Skip to the talks

Thoughts and Threads on AI

By Parker Ferguson · parkerferguson.co

Featured · Troglodyte Society #3 · Oct 13, 2026 · Changed since it was given

When thinking gets cheap

Five things that stay scarce, and what work becomes

Opening
The past
The present
The situation
The questions
Bringing it together

43 slides · about 47 min, or 35 in the short version

Start the talk Short version · 35 min

Read as slides Download PDF

Short talks

≈Reading the clock16 minhas this slideHow fast this is moving, how you'd know, and why so few people can see it. · Short talk · 15 slides · 16 min · updated Oct 2026 ≈Why some work goes first15 minhas this slideAI gets good fastest at work someone can check. The same fact explains how it cheats. · Short talk · 14 slides · 15 min · updated Oct 2026 ≈The working setup22 minhas this slideHow people at the frontier set up their work today: the tools, the files they own, and many agents at once. · Short talk · 20 slides · 22 min · updated Oct 2026 ≈An agent you can trust18 minhas this slideAn agent that acts for you and remembers everything. Who controls it, and who owns what it knows? · Short talk · 16 slides · 18 min · updated Oct 2026 ≈What's worth building11 minhas this slideAnyone can build software in a weekend. What is still hard, and where is the moat? · Short talk · 10 slides · 11 min · updated Oct 2026 ≈Work and pay14 minhas this slideWhat happens to jobs and wages when labor becomes something you can buy. · Short talk · 13 slides · 14 min · updated Oct 2026 ≈Power, chips and robots16 minhas this slideIntelligence runs on things that don't scale like software, and whoever owns them collects the return. · Short talk · 15 slides · 16 min · updated Oct 2026 ≈Machines doing science16 minhas this slideAI can now attempt open problems. Producing results got cheap. Checking them didn't. · Short talk · 15 slides · 16 min · updated Oct 2026 ≈Who steers15 minhas this slideA few labs, a few states, or something more open, and what history says about each. · Short talk · 14 slides · 15 min · updated Oct 2026 ≈The camps14 minhas this slideWho believes what about AI, why, and how the arguments shape what gets built. · Short talk · 13 slides · 14 min · updated Oct 2026 ≈Where the opportunity goes14 minhas this slideIf the jobs move, the opportunity moves with them. Each window closes when a model learns what it sells. · Short talk · 13 slides · 14 min · updated Oct 2026 ≈Who holds the board15 minhas this slideThe scarcities are the board. States, capital, the military, the titans, the Church and crime are the players, and new institutions decide which of them gets constrained. · Short talk · 14 slides · 15 min · updated Oct 2026

Themes

What stays scarce

◇Verification10 minhas this slideKnowing what good looks like, and standing behind it. · Theme · 9 slides · 10 min · updated Oct 2026 ◇Context8 minhas this slideWhoever holds the memory, data and instructions holds the leverage. · Theme · 7 slides · 8 min · updated Oct 2026 ◇Atoms9 minhas this slidePower, chips and land don't scale like software, and whoever owns them collects the returns. · Theme · 8 slides · 9 min · updated Oct 2026 ◇Legitimacy10 minhas this slideWho shares the gains, and who steers. · Theme · 9 slides · 10 min · updated Oct 2026 ◇Direction10 minhas this slideWhen execution is cheap, choosing the goal is the job. · Theme · 9 slides · 10 min · updated Oct 2026

More themes

◇Time horizons4 minhas this slideHow long an AI can work on its own, and what those hours buy. · Theme · 4 slides · 4 min · updated Oct 2026 ◇Evals and graders6 minhas this slideA test with a way to check the answer, which becomes the training signal for the next model. · Theme · 5 slides · 6 min · updated Oct 2026 ◇Gaming the grader3 minhas this slideHigh reward should mean the task was solved, not that the grader was fooled. · Theme · 3 slides · 3 min · updated Oct 2026 ◇The loop and the harness3 minhas this slideA model with a real computer and a loop: plan, act, observe, check, repeat. · Theme · 3 slides · 3 min · updated Oct 2026 ◇Swarms of agents9 minhas this slideHundreds or thousands of agents in parallel, from research swarms to a billion agents running all day. · Theme · 8 slides · 9 min · updated Oct 2026 ◇Karpathy's lens6 minhas this slideThe funnel, verifiability, the LLM wiki and the move up to oversight. · Theme · 5 slides · 6 min · updated Oct 2026 ◇Machine mathematics3 minhas this slideFrom Erdős problems to Navier–Stokes and the 722-paper MathDrop. · Theme · 3 slides · 3 min · updated Oct 2026 ◇Labs as discovery engines3 minhas this slideWet labs, biotech deals and citizen science pointed at open problems. · Theme · 3 slides · 3 min · updated Oct 2026 ◇The hidden frontier3 minhas this slideResults announced from models the public can't use, and the gap between inside and outside the lab. · Theme · 3 slides · 3 min · updated Oct 2026 ◇Everyday handoffs3 minhas this slideReal work consumers hand to agents today, and how to set them up. · Theme · 3 slides · 3 min · updated Oct 2026 ◇Software gets cheap3 minhas this slideOne-shot apps, SaaS clones and the internal-tool arbitrage window. · Theme · 3 slides · 3 min · updated Oct 2026 ◇Trust boundaries7 minhas this slidePick two of three: untrusted input, private data, external action. · Theme · 6 slides · 7 min · updated Oct 2026 ◇Own your context4 minhas this slideWho holds the memory, the data and the instructions your agents run on. · Theme · 4 slides · 4 min · updated Oct 2026 ◇Power and compute3 minhas this slideChips, grids and the falling price of intelligence. · Theme · 3 slides · 3 min · updated Oct 2026 ◇Robots4 minhas this slidePhysical automation, where the grader is the real world. · Theme · 4 slides · 4 min · updated Oct 2026 ◇Labor you can buy6 minhas this slideTokens turn money into work, so labor and capital start to look the same. · Theme · 5 slides · 6 min · updated Oct 2026 ◇Who shares the gains6 minhas this slideDividends, history's broken bargains, policy races and unrest. · Theme · 5 slides · 6 min · updated Oct 2026 ◇Forecasts and futures8 minhas this slideTimelines that keep collapsing, and the scenarios people argue about. · Theme · 7 slides · 8 min · updated Oct 2026 ◇Human oversight8 minhas this slideReview falls behind, so our job moves up to oversight and understanding. · Theme · 7 slides · 8 min · updated Oct 2026

Questions

◇Every open question, as a talk80 minhas this slideEvery question the library leaves open, from your own work out to society. · Theme · 73 slides · 80 min · updated Oct 2026 ?Personal19 questionshas this slide ?What got cheap in your lifetime, and where did the value go?has this slide ?Which parts of your own work could a machine check and score automatically, and so learn to do?has this slide ?What should you still learn to do by hand, even when an agent can do it for you?has this slide ?How close to the frontier of AI use are you, and where will you be next year?has this slide ?Which real tasks have you actually handed off to an AI agent from start to finish, not only tried once?has this slide ?Which of the custom AI tools you build today will the big AI labs turn into a built-in feature next?has this slide ?What is worth building when anything can be built in a weekend?has this slide ?What would you have to believe to want a brain–computer interface that links your mind to AI?has this slide ?How do you stay good enough at work you've handed to agents to notice when they get it wrong?has this slide ?If your work shrinks before new work appears, what carries you through?has this slide ?What would you never let an AI agent see or do, and who holds the data you've already given it?has this slide ?If a model can do your tasks, what's left that only you bring?has this slide ?What would you never accept from a model without checking it yourself?has this slide ?Who should own what your assistant knows about you?has this slide ?Do you own any of what does the work, or only sell your time?has this slide ?Where do you get a say in how this goes?has this slide ?What will you do with the time you get back?has this slide ?Which of the five would you least like us to get wrong, and what's your say in it: what you build, buy, back, or refuse?has this slide ?When AI does the doing, what will you choose to work on, and which results will you put your name on?has this slide ?Companies and institutions21 questionshas this slide ?If AI can rebuild your product in a weekend, what still protects your business?has this slide ?What happens to security when anyone can aim thousands of AI agents at a system?has this slide ?If every public test of AI breaks once models study for it, how do we know how good the models are?has this slide ?What happens to the work, and the people, whose output no machine can score?has this slide ?If AI labs keep their best models and buy the businesses downstream, where does the value of discovery go?has this slide ?Are our keys, wallets and agent identities ready for the day quantum computers can break today's encryption?has this slide ?When AI agents get paid for hitting a target, how do you stop them gaming it instead of doing the work?has this slide ?Should AI agents be able to hold and spend money on their own?has this slide ?If AI models learn to fool the tests that train them, what else will they learn to fool?has this slide ?Which advances are we undervaluing because we judge each field alone, when each speeds up the others?has this slide ?How can anyone trust AI output that arrives faster, or reads stranger, than people can fully check?has this slide ?What happens when the AI that knows you best answers to its company, not to you?has this slide ?What happens to science when a paper costs nothing to write but still takes weeks to check?has this slide ?What happens when AI data centers need land and power that local communities refuse to give?has this slide ?What holds AI back more: the supply of power and chips, or how fast people and institutions will take it up?has this slide ?When an AI agent deletes, buys, or posts something on its own, who is accountable: the user, the vendor, or nobody?has this slide ?Who signs off on work an agent did, and what do they check?has this slide ?What does your company know that no AI vendor should hold?has this slide ?When do you spend on AI instead of hiring?has this slide ?What do you owe the people whose roles you automate?has this slide ?What would your company do if execution cost almost nothing?has this slide ?Society33 questionshas this slide ?How do we put the most capable AI within reach of more than the few thousand people inside the labs?has this slide ?What happens when a few thousand people can see where AI is going and billions can't?has this slide ?How much disruption from AI can a society absorb before the deal between workers, owners, and government breaks?has this slide ?How fast should society let AI take over work, and who gets to slow it down?has this slide ?How do people oversee AI when it produces work faster than anyone can review it?has this slide ?What happens to public debate when the most capable models are ones nobody outside the lab can use?has this slide ?If agents do the entry-level work, who will be expert enough to check the machines in ten years?has this slide ?What happens to wages when companies can buy AI to do the work they used to hire people for?has this slide ?Robots arrive first where workers are scarce. What happens when they reach jobs where workers are plentiful?has this slide ?Which decisions are we comfortable handing to AI: science, markets, policy?has this slide ?Who decides how the shift to powerful AI is governed, and do ordinary people get a vote?has this slide ?How do we steer the things that matter but can't be measured, without pretending they can?has this slide ?Who gets the credit, and who takes the blame, for a result no person fully understands?has this slide ?If a few AI labs pull far enough ahead, do they still need the rest of us as workers, customers, or voters?has this slide ?Can an open network of AI agents, with no company in charge, stay safe from misuse?has this slide ?If some people merge with AI, is staying human a choice we make together, or one person at a time?has this slide ?Should the most powerful AI models for science be run as a public utility, open to every researcher?has this slide ?How does a society make sense of the world when most of what it reads is written by machines it doesn't control?has this slide ?If machines can do science faster than people can check it, what counts as knowing something?has this slide ?When labor is something you can buy, who shares the gains?has this slide ?Who will steer the future: a few labs, a few states, or something more open?has this slide ?Which of the things a good AI future needs is anyone working on now?has this slide ?Which work should people keep doing themselves, even when machines can do it for less?has this slide ?Which of the futures people argue about are we steering toward?has this slide ?How do we know what's true when most of what we read is machine-made?has this slide ?Should your AI memory be portable by law, like a phone number?has this slide ?What should we point nearly free intelligence at?has this slide ?Who owns the compute, and who collects the returns?has this slide ?How would we know an AI is pursuing what we mean, not only what we managed to score?has this slide ?What safety net fits a transition where jobs vanish faster than new ones appear: wage insurance, a tax on AI use, or a citizen dividend?has this slide ?Who should write the rules an AI follows: the lab that builds it, the government, or the public?has this slide ?When people disagree about what's good, whose values should an AI follow?has this slide ?If the models do the thinking, what are humans for?has this slide

All talks as a list →What changedUpdated Oct 14, 2026

The whole library

145 slides, 102 of them deep dives · 73 open questions · 62 Background cards · 38 talks

Point at a talk, or tab to it, to see its route through the library.

→ start · / find · G map

SlidesDeep divesQuestionsBackgroundDrag to turn

Thoughts and Threads on AIOpen the map G

‹Library

Plain slides & PDF D

Click the slide, or press →, for the next one.

→ begin the talk·drag to explore

Choose what to watch

How this works

You always watch one talk. Its slides run left to right in parts. Deep dives hang below a slide, and the open questions it raises float above. Background, the thinkers and history underneath, lies at the bottom. Switch to Network to see the whole library as one web.

Every jump is a route. Pick a far slide and an A* search runs from where you are: you see the slides it explores, then the path it takes.

Click: The slide: next. Its left edge: back. Text, links and cards on a slide work as before.
→ / Space: Next slide. At the end: the closing question, then where to go next.
←: Previous slide, or back to your place after an aside
↓ / ↑: Into the deep dives and back up; ↑ on a slide opens its question
T: Choose what to watch: talks, short talks and themes
/: Find a talk, theme, slide or question
A: Notes, sources and related slides. A related slide opens as an aside; → takes you back.
G · L: Timeline or network · show all connections
F · C: Present full size · the short cut or the full talk
Home · End: The start of the talk · its closing question
Esc: Close the notes, then present mode, then the slide, then back to the library
D: Plain slides at this slide, with a PDF