B origin re-dated to 2012 (Fralin bio) from v1's "~2013–2016".
• Incentive map re-derived: no lab/fund conflict; CIMC and Game-B-reputation discounts added; MFSA political commitment logged; no spouse conflict found.
• Identity guard: a Dana Foundation CIO "Jim Rutt" (confirmed distinct via the Vation council bio), a Palomar Ventures "Jim Rutt," and the "Rutt Bridges" fund all surfaced in investor searches and were excluded.
• Not found this cycle: his own November-2025 health post (Substack archive shows a two-month silence instead); the 2025-10-19 Win-Win transcript (only his own summary); EP 8's Game B origin account (site returned 503 twice); the A Journey to GameB text (Medium and mirrors blocked — cited by title, date and his own announcement only).
• Tier: annual — the record is closed; only a posthumous collection, a CIMC leadership change, or a due-date on his checks can move it. Confidence raised to medium-high: every affiliation and number now sits on a primary; the thinness is in outcomes, not sourcing.

Sources

• Death and memorials (2026): Obaugh Funeral Home — obituary (d. 2026-05-27) · Santa Fe Institute — In Memoriam (2026-07-06) · MIT Free Speech Alliance — Tribute (2026-09-05) · Lydia Laurenson — RIP Jim Rutt, philanthropy and culture-building (2026-06-03) · The Whirl — Remembering Jim Rutt (2026-06-03) · U.S. Transhumanist Party notice (2026-06-08) · Fralin Biomedical Research Institute — advisory board, in memoriam
• His own final essays (Substack, 2025–26): Will Kirk's Assassination Trigger an Escalating Spiral of Violence? (2025-09-11) · Me on Liv Boeree's Win-Win podcast (2025-10-09) · Psyop, Insanity, or…? Peter Thiel, the Antichrist, and Our Collapsing Epistemic Commons (2025-10-15) · Jevons' Paradox and the Fate of Software Developers in the Age of AI Coding (2026-03-23) · Values Are Neither Universal Nor Arbitrary (2026-04-01) · Moloch Comes to Virginia (2026-04-21) · The Silent Sky and the Test Ahead (2026-05-04) · Liquid Democracy: Governance After Red vs. Blue (2026-05-14)
• Podcast episodes cited (his own words on air): EP 181 Forrest Landry on AI Risk (2023-03-23) · EP 231 Vance Crowe Interviews Jim Rutt on AI Risk (2024-03-27) · EP 284 Jordan Hall on AI, the Commons, and the Church (2025-02-13) · EP 327 Nate Soares on Why Superhuman AI Would Kill Us All (2025-10-15) · EP 332 Worldviews: Jim Rutt, interviewed by Brendan Graham Dempsey (2026-02-17) · EP 342 Worldviews: Jordan Hall (2026-05-06) · EP 346 Cory Doctorow (2026-06-03, posthumous) · Substack archive (episode dates; read 2026-09-21) · Rephonic — show metrics
• Bio, works, institutions: SFI profile (Trustee Emeritus, Distinguished Fellow; ADA; JPR Ventures) · Works page (essays, talks, films, MFSA, The WELL) · Podcast appearances (through 2025-10) · CIMC — The Beginning of the California Institute for Machine Consciousness (2025-06-24) · Deep Tech Week — CIMC listing (undated; "Board Chairman Emeritus") · Win-Win #48 episode page (2025-10-19) · Quillette author page (Musk and Moderation, 2022-04-27) · Wikipedia — Jim Rutt
• Network Solutions / VeriSign (primary): VeriSign Form 8-K (filed 2000-06-19): acquisition completed 2000-06-08, ~78M shares, 1.075 ratio · Wikipedia — Network Solutions (CEO May 1999–Feb 2001, citing WaPo 2001-02-27; $21B, citing ZDNet 2000-03-07) · EE Times — Synopsys buys Analog Design Automation (2004)
• Game B community status (read 2026-09-21): Game B wiki (main page last edited 2021-08-29) · game-b.org · Game~B film (executive producer; no release date) · His announcement of A Journey to GameB (2020-01-13) · A Journey to GameB (Medium, 2020 — not retrievable this run)
• Bench cross-references: Jordan Hall (v1, reviewed 2026-06-10 — still lists Rutt as living) · Chris Olah (interpretability ground-truth checks, 2026-02) · Erik Brynjolfsson / Carl Benedikt Frey (Canaries payroll data) · Game A vs. Game B theory stub
• v1 sources carried: The Jim Rutt Show (podcast tier); early Game B essays/dialogues with Jordan Hall and collaborators (video/forum tier); GameB community materials (forum tier — low, per CONVENTIONS rule 4); SFI board role; Network Solutions tenure — all now dated and linked above

migrated v1→v2 2026-09-21 · backup: _archives/jim-rutt.html.bak-20260921 · converts-from: people/jim-rutt.md · AI & Society domain