Jobs contain tasks, and AI affects each differently. AI sometimes invent information depends on laboratories, engineers, suppliers and the physical systems that turn an idea into something reliable and affordable.
Reading confidence0% of the way to explored
The deeper explanation
The answer in detail
Generative AI is optimized to produce a plausible continuation, not to verify every claim against reality. When training patterns or supplied context are weak, it can confidently assemble details that sound right but are false.
Why it matters
The practical question is which parts of your work change—and whether you receive time to adapt.
The infrastructure behind AI sometimes invent information
Standards, software, data, energy and specialised equipment determine who can use ai sometimes invent information and whether different systems work together.
Investment and regulation shape the pace: research funding and large customers pay for scale, while safety, privacy or export rules determine where the benefits and risks land.
System lens
What makes this more complicated?
Three layers that help explain what changes the outcome.
Starting conditions
Evidence, available capability and access set the starting position before the technology reaches everyday use.
Limits and trade-offs
Rules, capacity and competing priorities limit which responses are practical, even when several options look possible.
Choices and distribution
The outcome depends on who can change those conditions and how its consequences are shared.
Can AI take your job—or change it?
Shared background, people, references and other questions for this story collection.
Read the main storyThe Price of the Choice
AI can make routine work faster, but who gains depends on who controls the saved time, income and decisions.
01
Who gains? Who loses?
One choice can shift advantages and costs in different directions.
Employers adopting AI
EconomicGainAutomating routine tasks can lower costs and increase output.
CostPoor deployment adds errors, privacy risks and checking work.
Workers in exposed tasks
Economic / socialGainAI can remove repetitive work or help people produce more.
CostIf employers keep the savings, workers may face weaker bargaining power, fewer hours or job loss.
Customers
AccessGainFaster or cheaper services may become available.
CostAutomated mistakes can be hard to challenge when no person is accountable.
02
What’s the catch?
Each apparent fix solves one problem—and may worsen another.
Ban AI at work
SolvesBlocks some immediate automation and data risks.
ButIt also prevents useful assistance and may push adoption out of view.
Let firms automate quickly
SolvesCan capture productivity gains sooner.
ButWorkers may lose income or tasks before retraining and protections arrive.
Require human review
SolvesKeeps a person responsible for consequential decisions.
ButReview becomes a bottleneck or rubber stamp if workloads stay high.
The unresolved choiceWill AI’s productivity gains give workers more time for skilled work—or simply raise the amount expected of them?
Meet the players
The people and organisations that can change what happens in “Can AI take your job—or change it?”. Each card explains a goal, a source of leverage and a constraint.
Companies
Technology companies
Organisations building computing & ai systems
Wants toturn technical capability into products
LeverageControl of product design, data, infrastructure and distribution.
Read profile: Technology companies
Companies decide which designs, tools and services reach the public. Their choices make the question in this story concrete for customers, workers and competitors.
They can invest, release, restrict or redesign a product, but depend on chips, cloud infrastructure, skilled people, capital and public trust.
What to watchWatch the decisions, resources or constraints that can materially change this actor’s position.
Sources: [1]
Operators
Infrastructure providers
Cloud, network, power or hardware operators
Wants tokeep the physical systems running
LeverageAccess to scarce physical capacity.
Read profile: Infrastructure providers
Infrastructure providers supply the computing, networks, energy or specialist equipment that the technology needs. A shortage or outage can stop a much larger service.
Their leverage comes from hard-to-replace capacity, while geography, investment lead times and regulation constrain expansion.
What to watchWatch the decisions, resources or constraints that can materially change this actor’s position.
Sources: [1]
Scientific community
Researchers and engineers
People producing the evidence and designs
Wants tomake the system work reliably
LeverageEvidence, specialist knowledge and technical standards.
Read profile: Researchers and engineers
Researchers and engineers test what is technically possible, measure limits and solve practical failures. Their work changes what companies can build and what users can expect.
Experiments, uncertainty, safety requirements and the time needed to verify results limit how fast ideas become products.
What to watchWatch the decisions, resources or constraints that can materially change this actor’s position.
Sources: [1]
Institutions
Public regulators
Authorities setting rules for deployment
Wants toallow useful applications while limiting harm
LeverageLicensing, standards, investigations and enforcement.
Read profile: Public regulators
Regulators translate concerns about safety, privacy, competition or security into rules that companies and users must follow. Their decisions affect which systems can launch and how they are monitored.
They need evidence and technical capacity, and must balance innovation with public accountability.
What to watchWatch the decisions, resources or constraints that can materially change this actor’s position.
Sources: [1]
Community
Users and affected organisations
People, firms and public services that adopt or reject it
Wants toget useful results without unacceptable costs or risks
LeverageAdoption, spending, feedback and refusal.
Read profile: Users and affected organisations
Users decide whether the technology solves a real problem in homes, workplaces, schools or public services. Their behaviour creates demand and exposes failures.
They have purchasing power and practical knowledge, but often cannot inspect or replace the systems they depend on.
What to watchWatch the decisions, resources or constraints that can materially change this actor’s position.
Sources: [1]
Sam Altman
Executive making release and investment choices in generative AI
Wants toscale useful AI products while maintaining public and investor support
LeverageProduct decisions, capital allocation and public commitments.
Read profile: Sam Altman
As OpenAI’s chief executive, Sam Altman has influenced how powerful models are funded, released and integrated into products. Those decisions affect competitors, users and regulators in this story.
He can direct investment and product priorities, but depends on computing suppliers, research teams, partners and public trust. A change in release policy can alter the pace and direction of the wider market.
What to watchWatch the decisions, resources or constraints that can materially change this actor’s position.
Sources: [1]
Portrait: image source
Decisions travel between them
Technology companiesInfrastructure providersTechnology companies and Infrastructure providers shape the outcome through the decisions and limits described in this story.
Technology companiesResearchers and engineersTechnology companies and Researchers and engineers shape the outcome through the decisions and limits described in this story.
Technology companiesPublic regulatorsTechnology companies and Public regulators shape the outcome through the decisions and limits described in this story.
Technology companiesUsers and affected organisationsTechnology companies and Users and affected organisations shape the outcome through the decisions and limits described in this story.
Technology companiesSam AltmanTechnology companies and Sam Altman shape the outcome through the decisions and limits described in this story.
Sources & visual credits
Flags and logos identify the actors; they do not imply endorsement.
Story collection
6 questions to explore
Open a question to read its explainer and follow the linked questions in this collection.
A quick glossary for this collection.
AutomationTechnology completing a task with limited human input.
AugmentationTechnology helping a person perform a task.
Generative AISystems producing text, images, audio or code.
AlgorithmA set of instructions a computer follows to solve a problem.
Training dataExamples used to teach an AI system patterns.
PromptAn instruction or question given to an AI system.
Sources for the whole story
How we work ↗Sources to start with
These references support the story collection as a whole.
Take the question with you
Visit the shop →Make complex ideas conversation starters.
Explore humorous products inspired by stories such as Can AI take your job—or change it?.