Follow material in an order that builds understanding.
inAiLoading page content...
Unit 10 · institutions and uncertainty
Plan for this page
After this unit, you should be able to
Terms you will use
These definitions prepare you for the reading; you do not need to memorize them.Technology interacting with people, institutions, incentives, rules, infrastructure, culture, labor, and affected environments.
A structured possible future built from explicit assumptions, used to examine consequences rather than predict one certain outcome.
A contested family of terms for broad or general machine capability; definitions, tests, timelines, and implications differ substantially.
AI may automate part of a task, augment a worker, add review work, change quality expectations, redistribute responsibility, or enable a new workflow. A job contains tasks and relationships; an organization contains incentives, processes, capital, customers, law, and power.
Evidence that a model performs a benchmark or task component does not directly establish adoption, productivity, employment, wages, quality, or distribution. Ask who adopts, which process changes, who gains control, who bears errors, and over what period.
AI can support explanation, translation, accessibility, feedback, creation, and research while also changing assessment, authorship, privacy, inequality, dependence, and the information environment. Policies must account for age, role, purpose, data, evidence, and consequences.
Media systems and generated content can scale persuasion, error, fraud, and synthetic evidence, but institutions also develop authentication, verification, provenance, literacy, standards, and enforcement. Outcomes depend on design and collective response, not capability alone.
Training and operating AI systems use computing hardware, data centers, networks, energy, water, materials, capital, and labor. Measurements and projections vary by system boundary, geography, workload, efficiency, energy mix, and future demand assumptions.
Benefit and cost are distributed unevenly. Ask who owns infrastructure and data, who can participate, which languages and communities are served, who performs hidden labor, who can contest decisions, and where environmental and economic effects appear.
One speaker may mean human-level benchmark breadth, another economic task performance, autonomous research, flexible learning, or an integrated system of models, tools, memory, institutions, and infrastructure. A timeline forecast is not a definition; a definition is not evidence that the threshold has been reached.
Separate current evidence, extrapolation, assumptions about scaling or architecture, scenario consequences, values, and policy choices. It is reasonable to prepare for plausible futures while remaining explicit about uncertainty and competing positions.
Interactive model
What is directly observed, measured, reproduced, or documented within a stated current scope?
Misconception to correct
Active checks
Explain back and transfer
Source context
Human-centred, ethical, technical, application and creation competencies; designed for education systems and requiring adaptation for broad self-directed adult learning.
Measured energy-system evidence and scenarios with explicit assumptions; projections are not guaranteed outcomes and must retain date, geography and methodology.
Prepared inAi thesis proposing a system-level view of AGI; clearly a company position, not established scientific consensus.
