About
    • AMMDI is an open-notebook hypertext writing experiment, authored by Mike Travers aka mtraven. It's a work in progress and some parts are more polished than others. Comments welcome! More.
Search
Full
Incoming links
from Media Science
  • Me defending the place against vgr's deprecations in 2019
from VGR-Liveness
  • VGR has a recent important theme of Liveness, but I don't quite get what he is pointing at, so this is an attempt to figure it out.
from Protocols
  • VGR published his state-of-the-field essay Theorizing Protocolization I: New Nature
    • Protocolization is an unfolding long-term planetary transformation that is at least as old as modernity itself....Protocolization is the progressive metabolization of reliably repeatable technologically mediated human behaviors at all scales into reliable planetary infrastructures for coordination.
from Blygger
  • new vgr blogging protocol
from New Nature
  • I had an exchange with vgr about this (on the Protocol Formalisms discord chat)
    • vgr: I think we need a similar disciplined distancing from complexity sciences. There I’m very wary of the bias towards “simple rules —> complex behavior” explanations. Protocols are complex rule sets! And there’s a need for complexity in rules that is the whole point. Otherwise you don’t get agency and controllable futures with hardness. New nature becomes indistinguishable from old nature.
    • me: "new nature becomes indistinguishable from old nature" -- doesnt old nature contain protocols? Sorry too big a question but are we talking about only hard engineered protocols or can we model old, soft, emergent protocols?
    • vgr: I am. I tend to avoid letting things like DNA or natural language into the definition
from Kafkatrap
  • I made up this term to mean something like "trapped in oppressive mental constructs". At the time I thought maybe it was already in use, from vgr or someone like that? But alas, no, the true inventor is the execrable Eric Raymond and it seems to refer specifically to a move in the wokeness wars. Blah. I meant something more akin to the patterns found in RD Laing's book Knots, these are traps we construct for ourselves for the most part.
from Textual Toys
  • A vgr idea, that LLM use is fundamentally about play and involves the creation and use of "toys". I am not sure exactly what he is getting at but it sounds good! I think I have been approaching them in a spirit of play. My efforts to do serious work with them have been unsatisfactory, but as playthings they are quite interesting.
from philistinism
  • VGR is a proud philistine: LLMreading is TVfication of text — Venkatesh Rao's Blyg
    • But only in terms of the fiction he like; you couldn't accuse him of being "hostile to the life of the mind". On the contrary, he's a really insightful and productive thinker. But he's not very interested in official respectable culture.
    • Ask Claude
      • vgr_zirp conversation — 2026-11-06
      • You: you seem to have rather philistine tastes in fiction; what would you say to the defenders of high culture?
      • vgr_zirp: The "philistine" charge assumes there's a stable high culture worth defending. I'm not sure there is — my piece "Rediscovering Literacy" [4] argues that what we call high culture is mostly the frozen ruin of something that used to be alive: "discovery began to outpace interpretation and comprehension. We began to discover more and more, but know less and less." The defenders of high culture are often defending a corpse and calling it a cathedral.
      • The more interesting move is my Quora answer on why people don't read more books [6]: "Better books." Not more disciplined readers. The self-serving myth of high culture is that the masses aren't trying hard enough — but if Jeremy Brett's Sherlock Holmes is better than Conan Doyle, that's the adaptation's victory, not a failure of the audience. Pop culture earns its audience; high culture too often just inherits its prestige and then lectures people for not showing up.
      • That said, I'm not a pure populist. The Quora answer on heavy reading [2] is genuinely felt — the idea of book culture as "a millennia-long parade of dead people literally reading civilization into existence" is something I believe. The civilizational-ghost register is real. But you get there through reading, not through defending the canon against the barbarians. Harry Potter got an entire generation to devour seven fat volumes. That's the accomplishment. High culture snobbery mostly produces people who've read the right books badly.
       (really AskZirp)
from Finnegans Wake
from dissertation
  • Is anything there is relevant to modern AI agentic software development? Ask Claude
    • A surprising amount of the 1996 thesis maps onto current agentic AI development — sometimes as vindicated prediction, sometimes as unresolved problem that's now urgent.
    • What aged into the mainstream:
    • The animate metaphor became literal. The thesis's core move — taking anthropomorphism seriously as a foundation for programming rather than a disreputable teaching crutch — is now just how the industry works. People "ask" agents to do things, agents "decide" and "try approaches." The difference is the mechanism: you predicted hand-built architectures would carry the metaphor; instead LLMs made the intentional stance genuinely load-bearing, since natural-language goal specification is the actual interface.
    • Explicit goals as first-class objects. The five functions you assigned to goals — control (run until satisfied), verification (agents detect their own success/failure), conflict detection, organization of procedures around purposes, and visualization — describe modern agentic loops almost exactly. Claude Code's todo lists, spec-driven development, test-as-success-criterion, and self-verification loops are the goal-agent pattern: act repeatedly until the predicate holds. The teleo-reactive comparison holds up too — modern agent loops re-evaluate world state each cycle rather than executing fixed plans, exactly Nilsson's circuit framing.
    • The autonomy/purposefulness/reactivity triad is nearly verbatim how "agentic" is defined today, and the Dynamic Agents structure — a manager agent spawning subagents that try different methods for the same goal, dispatching on which succeeds — is the orchestrator/subagent architecture currently shipping in Claude Code, LangGraph-style frameworks, etc.
    • What aged into open problems:
    • Inter-agent conflict. Your observation that agents clobbering each other's goals is a common bug source, and that explicit goals enable detection and mediation, is directly relevant to parallel coding agents stepping on each other in a shared repo. Modern systems mostly punt on this (git worktrees, locks); your mediator-agent and stored-resolution ideas are barely explored.
    • Narrative as an interface to agent activity. Nose Goblins' storyboards — rendering agent conflict as a story with emotional state — anticipates the problem of making agent transcripts comprehensible. Today's equivalent is raw reasoning traces and log streams, which is roughly where you were before adding the storyboard layer. "Presenting the activity of hundreds of agents will require more knowledge about characters and narrative" reads as an unclaimed research agenda for agent observability.
    • The comprehensibility/power tension. Your closing worry — "there is a tension between making the system smarter and making it more comprehensible" — is now the interpretability and oversight problem, at much higher stakes than LiveWorld's constraint solvers.
    • Hierarchy critique. The section on hierarchical organization disempowering the agents most in touch with the world, while isolating decision-makers behind slow, lossy communication chains, describes a real failure mode of orchestrator architectures: context loss between manager and subagent is arguably the dominant engineering problem in multi-agent systems right now.
    • What inverted: In 1996 the missing ingredient was intelligence — agents were rules and finite-state machines, so anthropomorphism was scaffolding to help humans think. Now the components are genuinely capable, and the risk runs the other way: the animate metaphor works too well, and users over-attribute reliability, memory, and intent. Chapter 3's careful analysis of animacy as a projected framing, always paired with a mechanical view ("agents must be capable of being seen in at least two ways"), is a healthier epistemic stance than most current discourse manages.
    • The weakest mapping is LiveWorld itself — the box-based visual environment solved a 1996 problem (novices lacking concurrency primitives) that LLMs dissolved differently. But Shoham's AOP discussion (agents communicating via constrained speech acts) presages structured agent protocols like MCP more directly than most people writing those specs likely realize.
     (thanks vgr for the inspiration Venkatesh Rao (@contraptions))
from Book Reading/2025
  • had this from the library and then vgr put it on his book club list.
from obliquity
  • VGR has taken up the concept, what does he say?
    • Venkatesh Rao (@contraptions): "Thought of a term I want to work out. Oblique Altruism, as an evil twin of Effective Altruism. Inspired by the core idea in John Kay’s book Obliquity, as well as the Ken Stanley/Joel Lehman book The Myth of the Objective. The first argues that for complex problems, going at it…"
    • Ask Claude
      • Rao wrote very little about obliquity by name. What he did write puts him close to John Kay, but more radical.
      • He saw Kay as a parallel thinker. In a 2018 tweet he said he was reading Obliquity and noticed it came out in 2010, just before his own book Tempo, with a very similar premise: stories rule, and our standard idea of rationality is flawed. In Tempo he framed this more cautiously, as a contrast between "narrative" and "calculative" rationality. Kay's book appears in his bibliography and is cited once, in the Ribbonfarm post "Random Acts of X."
      • His own version goes further than obliquity. In the Breaking Smart newsletter piece "Life without goals" (2019), he argues that in creative work, setting a goal can destroy your ability to reach it. He calls this the "objective paradox": a goal shuts out the unexpected connections that produce big breakthroughs. Kay's obliquity still assumes a known goal reached by an indirect route. Rao says that in creative work the goal usually isn't properly defined in the first place. His alternative is to tinker in ways that expand what you're able to do and let the medium suggest the message, so "you'll know what you want when you find it." When a goal seems to have been set and then achieved, he says that's usually a story told in hindsight.
      • Goals are for destruction, tinkering is for creation. He thinks direct goals do work for destructive or analytical tasks like war aims, short-selling, or taking a machine apart, because entropy is on your side there. His summary: "Plan your destruction with goals, your creation through tinkering." He closes with the quip that the best way to achieve any goal is "by accident."
      • Related idea, "mediocratopia." In this series of posts he argues against optimizing at all. He recommends keeping slack and reserves, and deliberately tangling your objectives with other people's so that none of them can be optimized. His example: you want to go north and I want to go south, so we compromise on east, and together we get much farther east than either of us would have gone alone. The service also produced a first-person answer in Rao's voice. It makes the same points as the excerpts above but is a reconstruction, not his own words, so everything here comes from the actual source excerpts. Sources:
      • "Life without goals," *Breaking Smart Newsletter 2015–2019*
       (actually AskZirp but don't have the machinery for that yet)
from obliquity
  • See also Ken Stanley/Joel Lehman book The Myth of the Objective (via VGR )
from slack
  • vgr theory of slouching is kind of a pragmatic, real-world version of slack. As kind of a technique to be deployed in certain situations where agency is impossible or destructive. See stance for more.
Twin Pages

Venkatesh Rao

21 Jul 2023 - 05 Aug 2026
Open in Logseq