Dr. Letitia Wright’s experience offers a useful answer to the fear that AI automatically makes a person’s work obsolete: in her account, it reduced repetitive preparation and created more room for creativity, strategy, experimentation and human connection. That is a personal account, not a guarantee for every occupation.
The broader evidence points in the same direction but with important limits. The International Labour Organization (ILO) finds that generative AI is more likely to transform many jobs than eliminate them because most occupations still require human judgment and interaction. Whether that transformation becomes a better job depends on the tasks, the way the technology is introduced and who controls the resulting time.
As an Amazon Associate I earn from qualifying purchases.
What Wright’s story actually says
A career built around change
The essay in The AI Journal follows Wright through several kinds of work: healthcare, television, crowdfunding, streaming media and live social selling. Those moves matter because they show a worker repeatedly carrying experience from one setting into another rather than treating a career as a fixed list of duties. Her guiding question is, “What can I learn that I didn’t know yesterday?”
Free tools Windows power users keep installed
One-click scans. No signup required.
That question changes the meaning of AI assistance. Instead of asking only whether a system can perform a task, Wright asks what new work becomes possible when routine parts take less effort. In her telling, the technology supports preparation, research, drafting and experimentation while she remains responsible for decisions, relationships and the direction of a project.
AI as a collaborator for routine work
Wright characterizes the change this way: “AI doesn’t eliminate the work, but removes much of the repetition surrounding the work, which matters.” Repetition can include organizing information, producing early drafts, generating alternatives or preparing material for a presentation. Those activities are often necessary, but they can consume the time a small-business owner or creator needs for judgment and discovery.
Reducing that friction does not make the remaining work automatic. Someone still has to decide which information is relevant, check whether an output is accurate, adapt it to a real audience and accept responsibility for the result. The benefit Wright describes is the option to spend more of her limited attention on those higher-value choices.
Why live selling still depends on people
Her live-selling experience makes the boundary especially clear. AI can help a seller prepare talking points, organize product information or rehearse possible questions. During a live interaction, however, success depends on listening, timing, credibility and an ability to respond to another person’s concerns. Those qualities are not interchangeable with a generated script.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Wright therefore presents AI as preparation for human connection, not a substitute for it. She does not claim that the technology caused better sales, and her essay names no particular model or application. The account is valuable as an illustration of changed possibilities, not as a controlled productivity test.
Three different questions that are often confused
“Will AI replace my work?” combines several questions that need separate answers. A job can contain exposed tasks without being removed. A worker can save time without producing more measurable output. An employer can announce an intention to automate without carrying it out.
| Question | What the evidence can show | What it cannot establish by itself |
|---|---|---|
| Is an occupation exposed? | A modeled assessment of how many tasks could be affected by current generative AI. | That every worker in the occupation will lose a job or that adoption will occur soon. |
| Did a worker’s routine become easier? | A personal account, survey response or experiment may report time saved or a different workflow. | That the same result will occur in another workplace, sector or country. |
| Will employment change? | Employer plans and economic projections can indicate possible directions. | Observed layoffs, new hiring or wage changes unless those outcomes are measured directly. |
Keeping these categories apart prevents both overconfidence and unnecessary alarm. Wright answers the first-person question—what changed in her own work—while labor studies address patterns across occupations and populations.
What the ILO’s evidence says about replacement
Broad exposure is not broad redundancy
The ILO’s “Generative AI and jobs: A 2025 update,” published on 20 May 2025, estimates that one in four workers worldwide are in occupations with some degree of generative-AI exposure. Its central qualification is just as important as the headline number: most jobs are more likely to be transformed than made redundant because human input remains necessary.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchThe highest-exposure category is smaller
The ILO’s Working Paper 140, “Generative AI and Jobs: A Refined Global Index of Occupational Exposure,” also published on 20 May 2025, places 3.3% of global employment in its highest exposure category. This is a modeled occupational estimate, not a count of layoffs. The paper reports differences by gender and country income, so a global percentage should not be read as a prediction for an individual labor market.
Tasks determine the practical effect
The ILO’s companion explanation, “How might generative AI impact different occupations?” describes an assessment that scores tasks within occupations. Few jobs consist entirely of tasks that current systems can automate. Nearly all occupations include activities requiring human input, such as accountability, physical action, contextual judgment, negotiation or care.
That task-based approach explains why a job title alone is a poor guide. Two people with the same title may spend different proportions of their time on documentation, analysis, customer contact or hands-on work. Their exposure, and the opportunity to redirect saved time, can therefore differ substantially.
The numbers to keep in context
| Publisher and date | Figure | How to read it |
|---|---|---|
| International Labour Organization, 20 May 2025 | One in four workers worldwide are in occupations with some generative-AI exposure. | A broad modeled exposure measure; the ILO says transformation is more likely than redundancy for most jobs. |
| International Labour Organization Working Paper 140, 20 May 2025 | 3.3% of global employment is in the highest exposure category. | An occupational index, not observed job losses or a forecast for a particular person. |
| World Economic Forum, Future of Jobs Report 2025, announced 7 January 2025 | Employers projected 170 million new jobs and 92 million displaced jobs by 2030. | A projection across multiple global trends, not a generative-AI-only causal estimate. |
| World Economic Forum, 2025 employer survey | 77% planned to upskill workers; 41% planned workforce reductions as AI automates certain tasks. | Reported intentions, not completed training, layoffs or redeployment. |
The World Economic Forum figures describe what surveyed employers expect and plan. They should not be presented as outcomes that have already happened, nor as a precise forecast of how one person’s role will change.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →What productivity evidence can—and cannot—say
The ILO’s “The impact of GenAI on jobs, productivity and work organization: a review of the empirical evidence,” published on 1 June 2026, examines experiments, firm-level data, platform studies and worker and firm surveys from Australia, Denmark, Germany, Korea, Kuwait, the United Kingdom and the United States.
Its conclusion is measured: productivity gains appear real in some settings, but they are uneven and often unverified. Workers sometimes report saving a few per cent of their working hours, yet the reviewed evidence has not consistently shown corresponding increases in measured output, earnings or employment. This does not invalidate Wright’s experience. It means that personal time savings and economy-wide productivity are different measurements.
A saved hour can be used in several ways: a worker may take on more creative work, serve customers better, learn a new skill, reduce stress or simply finish earlier. An employer may instead increase the workload, capture the time as cost reduction or leave the process unchanged. The technology alone does not determine which result occurs.
A practical way to test whether AI expands your role
Rather than starting with a tool, start with the work. The following audit keeps the human contribution visible and makes it easier to judge whether an apparent gain is genuine.
Rank #4
- List recurring tasks. Record a representative week of activities such as searching, summarizing, drafting, scheduling, checking data, answering routine questions and speaking with customers.
- Separate preparation from judgment. Mark which steps produce a first draft or organize material and which steps require interpretation, accountability, empathy, negotiation or a decision with consequences.
- Choose a low-risk pilot. Begin with a reversible task where an incorrect output will not harm a customer, patient, employee or legal obligation.
- Define a quality check before using AI. Decide what must be verified: factual accuracy, calculations, source coverage, tone, privacy, accessibility or compliance with a professional rule.
- Measure more than minutes. Track correction time, error rates, customer response, worker control over the process and whether the saved time is actually redirected to a useful activity.
- Review the effect with the people doing the work. Ask whether the process reduces drudgery or merely adds monitoring, hidden editing and pressure to produce more.
| Evaluation axis | Questions to ask |
|---|---|
| Task fit | Is the activity repetitive and information-based, or does it depend on physical presence and situational judgment? |
| Human contribution | Who makes the final decision, handles exceptions and maintains the relationship? |
| Quality and verification | Can an expert check the output against reliable records before it reaches another person? |
| Autonomy and job quality | Does assistance give the worker more discretion, or does it increase surveillance and pace? |
| Realized benefit | Does reported time saved become learning, innovation, service quality or another measured improvement? |
How this framework fits entrepreneurial work
Research and preparation
An entrepreneur can use an AI system to assemble questions, sort notes or propose an outline before investing time in a polished draft. The owner still needs to inspect sources, remove unsupported claims and decide what matches the audience. The practical gain is a faster starting point, not permission to skip expertise.
Drafting and experimentation
Generating several rough approaches can make experimentation cheaper. A creator might compare different explanations, offers or formats and then select one that reflects a real understanding of the audience. The human work lies in choosing the experiment, interpreting feedback and changing course when the evidence disagrees with the first idea.
Live interaction
In a live-selling setting, preparation can include an organized product brief, likely questions and a short list of facts to verify. Once the broadcast or conversation begins, the seller must notice confusion, answer honestly, disclose relevant limits and adapt to the person in front of them. A generated answer that sounds confident but is wrong can damage trust faster than a pause to check.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Protecting the possibilities AI creates
Expanded capacity is not automatically expanded opportunity. Workers and managers should watch for predictable failure modes:
Recommended Free Tools
- Automation bias: treating fluent text as accurate. Require source checks and an accountable reviewer.
- Hidden workload: counting only generation time while ignoring editing, fact-checking and exception handling. Measure the entire process.
- Deskilling: removing practice that workers need to develop judgment. Keep people involved in decisions and rotate opportunities to perform the underlying task.
- Privacy leakage: placing customer, patient or confidential business information into an unapproved system. Use only tools and data practices authorized by the organization.
- Unequal access: giving assistance to some workers while judging others by old output standards. Explain who receives the technology, training and time to learn it.
- Work intensification: converting every saved minute into additional assignments. Agree in advance how reclaimed time will support quality, learning, recovery or innovation.
These safeguards turn AI from an individual workaround into a deliberate work-design choice. They also make it possible to tell whether a benefit belongs to the worker, the customer, the organization or some combination.
Best Value
- Author: Bungay Stanier, Michael.
- Publisher: Page Two
- Pages: 244
- Publication Date: 2016-02-29
- Edition: 1
Will our skills still matter?
Wright names the anxiety directly: “whether our skills will still matter.” The task evidence suggests that the answer depends on which skills are being used and how they connect. Skills that frame a problem, judge evidence, build trust, handle exceptions and take responsibility remain central even when a system produces a draft. Technical fluency also matters, but it is only useful when paired with domain knowledge and verification.
A practical response is to keep learning in two directions: understand enough about AI systems to specify and check their work, and deepen the human capabilities that cannot be reduced to text generation. Wright’s question—“What can I learn that I didn’t know yesterday?”—is a durable career habit because it applies whether the next change comes from AI, a new medium or a different customer need.
The most defensible conclusion
Wright’s essay does not prove that AI expands every worker’s prospects, and the ILO and World Economic Forum findings do not disprove her experience. They answer different questions. Her story shows how one person used assistance to move attention away from repetition and toward creative, strategic and relational work. The labor evidence shows that exposure is widespread but highest exposure is much narrower, that most occupations still contain human-dependent tasks and that productivity results remain uneven.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesThe useful test is therefore not whether AI “replaces” a job in the abstract. Ask which tasks change, who retains judgment, how quality is verified and what happens to the time that becomes available. When those choices favor learning, autonomy, experimentation and genuine human connection, AI can expand a person’s possibilities. When they simply raise the pace or remove accountability, the same technology can narrow them.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
