Week 8 Study Guide: Futures, AI Narratives, and Automation

Big idea

1. Talking about the future is never neutral. Futuring (predicting, imagining, planning, building) shapes what gets built, funded, feared, and regulated
2. Futures are told through reusable narrative templates (1984, Terminator, the Hero, "abundance")
3. Data tech doesn't just predict. It remixes the past, steers action, and preempts futures (Hildebrandt)
4. AI futures come mostly from sci-fi and transhumanism: doom, merging with machines, immortality, or abundance
5. Automation is not just a machine. It's a social relationship between workers, owners, and machines (Noble, Marx)
6. Technology is chosen, not inevitable. Design reflects who has power (Noble, Winner in Week 1)
7. Always ask: whose future is this, who benefits, and who is left out?

Hildebrandt, Smart Technologies and the End(s) of Law (Intro)

Core idea: Smart environments regulate behavior through design, without any of the democratic features that make law legitimate. Hildebrandt opens with a story to make us feel this.

Diana's "onlife" day (onlife = no online/offline divide)

• Her PDA Toma tunes her alarm to her mood, advises transit (nervous driver → higher insurance), schedules meetings around her energy, and withholds a message to protect her calm
• Daughter Lindsay is biometrically monitored at school and matched to risk "group profiles"
• Grandfather Jacob's care robot Henry balances insurer, hospital, and family interests and doesn't report his stroke. He dies.

Why a story? It imitates scenario studies: extrapolate from real trends to sensitize readers, not a Luddite fantasy or a corporate "pink" ad. (L19 uses this as the example of speculative futuring that sets us "into action.")

Law vs. technological regulation

	Law	Tech regulation
Made by	Democratic legislator	Engineers/companies, often as a side effect
Can you violate it?	Yes	Often designed so you can't
Can you contest it?	In court	Often invisible, no court

• Example: cookies were built for e-commerce, became tracking, and nobody voted on it
• Surveillance metaphor: Solove says Kafka's The Trial (anonymous accusation, chilling effect) fits better than Big Brother
• Pre-emptive computing: systems act on your inferred future. Fix proposed: Legal Protection by Design
• She deliberately skips political economy (that's what the labor material adds)

Le Guin, "The Carrier Bag Theory of Fiction" (1986)

Core idea: Our stories about technology and human progress center the Hero and the weapon. A better story centers the container: gathering, holding, and relating.

• Early humans mostly gathered (65-80% of food, about a 15-hour work week). Yet we tell the mammoth hunter story, because it's gripping: action plus a Hero.
• Drawing on Elizabeth Fisher, the first cultural tool was probably a bag or sling, not a spear
• "Killer story" = spear/arrow shape, conflict as the point of narrative, "Ascent of Man." She references 2001's ape with a bone.
• A novel is a bag: it holds people and things in relation. Conflict is one part, not the whole. A Hero "looks like a potato" in a bag.
• Sci-fi as the mythology of modern tech is Promethean and tragic (triumph → apocalypse). Redefine tech as a cultural carrier bag, not a weapon of domination, and sci-fi is less apocalyptic and more realistic.

Use it for: alternative narrative template. It critiques the Hero/conquest framing in transhumanism and Andreessen.

Andreessen, The Techno-Optimist Manifesto

Core idea: Technology plus free markets is an endless engine of growth and abundance. Slowing it down is the real danger.

• Growth = progress. Tech is the only perpetual source of growth ("everything good is downstream of growth")
• Techno-capital machine (Nick Land): markets + innovation spiral upward forever. Falling prices = rising real income. Productivity raises wages. Human wants are infinite, so jobs never run out.
• Intelligence + energy are the key resources. AI is a "universal problem solver." Delaying it "costs lives."
• Against UBI ("zoo animals"). For markets over central planning (Hayek). Wants nuclear power and 50B+ people.
• "Enemies": existential risk, ESG, sustainability, tech ethics, precautionary principle, experts, "stagnation," Nietzsche's Last Man
• Style: conquest, "technological supermen," rejects the Prometheus myth as a story of fear

How to analyze it: it is itself a narrative template (the boomer side of doomers vs. boomers). Ask whose interests it serves (a VC), what it assumes (growth is good, markets distribute fairly), and what it dismisses (labor power, ethics, regulation). It is the opposite story from Noble, Marx, and Le Guin.

Noble, "Social Choice in Machine Design" (numerically controlled machine tools)

Core idea: The automation of machine tools was a social choice, not an inevitable technical path. Design reflected who held power, between firms and between management and labor.

Two ways to automate

• Record-playback: a machinist makes the part once, the machine records and repeats it. Skill stays on the shop floor.
• Numerical control (N/C): engineers program the part from blueprints. Skill moves to the office.
• N/C won, even though record-playback was technically as reliable and suited small shops.

Why N/C won

1. Horizontal relations (between firms): the Air Force funded it ($62M+), created the market, and pushed the complex APT language as the standard. Big firms could afford it, small shops couldn't.
2. Vertical relations (labor vs. management): N/C removes the machinist as the "source of intelligence," ends pacing/soldiering, lets managers set rates by "tape time," and shifts control to management. GE called it a "management system." Postwar strikes made this appealing.
3. Ideology of control: engineers' distrust of "human error" mirrors capital's distrust of labor. Automation is driven by faith that replacing labor with capital is good, not just by profit. Drucker: automation is "a logical extension" of Taylorism.

Reality on the shop floor: desire ≠ satisfaction. Deskilling mostly didn't happen as planned, because machines are unreliable and costly. Unions fought over classifications. "The intelligence of production... remains in the possession of the work force."

Argues against: technological determinism (the story that N/C automatically led to concentration and managerial control).

Lecture 19: Data Futures

Core idea: Narratives do work, and futuring is a fiction that is always about the present too. Ask whose narrative it is and whose it isn't.

• Narratives = stories in time that explain what matters. Selective, meaningful, descriptive and directive. Everywhere, including in technologies (1860 sewing machine will "banish ragged humanity").
• Data tech works with the future beyond prediction: recycle/remix the past, shape and guide actions, prevent/preempt futures (Minority Report)
• Two purposes of futuring: impose order and control (predict, steer, preempt) vs. open up imagination (sensitize, build common ground, make a future "real enough" to take seriously)
• Routine futuring examples: use-case brainstorming, Songdo smart city (everyday utopianism), Diana's world (speculative scenario), Black Mirror (dystopian), The Matrix (inventive, not "likeness-to-life")
• Narrative templates: like cookie cutters pressed onto reality, then decorated. 1984 = "Orwellian," "Big Brother," "thought police." Alternative: Afrofuturism (Octavia Butler, Black Panther, Oakland's "Mothership").
• Beyond fiction: forecasting, scenario planning, risk assessment, VC, making companies, engineering, social action, art

Takeaway: examine how futuring works, denaturalize narratives you're handed, and look at how they universalize one vision of what's human.

Lecture 20: AI Futures

Core idea: "AI" is more than a technical system. It's a future framed by powerful sci-fi and transhumanist templates, mainly "AI is coming for us."

Template family: AI controls/enslaves/replaces us (Terminator) · we fuse with AI (2001's ending, singularity) · AI becomes our companion (Her). They're stories about what it means to be human.

Existential risk / longtermism (the EA and Rationalist world)

• Yudkowsky ("shut it all down"), Bostrom (Superintelligence), MacAskill (What We Owe the Future)
• Doomers vs. boomers: Yudkowsky vs. Andreessen/e/acc

Transhumanism (old, not new)

• Huxley coined it. Also Bernal, H.G. Wells (Men Like Gods).
• I.J. Good (1966): "ultraintelligent machine" → intelligence explosion → the "last invention"
• Extropy, Kurzweil's The Singularity Is Near
• Sci-fi becomes blueprint: Neuromancer → Neuralink; the "Torment Nexus" meme (author warns, company builds it)
• Elite voices: Musk (humans as "biological bootloader"), Altman (humans "train" too), Thiel

Longevity/immortality: Bryan Johnson, cryonics, Amodei's Machines of Loving Grace (doubling lifespan), San Junipero. Class angle: Blade Runner's Roy Batty ("I want more life"), "rise of the robots" as class-conflict allegory.

AI as mirror of the human condition: Data on trial (Star Trek), replicants, ELIZA, Pygmalion, Her. Prometheus myth: Frankenstein, Alien: Earth.

Alternatives

• Le Guin's carrier bag (see reading)
• Haraway's cyborg: not reverent, wary of holism, "illegitimate offspring" of militarism and capitalism but unfaithful to its origins

Also mentioned: obligatory futuring ("Something Big Is Happening," hype memes), "AI going rogue" headlines, "permanent underclass" fear, Pope Leo XIV encyclical on human dignity.

Takeaway: these narratives are technologically determinist and can be read through gender, class, and labor. Look at what's left out and who they serve.

Lecture 21: Automation and Labor

Core idea: Automation is a sociotechnical relationship, best understood by treating technology as capital. AI narratives about mass unemployment shape what action people consider.

Two basic narratives: automation → abundance (Brynjolfsson/McAfee, Fully Automated Luxury Communism) vs. automation → unemployment, low wages, inequality (Wiener). The real answer: it depends (jobs, wages, labor share, skill, power, which workers).

What counts as automation? A self-moving machine · any tech that raises labor productivity · tech that embodies a human skill · a machine that displaces control from the operator · a productive technology that functions as capital.

Technology as capital (Marx, Noble, Braverman)

• Automation sits within labor vs. capital, a division of labor, and who owns the machines
• Machines substitute capital for labor and displace control over the work process
• Division of labor → machine: Prony's logarithm project (mathematicians / calculators / workers) → Babbage's Difference Engine embodies that hierarchy

Industrial revolution pattern: longer hours, more intense work, downward wage pressure, deskilling (women and children enter), loss of control, surveillance (factory as panoptic), cycles of unemployment, and huge growth. Marx: machinery is "the most powerful weapon" against strikes. Luddites (1811-16) fought over control and conditions, not technology itself. Taylorism, Modern Times, CNC machines continue this.

Today

• Evidence: Brynjolfsson et al. ("Canaries in the Coal Mine"): early-career workers in AI-exposed jobs diverge. Autor & Thompson: jobs are bundles of tasks, so what matters is which tasks (expert vs. non-expert) get automated.
• AI intensifies work (HBR; agent fatigue), echoing More Work for Mother
• Worker sentiment on AI is falling (Glassdoor). Altman himself says capital/labor balance "could easily get messed up."
• Solow's paradox: "computers everywhere but in the productivity statistics." Wages lag productivity due to union decline and Baumol's cost disease (services stay "technologically stagnant").
• Value of AI: factors (data, compute, expertise, labor, energy), strategies of capture (frontier models, hyperscalers, conglomerates, open source). Who controls each?
• Hidden labor: Kenyan workers paid under $2/hr to make ChatGPT less toxic; laid-off experts training AI to replace themselves
• Responses: UBI, "luxury communism," AI-enabled planning (Cybernetic Revolutionaries), billionaire tax fights

Takeaway: firms' AI bets assume huge labor-productivity gains. Understand automation through the Industrial Revolution precedent, not just as a tool.

Quick Reference

Concept	Definition
Futuring	Any way of working with the future: predicting, imagining, planning, building
Narrative template	A reusable story pattern "everyone" knows (1984 → "Orwellian")
Obligatory futuring	Pressure to have a take on AI ("Something Big Is Happening")
Onlife	No online/offline split. Environment saturated with data-driven agents
Technological regulation	Tech shapes behavior by design, without democratic enactment, violation, or contestation
Carrier bag theory	Culture began with gathering and containers, not weapons. Stories should hold, not conquer
Techno-capital machine	Andreessen: tech + markets = endless growth and abundance
Transhumanism	Using tech to transcend human limits: merging, longevity, immortality
Existential risk / longtermism	Prioritizing humanity's long-run survival; the doomer camp
Cyborg (Haraway)	Alternative myth. Unfaithful offspring of militarism and capitalism
Technology as capital	Machines owned by capitalists, used to replace labor and shift control
Deskilling	Moving skill from worker into machine or management
Record-playback vs. N/C	Skill stays with machinist vs. skill moves to office (Noble)
Technological determinism	Tech has its own inevitable path. Noble and Winner argue against it
Solow's paradox	Computers visible everywhere except productivity stats
Baumol's cost disease	Services stay "stagnant," so costs rise

How it connects

• Hero vs. bag: Le Guin's critique applies to transhumanism, Prometheus, and Andreessen's "conquerors"
• Optimists vs. critics: Andreessen (growth, productivity raises wages) vs. Noble/Marx (control, deskilling, class)
• Tech isn't inevitable: Noble's design choices ↔ Hildebrandt's design as regulation ↔ Winner's "artifacts have politics"
• Who's left out: gatherers (Le Guin), Afrofuturist voices, hidden data workers, small shops (Noble), anyone Diana's PDA profiles