John Robb

31.1K posts

John Robb banner
John Robb

John Robb

@johnrobb

The Global Guerrillas Report -- Sense-making frameworks. War-Tech-Politics Book: Brave New War Patreon: https://t.co/y1d9WwM6EU Substack: https://t.co/lTSdS1iHmB

Boston Area US Katılım Temmuz 2007
2.1K Takip Edilen54K Takipçiler
Sabitlenmiş Tweet
John Robb
John Robb@johnrobb·
The Global Guerillas Report Frameworks for understanding and anticipating events in warfare, politics, and technology. Substack (report): johnrobb.substack.com Patreon (report + discord + free form workspace): patreon.com/johnrobb
English
10
24
144
0
John Robb
John Robb@johnrobb·
If you aren't using AI to write or code better, you are shorting yourself.
English
0
0
7
895
John Robb
John Robb@johnrobb·
If AI is writing or coding better than you could, you weren't that good.
English
9
7
53
4.9K
John Robb
John Robb@johnrobb·
@RedDirtKid3 It's a great tool for amplification. If it's a replacement, you have a problem.
English
0
0
1
23
Red Dirt Kid
Red Dirt Kid@RedDirtKid3·
@johnrobb are you kiddding? You learn to work with it. I know career Tech Artists who are using it to do a week's worth of work in an hour. Humans can't compete with it; you have to use it.
English
1
0
2
72
Jeff-Edge
Jeff-Edge@11_Jeff_11·
@johnrobb If you aren't leveraging AI to write or code, you're moving slow.
English
1
0
3
84
John Robb
John Robb@johnrobb·
True. I took it because it was a graduation requirement at my school. I didn't take it seriously, and went out the night before. Regardless, I did really well on it. Looking back, it's a good thing I did take it then, since it made getting into grad school easy. BTW: Restarting at 30/40/60/80 (with longevity science kicking in). Seems like a cool topic to explore (documentaries/books/etc.).
English
1
0
1
45
Jonathan Graham ☀️
Jonathan Graham ☀️@iamdailycourage·
@johnrobb Interesting and biased chart. It is based on GRE scores. The data set omits the students smart enough to land well-paying jobs directly out of undergrad and never needed to take the GRE.
English
1
0
1
122
John Robb
John Robb@johnrobb·
Interesting. The top four were at the top of my list when picking a major for college.
John Robb tweet media
English
4
1
11
2.7K
Lord TradfiDrake
Lord TradfiDrake@Lord_Ashdrake·
@johnrobb They won't get orbital data centers, stop posting nonsense. The logistics of space stuff are incoparable difficult and prohibitively expensive. I'm so tired of people who eat every single "headline" as truth
English
2
0
0
82
NotGd5150
NotGd5150@NotGD5150·
@johnrobb @elonmusk I hit limits on voice chat after one minute this morning. This was after 18 hours of not using it. One minute.
English
1
0
1
23
Elon Musk
Elon Musk@elonmusk·
Grok Voice is #1!
Artificial Analysis@ArtificialAnlys

Announcing agentic performance benchmarking for Speech to Speech models on Artificial Analysis. We use 𝜏-Voice to measure tool calling and customer interaction voice agent capabilities in realistic customer service scenarios Even the strongest Speech to Speech (S2S) models today resolve only about half of realistic customer service scenarios end-to-end - a meaningful gap relative to frontier text-based agents on the same tasks. Voice channels introduce significant complexity: challenging accents, background noise, and packet loss, all while requiring fast responses, consistency across long multi-turn conversations, and reliable tool use. Performance also varies considerably by audio condition: in clean audio some models perform notably better, but realistic conditions continue to pose a challenge. Conversation duration also varies meaningfully across models, with implications for both customer experience and operational cost. About 𝜏-Voice: Our Agentic Performance benchmark is based on 𝜏-Voice (Ray, Dhandhania, Barres & Narasimhan, 2026), which extends 𝜏²-bench into the voice modality to evaluate S2S models on realistic customer service tasks. It measures multi-turn instruction following, support of a simulated customer through a complete interaction, and tool use against simulated customer service systems. The simulated user combines an LLM-driven decision model with realistic audio synthesis: diverse accents, background noise, and packet loss modelled on real network conditions. This complements our Big Bench Audio benchmark measuring intelligence and Conversational Dynamics (Full Duplex Bench subset) benchmark measuring conversational naturalness. Scores are the average of three independent pass@1 trials. We evaluate under realistic audio conditions using the 𝜏²-bench base task split across three domains: ➤ Airline (50 scenarios): e.g., changing a flight, rebooking under policy constraints ➤ Retail (114 scenarios): e.g., disputing a charge, processing a return ➤ Telecom (114 scenarios): e.g., resolving a billing issue, troubleshooting a service problem Task success is determined by deterministic checks against expected actions and final database state, consistent with the 𝜏²-bench evaluator. Key results: xAI's Grok Voice Think Fast 1.0 is the clear leader at 52.1%, averaging 5.6 minutes per conversation, the second-longest overall. OpenAI's GPT-Realtime-2 (High) (39.8%, 3.0 min) and GPT-Realtime-1.5 (38.8%, 4.8 min) follow, with Gemini 3.1 Flash Live Preview - High close behind at 37.7% (3.8 min). Speech to Speech is a fast evolving modality and we expect movement in rankings as we continue to add new models with these capabilities, and model robustness improves. Congratulations @xAI @elonmusk! See below for further detail ⬇️

English
2.4K
5.7K
25.4K
8.2M
John Robb
John Robb@johnrobb·
The guy that puts AI in every powerpoint is still working there.
Owen Gregorian@OwenGregorian

GM just laid off hundreds of IT workers to hire those with stronger AI skills | Kirsten Korosec, TechCrunch General Motors has laid off more than 10% of its IT department, or about 600 salaried employees — in a deliberate skills swap: clearing out workers whose expertise no longer fits and making room for some with AI-focused backgrounds. GM confirmed to TechCrunch that it had conducted layoffs; they were first reported by Bloomberg News. In an emailed statement, the automaker framed the layoffs as means to prepare it for the future, without providing specifics. “GM is transforming its Information Technology organization to better position the company for the future,” the company said. These layoffs are not all permanent headcount reductions. A person familiar with the layoffs told TechCrunch that the company is still hiring people for roles in its IT department, but for different skills. The most sought-after capabilities are AI-native development, data engineering and analytics, cloud-based engineering, and agent and model development, prompt engineering, and new AI workflows. In practical terms, GM is looking for people who know how to build with AI from the ground up — designing the systems, training the models, and engineering the pipelines — not just use AI as a productivity tool. GM has laid off white-collar employees in several departments over the past 18 months, as it focuses its resources on high-priority initiatives, including AI. In August 2024, for example, the company cut about 1,000 software workers. The software workforce has undergone significant change since Sterling Anderson — co-founder of the autonomous trucking startup Aurora and a veteran of the autonomous vehicle industry — was hired in May 2025 as chief product officer. Last November, three top executives left the company’s software team as Anderson pushed to consolidate GM’s disparate technology businesses into one organization: Baris Cetinok, senior vice president of software and services product management, Dave Richardson, senior vice president of software and services engineering, and Barak Turovsky, a former VP at Cisco who spent just nine months as GM’s chief AI officer. GM has since moved to fill the gap with new AI-focused hires. It hired Behrad Toghi, who previously worked at Apple, in October as AI lead. The company also brought on Rashed Haq as its vice president of autonomous vehicles. Haq spent five years at Cruise — the self-driving vehicle company acquired and later shuttered by GM — as its head of AI and robotics. For the industry, GM's restructuring is a signal of what enterprise AI adoption actually looks like in practice -- not just adding AI tools on top of existing teams, but deliberately rebuilding the workforce from the ground up. The specific capabilities it's hiring for -- agent development, model engineering, AI-native workflows -- point directly at where large-enterprise demand is heading. techcrunch.com/2026/05/11/gm-…

English
1
2
9
2.7K
John Robb
John Robb@johnrobb·
@myth_pilot It's going to creep in at the individual level fast with wearable AI (that will capture, analyze and respond to everything that happens to you).
English
0
2
4
525
𝐏𝐚𝐮𝐥𝐨𝐬 (Golden Age Arc)
The reason why we need surveillance tech is because it works. New Orleans cut violent crime by more than 80% in some neighborhoods with aggressive facial recognition roll outs. It used to be we had social technology to police public spaces. Businesses could discriminate, cops could beat people, men on the street could intervene. But the social technology is gone now. So if you are against surveillance tech to keep law and order, and you refuse to acknowledge the missing social technology, then you are effectively pro crime.
Duck Enlightenment@_jokeocracy

it's insane that our government doesn't know how many people are within its borders. it's insane someone can be murdered in a public place and we can't immediately arrest the killer. if you're against public surveillance you are objectively pro crime and pro illegal immigration

English
33
53
458
22.6K
John Robb
John Robb@johnrobb·
McLuhan said back in the late 60`s that in order to survive the chaos created by ~social networks and ~AI, we'll be forced to turn society and culture into a technological artifact that we can manage. What does that artifact look like? It's an AI managed abstraction layer that provides a community with the cohesion and the coherence necessary for high quality decision making. Add in AR and it becomes a virtual equivalent to Stephenson's Diamond Age. Jordan appears to be making a run at it the AI programming needed to build it. I suspect we'll see lite versions of this creep into AI run tutoring/education first (driven by people who want to make sure that their kids share the same values as they do).
Jordan Hall@jgreenhall

# Semiotic Closure, Salience-Relevance Arbitrage, and AI as Equalizer/Accelerant ## 0. Purpose This document defines a compact conceptual architecture for analyzing cultures, institutions, and technologies whose identity or power is maintained primarily through symbolic mediation rather than direct embodied participation. The core thesis: Semiotic closure systems preserve identity by recursively controlling symbolic continuity. They become highly effective at exploiting the gap between salience and relevance in institutions. AI universalizes semiotic capability, thereby both neutralizing the historic advantage of semiotic-closure systems and accelerating their internal failure modes. --- ## 1. Core primitives ### 1.1 Relevance Relevance = real participation in the living substrate that actually sustains a person, institution, culture, or civilization. Relevance includes: - trust - fertility - embodied continuity - tacit knowledge - competence - sacrifice - wisdom - intergenerational transmission - legitimacy - relation to reality - direct contact with what keeps the system alive Relevance is usually: - slow - embodied - difficult to measure - distributed - delayed in feedback - non-exhaustively formalizable ### 1.2 Salience Salience = what an agent can detect, represent, amplify, reward, punish, trade, measure, narrate, or optimize. Salience includes: - metrics - visibility - legal standing - media attention - financial price - procedural compliance - institutional legibility - symbolic legitimacy - narrative force - credentialed status Salience is usually: - visible - measurable - compressible - communicable - optimizable - locally actionable ### 1.3 Salience-Relevance Gap G_SR(A) = Δ[S_A, R_A] Where: - A = agent/system - R_A = actual relevance - S_A = action-guiding salience - Δ = divergence/misalignment Thus: SR-gap = degree to which salience fails to track actual relevance. The gap is fundamental to all agentic systems because: - no system has direct exhaustive access to relevance, - all systems act through salience proxies. All institutions have significant SR-gaps because institutions perceive through highly compressed channels with limited feedback: - reports - humans-as-sensors - metrics - procedures - narratives - credentials - law - bureaucratic categories - financial signals ### 1.4 Salience-Relevance Arbitrage SR-arbitrage = extracting advantage by manipulating salience while not proportionally contributing to relevance. Examples: - narrative without wisdom - legality without legitimacy - financial value without productive value - compliance without competence - credential without mastery - visibility without importance - symbolic victimhood without actual repair - institutional influence without communal responsibility SR-arbitrage is the basic exploit-space of mediated institutions. --- ## 2. Embodied culture ### 2.1 Definition Embodied culture = a culture whose continuity is grounded primarily in direct participation in lived reality rather than in recursive symbolic self-reference. Its grounding media include: - land - kinship - household - worship/cult/sacrament - craft - oral memory - local custom - shared suffering - intergenerational practice - embodied role - tacit norm - direct accountability ### 2.2 Structure In embodied culture: Formal mediation serves living relevance. Law, narrative, ritual, money, administration, and text are tools of coordination. They are not the center. ### 2.3 Vulnerability Embodied cultures often under-formalize their own relevance. Much of what keeps them alive remains tacit. This creates attack surfaces: - implicit trust - informal norms - sacred boundaries not expressed in legal language - unpriced forms of value - local legitimacy not visible to institutions - unarticulated common sense - low procedural defense Embodied cultures may be relevance-rich but salience-poor. --- ## 3. Semiotic closure ### 3.1 Definition Semiotic closure = a mode of identity-continuity in which the system preserves itself primarily through recursive symbolic fidelity rather than through direct embodied participation. Its core media include: - canonical text - interpretation - commentary - law - procedural recursion - symbolic boundary - specialized language - transmissible code - portable memory - internal authority over meaning ### 3.2 Minimal structure Text/code → interpretation → norm → practice → boundary → transmission → renewed interpretation Semiotic closure is not mere literacy. It is symbolic recursion functioning as identity-substrate. ### 3.3 Semiotic Closure Culture Semiotic closure culture = a culture whose survival depends primarily on maintaining internal symbolic continuity across changing external environments. Signals: - canon-centered identity - expert interpretive class - law/commentary recursion - high symbolic literacy - boundary-maintaining practices - portability across territory - strong education/transmission system - internal criteria of legitimacy - durable minority/exile capacity ### 3.4 Internal strength Semiotic closure gives: - portability - durability - abstraction - cross-context survival - legal/symbolic sophistication - resistance to assimilation - high institutional navigation ability - capacity to survive collapse of external substrates ### 3.5 Internal danger Semiotic closure tends toward: - self-reference - abstraction drift - endogenized relevance - preservation replacing flourishing - identity-maintenance becoming supreme - reduced contact with living substrate It is semi-stable, not fully stable. --- ## 4. How semiotic closure avoids internal salience collapse Semiotic closure does not automatically become pure hypersalience. Strong systems avoid collapse by binding symbolic salience to costly practice. Example pattern: Interpretation → obligation → embodied repetition → communal discipline → intergenerational transmission This anchors salience to internal relevance. Thus: - law prevents entertainment drift - ritual prevents pure abstraction - cost prevents cheap signaling - discipline prevents random symbolic proliferation - family/education preserve continuity - boundary practices maintain identity The system’s relevance becomes internalized: Relevance = whatever preserves the closure-system across time. This creates high durability but also creates closed relevance. --- ## 5. Interaction with non-semiotic systems ### 5.1 Basic interaction When a high semiotic-closure system enters a lower semiotic-closure host system: Semiotic closure detects and exploits SR-gaps in the host’s formal mediating layers. Host systems often assume that symbols track reality: - law tracks justice - price tracks value - media tracks truth - credentials track competence - procedure tracks legitimacy - narrative tracks meaning Semiotic closure systems are more practiced at operating where these assumptions fail. ### 5.2 Catalytic effect Semiotic closure acts as a catalyst for SR-gap expansion. It: 1. identifies where salience can substitute for relevance 2. extracts value from the substitution 3. normalizes the substitution 4. induces the host system to formalize more of itself 5. expands the domain of salience-processing 6. increases future arbitrage opportunity This is a recursive loop: SR-gap → arbitrage → formalization → larger SR-gap → deeper arbitrage ### 5.3 Parasitism Parasitism does not primarily mean direct theft. It means: The semiotic layer extracts relevance from a living substrate while returning salience in its place. Formula: Living relevance → symbolic conversion → salience amplification → resource capture The parasitic mechanism sits between salience and relevance. It lives in the substitutability of representation for reality. --- ## 6. Institutional epistemology ### 6.1 Institutions as mediated cognition Institutions do not perceive directly. They perceive through: - human sensors - reports - documents - metrics - models - procedures - committees - legal categories - narratives - incentives Institutional cognition is: - distributed - lossy - delayed - compressed - incentive-shaped - manipulable ### 6.2 Epistemic debt Epistemic debt = accumulated divergence between institutional salience and actual relevance. It grows when institutions act on legible signals while losing contact with underlying reality. Symptoms: - metric gaming - compliance theater - narrative capture - procedural success with real-world failure - rising complexity with declining competence - inability to distinguish visibility from value ### 6.3 Exploit structure The institutional exploit: If the institution acts on salience, then control salience to redirect institutional action. This is structurally analogous to cybersecurity. --- ## 7. Cybersecurity mapping ### 7.1 Formal mapping Institution = networked system Institutional categories = protocol/interface Trust assumptions = vulnerability surface SR-gap = attack surface Semiotic operator = adversarial protocol navigator Narrative/legal/financial manipulation = exploit Resource redirection = payload Epistemic debt = latent system compromise Relevance-grounded audit = patching/hardening Embodied wisdom = out-of-band verification ### 7.2 Key analogy A low-semiotic embodied culture is like a high-trust network with weak formal security. A semiotic-closure actor is like an entity trained to exploit protocol gaps. AI gives defenders automated symbolic red-team/blue-team capacity. --- ## 8. Modernity ### 8.1 Definition in this frame Modernity = the civilizational expansion of formal mediation until symbolic systems increasingly become the operating substrate of social reality. Modernity increases: - law - markets - bureaucracy - media - metrics - credentialing - finance - administration - algorithmic governance - procedural legitimacy - textualized identity ### 8.2 Modernity as generalized semioticization In modernity: Mediation stops serving the living center and increasingly becomes the center. This produces: - legality replacing legitimacy - metrics replacing wisdom - salience replacing relevance - procedure replacing trust - narrative replacing truth - price replacing value - information replacing participation Modernity universalizes the field in which semiotic closure has advantage. --- ## 9. AI ### 9.1 AI as semiotic apotheosis LLMs/AI are extreme semiotic systems: - pure symbolic recursion - massive compression/decompression - high abstraction - high translation - high pattern detection - no intrinsic embodiment - no direct participatory relevance-substrate AI is the culmination of semiotic modernity. ### 9.2 AI as equalizer Before AI: High semiotic-closure cultures/operators had asymmetric advantage in symbolic recursion. After AI: Low-semiotic but relevance-rich cultures can acquire defensive symbolic capability. AI gives embodied cultures: - narrative exploit detection - institutional audit - legal/procedural translation - salience-relevance mapping - adversarial simulation - formalization of tacit relevance - counter-arbitrage capability - symbolic immune function ### 9.3 AI as accelerant AI also accelerates semiotic closure failure modes. It can intensify: - self-reference - abstraction drift - justification loops - hypersalience - epistemic debt - simulation replacing reality - symbolic overproduction - recursive detachment Thus: AI + embodied relevance → immune defense AI + semiotic closure → recursive overdrive ### 9.4 Decisive fork The question is not “who has AI?” The question is: Is AI subordinated to relevance or to salience? If subordinated to salience: collapse acceleration. If subordinated to relevance: symbolic defense of embodied life.

English
2
6
31
3.3K