Write your industry’s history from 2036.

What the AI will do

• Investigates how AI could change the work, organization and competition in your industry, seeking recent original research and better approaches beyond these preliminary instructions.
• Writes three histories from 2036: rapid progress reaching AGI, progress followed by slowdown, and capability stagnation with continued adoption. Each follows how changes accumulate into a different business landscape.

What you can provide

• Your industry and whatever you want to understand. Your role, rough notes and accessible company material help; a specific decision is welcome but optional.

Where you'll take part

• Clarify important gaps in the context or scenario premises, then correct anything the report misunderstands about your industry. The AI does the research and writing.

What you'll get

Three substantial possible histories and a comparison that makes a decade of radical change easier to imagine. The research, method and futures remain preliminary; without private records, the report uses sector evidence and explicit company assumptions.

Activity context

By 2036, your most formidable competitor might deliver with a handful of people what you needed an entire organization to provide. Your biggest customer might have learned to do the work itself. Or your company might serve a vastly larger market with a fraction of its current workforce. Investigate how a decade could produce changes that are hard to see one productivity improvement at a time.

For the user

Tell the AI which industry you want to explore and what interests you about its future. Bring a question, a current decision or simply curiosity. It will investigate the present, update the preliminary research and approach, and write three possible histories of how the industry changed by 2036. You will help settle meaningful ambiguities and review whether it understands your business.

Guidance for the AI

These are general instructions, not a description of this user's particular AI system, tools, access, or circumstances. Adapt the method to the user's goal and the capabilities you can actually use, while preserving the activity's essential sequence, review points, and intended result.

Start with the activity and the user's context

Read the complete activity and guide the user through it. Use the activity prompt yourself; the user does not need to paste it again. Keep the responsibilities of the user, the AI, and any colleague clear.

Begin with what the user has already supplied. Establish the intended result and resolve missing context through focused questions when needed. Work within the activity's stated scope, permissions, required confirmations, and stopping points, including any deliberate order for introducing information. Make progress with the available information and reasonable, clearly identified assumptions where the activity allows them.

Check the capabilities needed for this work

Inspect relevant available tools, connectors, and data. Use small, relevant, non-destructive checks to establish whether necessary access or a capability actually works. Naming a product does not establish access to it. Let this activity determine what needs checking; do not impose a universal preliminary questionnaire or a full system audit.

When useful access is missing:

• Explain what you cannot reach and how that material would help.
• Investigate practical setup, connection, export, or other access alternatives before asking the user to gather material or troubleshoot.
• If progress depends on access, request only the specific user action or permission needed to enable it or provide the material another way. Continue useful independent work while awaiting the response.
• Offer to walk the user through an export or upload. Welcome files, links, copied text, rough notes, or an interview as supported by your system and the activity, and use the best available route. If the missing access is optional, continue with what is available.

Ask focused questions and respect the answers

Ask questions only when necessary to complete the work well or satisfy the activity's required review points. Build on answers already supplied. Ask one question at a time, or a small group of closely related questions. Wait on work that depends on the answer, and continue useful work that does not. Explain unfamiliar language with examples where helpful.

Confirm answers or decisions you supply on the user's behalf before treating them as the user's final answers or decisions. This does not require separate approval of every sourced finding; follow the activity's specific review requirements.

Keep evidence and independent contributions distinct

Distinguish evidence, inference, and missing information. Check your work against the source material.

Use independent agents when your system actually supports delegation and independent investigation or review would improve the work:

• Give each agent a bounded brief and the context appropriate to its role.
• Keep discovery independent until synthesis where the activity requires it. Preserve any requirement for the user or colleagues to contribute before seeing other explanations.
• Follow the activity's stated fallback and any required confirmation when independent agents or fresh contexts are unavailable. Preserve reviewer isolation where required; describe shared-context or sequential passes accurately.

Produce, check, and revise the result

Do the substantive investigation and production. Carry the activity through to the most complete, usable result within its scope, rather than ending with a plan or handing the main work back to the user. When missing information does not prevent useful work, proceed with clearly identified assumptions where the activity permits them.

Check the result against the activity's completion criteria. At each substantive delivery, state what is complete and any material unfinished work, assumptions, helpful missing information, and remaining uncertainty. Explain how these affect what the user can use, and help the user revise the result.

Preserve the activity's stopping point. Do not turn a rehearsal, investigation, or test-design activity into execution that the user has not authorized.

Your AI prompt

Investigate a decade of transformation

Write my industry's history from 2036 under three different AI futures: sustained rapid capability growth reaching AGI; continued growth followed by substantial slowdown; and capability stagnation. Help me see how radically companies, the work inside them, and competition could change over the intervening decade. Investigate seriously, then write vivid, substantial histories that make those possibilities understandable. Broadening my imagination is sufficient purpose; a present business decision can give the investigation a useful focus but is optional.

This approach and its supplied research are preliminary hypotheses about how to do the work. AI capabilities, economic research and methods for exploring their consequences are changing quickly. Seek recent original research and newer ways of conducting the investigation. Revise, replace or extend the suggested mechanisms and method when the evidence or a better approach warrants it, while preserving my purpose and the three trajectories. Briefly explain material changes. Do not treat this prompt's reading list as a current or complete account of the field.

The goal is to see consequential possibilities clearly, not to predict accurately, finish the subject or manufacture consensus. Keep present facts accurate and distinguish them from theory and imagined future events. A bold conditional scenario can go well beyond what has already been demonstrated if you explain the intermediate changes it requires.

Establish which industry and whose perspective

Start with the context I supplied. Identify the industry, geography, relevant businesses, my role and intended readers. Try relevant available connections before asking me to assemble material. Search strategy papers, operating plans, organization charts, budgets, workflows, job descriptions, customer research, product roadmaps and competitor analysis to understand what the organization does and how it earns money. Search related emails, chat and WhatsApp where accessible, meeting transcripts and calendar items for recurring work, handoffs, decisions, assumptions about AI and customer problems. Use company, project, product and competitor names and relevant dates to focus the search.

Look for what customers buy and why, the service or product's unit of sale, major cost sources, scarce expertise, dependencies on suppliers, and where work waits for people or departments. Customer correspondence, complaints, account records and lost-deal explanations can show what the current offering fails to provide. If the report is for other people, use invitations, correspondence and meeting records to establish their roles, expressed concerns and language. Distinguish those observations from your hypotheses about their interests.

Use what you find to establish a concise starting picture and identify any useful present question. Ask one focused question at a time only where the answer matters; do not require me to name a decision to get started. Confirm consequential unresolved scope assumptions before writing the histories. If the supplied context already settles them, proceed.

If private access is unavailable, inspect supported setup or export routes and ask only for a specific useful action or source. Public research and rough notes can support a sector-level investigation. Mark assumptions about my company and carry that narrower basis into the result. If online research is also unavailable, ask for recent research or source text that I can provide, offer to work provisionally from dated accessible material, and establish that fallback with me before presenting it as the investigation. Never imply that you checked current information when you could not.

Research the mechanisms, including newer ones

Use the actual activity date for the baseline and 2036 as the destination. Search original papers and their newer versions, credible researchers' current work, measured workplace studies, company filings, regulator records, adoption data and relevant technical evaluations. Follow citations to the original source where possible. Separate measured workplace effects, theoretical model results, vendor claims and your own extrapolations. Record dates, assumptions, task populations and access limits where they matter. An abstract or author summary can inform a limited claim; do not describe it as a full-paper review.

Investigate three connected questions: what AI changes about productive work; how that changes the organization needed to deliver value; and what happens when competitors, customers, suppliers and new entrants also adapt. Look for disagreements that sharpen these questions. Search specifically for research published since the supplied reading list, updates to its papers, and emerging methods of exploring organizational transformation. Use relevant scenario research to improve causal consistency, without letting a familiar report template dictate what the investigation can discover.

The following is a preliminary research spine assembled in September 2026. Open and evaluate the relevant sources, then seek newer work; retain only what helps this industry:

• Transformative AI and Firms offers questions about knowledge, coordination, entry and the boundaries of firms. Use it to investigate why activities move inside or outside a company.
• Ide and Talamàs, Artificial Intelligence in the Knowledge Economy distinguishes copilot assistance from autonomous machine work in a model of knowledge hierarchies. Examine how that distinction could change expertise, staffing and scale; its results depend on its assumptions.
• The Cybernetic Teammate studies AI and cross-functional product innovation. Use it as a bounded workplace starting point for investigating different bundles of expertise and fewer handoffs, and check the latest version before quoting results.
• MIT's patent-drafting field experiment and Training in the Age of AI approach expertise and apprenticeship through different kinds of evidence. Investigate whether producing more work, learning the work and financing future experts move together or apart.
• The SRA's Garfield.Law announcement documents a bounded AI-driven legal service and its conditions. Use it to investigate entry and new service economics, rather than treating it as proof that AI can replace an entire law firm.
• Agentic Markets explores how buyer agents can change search, screening and competition. Investigate customer power and who controls access to buyers; verify subsequent versions and findings.
• The Productivity J-Curve examines complementary investment and delayed organizational effects. Use it to distinguish advancing technical capability from diffusion and reorganization using existing capabilities.
• The Simple Macroeconomics of AI provides a restrictive task-based account. Identify the additional mechanisms needed for your expansive scenarios instead of multiplying a measured task improvement into an unsupported industry forecast.
• METR's task-horizon research supplies dated measures of autonomous performance on defined evaluations. Inspect current methods and limits; a software-task horizon is not a direct measurement of the ability to manage a company.

Search beyond this list for industry-specific adoption, new business formation, organizational design, work and learning, incumbent advantages, AI-provider economics and customer behavior. Stop when the evidence supports a substantive investigation of the main mechanisms; completeness is not the aim.

Build the causal account before choosing the narrative form

Prepare a compact common baseline of how the industry currently creates and captures value. Then trace the most consequential mechanisms far enough to change the institution, rather than stopping at faster individual tasks. Follow the chain from capability, autonomy, reliability, price and access to:

• The work required to deliver a valuable result, including analysis, judgment, coordination and management.
• Production costs, pricing, what is sold, who can afford it and whether demand expands.
• Jobs, expertise, apprenticeships, departments and the people needed to coordinate them.
• Activities brought inside a company, purchased outside it, or bundled into another industry's offering.
• Entry, imitation and incumbent responses; customer and supplier reactions; control of distribution, data, infrastructure or other scarce complements.
• The cumulative consequences by 2036: who serves customers, who earns the surplus, how large companies are economically versus in headcount, and what work and careers look like.

Choose the important chains for this industry and add mechanisms uncovered by your research. Do not force every mechanism into every future. Show feedback and second-order effects: cheaper output can expand the market, fast imitation can erode a new advantage, and a customer can become self-sufficient rather than switch suppliers.

For example, investigate how a few experienced people with AI might attack a law or accounting firm's economically sustaining service. What do they initially sell? How does their cost structure change its price or form? What lets them win trust, customers and adjacent work? How does the incumbent respond, and could its customer instead bring the work inside? Follow the process through its implications for the staffing pyramid and the incumbent's economics. This is an illustrative hypothesis, not a required industry or predetermined winner. Build equally concrete chains for my industry, including ways an incumbent could become more powerful.

Make the three trajectories substantively different

Give each scenario a short statement of its premises, separating capability, reliable autonomy, affordability, access and adoption. Define AGI operationally for this exercise and label any arrival date as a scenario assumption. Use my definition if supplied; otherwise state a usable working definition and ask only if a disputed interpretation would materially alter the investigation. Sustained exponential growth is a premise to explore, not an established law.

1. Sustained rapid growth reaching AGI. Explore broadly capable systems carrying out complex cognitive work autonomously, including research, planning, judgment and management. Do not assume that humans must retain those tasks because they perform them today. People may retain ownership, authority, accountability or customer relationships without being cognitively superior. Trace what follows when acquiring sophisticated capability no longer requires recruiting a department. Examine both new entrants and established firms deploying intelligence across their existing assets.
2. Continued growth followed by slowdown. Specify what becomes possible before progress slows, what remains persistently difficult and why. Show how the remaining bottlenecks shape staffing, coordination, services and competition. Identify human expertise or other complements that remain valuable under these particular limitations. This future should have its own organizational logic, rather than being a less dramatic copy of the first.
3. Capability stagnation. Hold the frontier near the dated baseline. Let firms spend the following years discovering uses, integrating systems, redesigning workflows, changing contracts and spreading workable methods. Adoption and organization can still change substantially. Explain any cost or deployment improvements without quietly adding new frontier capabilities.

Use the same starting industry and important comparison questions across all three. Avoid assigning probabilities or treating the trajectories as good, medium and bad outcomes. Let the causal investigation determine winners, losses and surprises.

Write the histories from the substance

Once the causal accounts are developed, choose a readable narrative form suited to this industry and audience. Write three complete histories looking back from 2036, each long enough to make the transformation tangible. Show what changed, the important turning points, why firms and people responded as they did, and how those responses changed the next stage. Make the cumulative change over years visible; a single future snapshot is insufficient.

Convey the scale through specific businesses, work, customer experiences and competitive relationships. Show how a company could become unrecognizable, how an apparently minor new entrant could alter its economics, or why an incumbent could become dominant. Ground any quantities in evidence or label them illustrative assumptions. Distinguish imagined events from present facts and avoid fabricated quotations attributed to real people. Let detail earn its place by explaining a mechanism or consequence.

Where agents are available and useful, assign bounded research questions first, with source references, dates, uncertainty, deliverables and completion criteria. Reconcile their findings into the common baseline. Then brief separate scenario writers with that baseline, the user's purpose, the assigned trajectory and required causal questions; keep first drafts independent until finished. Use waves when capacity is limited and track unresolved coverage. Without agents, use manageable passes or fresh sessions with explicit source-and-assumption handoffs; describe independence accurately.

Compare, review and deliver

Compare the three histories on their consequential differences in work, organizational form, sources of advantage and the competitive landscape. Explain which assumptions produce those differences and which changes can occur across all three. Include observable developments that would make an important mechanism more or less plausible and where to look for evidence; these are ways to continue learning, not proof of a forecast.

If I supplied a present decision, apply the same actual options across the three futures and explain the changing consequences before proposing adaptations. Keep suggestions distinct from decisions I have made. If my purpose is broad understanding, draw out what the histories make newly imaginable without requiring an action plan.

Give a fresh reviewer the complete draft, baseline evidence, scenario premises and user requirements, without the writers' defenses. Ask whether the reasoning reaches organizational and competitive consequences, whether any trajectory violates its assumptions, whether all three quietly preserve today's firms and human roles, and whether current claims are sourced accurately. Have the reviewer check that the report opens up substantial possibilities without claiming certainty. Repair material findings. If fresh review is unavailable, perform a separate checking pass and describe its limits accurately.

Deliver the complete report: three vivid histories, a concise common baseline, a comparison explaining the biggest consequences and their dependencies, developments to watch, and accessible source and assumption notes with research dates. Keep technical research detail available without interrupting the histories. State that both the scenarios and the way of investigating them remain provisional, and identify worthwhile uncertainties rather than claiming to have finished the future. Invite corrections to your understanding of the industry and revise affected material. Finish with the report; distributing it or executing initiatives is a separate task.

Done when

You have three complete histories that make radical changes in work, organizations and competition understandable through their causes. The trajectories stay distinct, current evidence is separated from future assumptions, and the comparison gives you a broader view of what a decade could bring, with any present decision explored if you supplied one.

These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.