Think about what twenty or thirty years of work has actually taught you. The negotiations that nearly fell apart. The office politics you learned to read. The crisis the manual did not cover. The client who was ready to walk, and stayed.
Now the uncomfortable part. When an AI assistant was rolled out to 5,172 customer-support agents, it raised the number of problems they solved per hour by 15% on average. The newest and least skilled agents gained the most, in speed and in quality. The most experienced and most skilled gained a little speed, and their quality dipped slightly.
The tool did not make experience more valuable. It handed some of it to the beginners. Bad news for the old hands. Good news for anyone willing to read on.
That is not the end of the story for experienced people, but it changes the question. Your edge on routine work is shrinking. The edge that is left is real, and you need to know exactly where it lies.
What the tests found
There are two loud stories about AI and work. In one, the machines take everything. In the other, nothing much changes. The evidence so far fits neither.
In a controlled test, 453 college-educated professionals were given writing tasks from their own fields, and half of them were given ChatGPT. Those with it finished 40% faster, and their work was rated 18% better. The gap between the stronger and weaker writers shrank.
The support agents showed the same pattern. The tool helped the newer agents learn the job, and they caught up. Customers were more polite to the agents, and less likely to ask for a manager.
So the honest summary is this. For work that has a pattern, AI is a fast teacher, and it lifts the people at the bottom far more than the people at the top. If your value is that you do routine work well, the tool is coming for your advantage. It will not send a leaving card.
The jagged line
The more interesting finding came from 758 consultants at Boston Consulting Group. On 18 realistic tasks that sat inside what AI handles well, those using GPT-4 completed 12.2% more tasks, 25.1% faster, and to a higher standard. Then they were given a business problem chosen to sit just outside it. There, the consultants using AI were 19% less likely to get the right answer than those working without it.
The authors called it a jagged frontier, and the image is exact. AI is brilliant on one side of a line you cannot see, and confidently wrong on the other. It does not tell you which side you are on. Ethan Mollick, one of the authors, turned what he learned into , a practical guide to working alongside it.
In New York, lawyers suing the airline Avianca filed a motion they had written with ChatGPT. It cited court cases that did not exist, complete with invented quotations. When they asked the chatbot whether the cases were real, it assured them they were. The court dismissed the case and fined the lawyers $5,000. Confident, fluent and wrong is a dangerous combination in a colleague. It is no better in software.
Knowing where that line runs is the experienced person's job. A marketing director can let the tool generate twenty campaign ideas and still know which three will survive this particular client. A project manager can let it track every deadline and still be the one who notices that the real problem is two people who will not talk to each other. The AI provides options. The human provides judgment about which ones are real.
What experience gives you
Some of what you have learned is hard for any tool to copy, because it was never written down.
Reading between the lines. You know when someone agrees in the meeting and opposes it in the corridor. You notice the early warning in a change of tone, the email that is a little too polite. No transcript catches that.
Patterns from context. You have seen a hundred versions of this problem and remember which fixes worked and which only sounded good. That is the crystallised side of intelligence, built from experience. Faster reasoning peaks early, around the late twenties. Knowledge and the judgment built on it keep growing well into later life.
Relationships. People who know your record take your word for things. Former colleagues call you first. That trust took years, and it does not transfer to a newcomer with a better tool.
Crisis instincts. You have been through budget cuts, reorganisations and projects that went wrong, and you know how to decide with half the information.
Here is a useful check, the conversation test. Could you sort out this problem by talking it through with a colleague, reading the people involved, the politics and the things nobody will say out loud? If yes, that is work the tool struggles with.
The job market has been moving this way for decades. Between 1980 and 2012, jobs built on dealing with people grew by nearly 12 percentage points as a share of the American workforce, while maths-heavy jobs with little contact shrank. The fastest growth was in jobs that needed both.
The brain is on your side here. Across 48,537 people, different abilities peaked at very different ages: some around the end of school, some in early adulthood, and others not until the forties or later.
Use the tool
None of this means ignoring AI. The consultants who did best inside the frontier were the ones using it. Your job is to become the sophisticated user, the one who knows what to ask for, what to check, and when to throw the answer away. Think of it as a gifted intern who never says "I don't know".
Use it on your own work every week. Give it the dull jobs first: the first draft, the summary, the spreadsheet. Then give it a problem you already know the answer to, and watch where it goes wrong. You are mapping the jagged line for your own field. Paul Daugherty and James Wilson's sets out how jobs are being redesigned around people and AI working side by side.
And do not try to out-code a 25-year-old. You will lose, and you do not need to win. Compete on judgment, on knowing which answer fits this client, this team and this moment.
Age bias is real
It would be nice to say experience always wins on the CV. It does not.
In a test with more than 40,000 job applications, sent with otherwise similar CVs at different ages, there was robust evidence of age discrimination in hiring against older women, especially those close to retirement age, and considerably less against older men. Unfair, and worth planning around.
So plan for it rather than pretend it away. Lead with outcomes, the problems you solved and what they were worth, rather than a long list of job titles and years. Use your network, because a recommendation from someone who knows your work gets past the filter that a CV meets. And consider contract or consulting work, where people are buying a result and may care less about your birth year.
If fear of getting it wrong is what keeps you from applying at all, Why Your 'High Standards' Are Holding Back Your Career has the difference between standards that help and the fear that does not.
Starting something
If no one will hire you for what you know, you can hire yourself.
The myth says the successful founder is in their twenties. The data say otherwise. Among the fastest-growing new firms in the United States, the one in a thousand that took off, the average age of the founder was 45. The pattern held in high technology and in the famous start-up hubs. And prior experience in the same industry predicted far greater success. The market rewards what you know, even when a hiring form does not.
How to Rewire Your Brain and Change Limiting Beliefs has Ray Kroc, in his fifties and selling milkshake mixers, walking into the McDonald brothers' restaurant in 1954.
Staying in the game pays a second way. Using surveys of older people in the US, England and 11 European countries, early retirement had a clear negative effect on mental sharpness in the early sixties. The authors used national pension rules to tell cause from coincidence, and the retirement itself came out as part of the reason. Work that stretches you keeps the mind stretched.
The worry has a cost too
Living with the fear of being replaced is not harmless. Across 13 sets of findings that measured both, job insecurity and unemployment were each strongly tied to worse mental health, and the insecurity was more strongly linked to physical symptoms. The not-knowing wears on you as well as the losing.
Worry without a plan keeps you in that state. A plan, even a small one, moves you out of it.
Your plan
Audit what you know. Write up the hard problems you have solved as short case notes: the situation, what you did, what it was worth. Focus on the ones where judgment, people or context made the difference. These are your evidence.
Reposition. Rewrite your profile and CV around outcomes and difficult problems, not years served. Say what you are good at in the AI era: deciding, persuading, fixing what the tool gets wrong.
Learn on purpose. Pick one AI tool used in your field and get properly good at it, the way you once got good at your job. Hard learning keeps working at any age: the rewiring article above has the older adults whose memory improved after three months of learning quilting or digital photography. Learning: The Key to Personal and Professional Growth, linked below, has the methods that make new skills stick.
Do not undersell yourself. Playing down your experience to seem cheaper wastes your main advantage. Cheap is not a selling point for judgment. Look for the roles and clients that pay for lower risk and better decisions.
Do not take a big step back. A move that cuts your responsibility and pay sharply can be hard to climb out of. Be patient and selective.
Use contract work as a bridge. It lets you prove your value somewhere new, see several organisations from the inside, and choose where to land. Scott Barlow's is built on testing a new direction in small ways before you leap. Powerful Signs You're on the Right Life Path, linked below, has what to weigh before you leave what you have.
Irreplaceable, honestly
The truth about AI and experience is less flattering than the usual pep talk, and more useful. The tool narrows the gap on routine work. It does not know where its own line runs, and it has none of your relationships, your scars or your sense of what people really mean.
Stop waiting for the perfect moment to reposition. The tools are already here, and the people who learn where they fail will be the ones others call. Be the person who knows when the machine is wrong.
This week, write up three hard problems you have solved, and for each one, note what an AI would have got wrong.
Thinking of changing direction? Powerful Signs You're on the Right Life Path — the regrets that last longest, what each decade asks, and what to weigh if you are near fifty.
Need to learn the new tools fast? Learning: The Key to Personal and Professional Growth — spacing it out, testing yourself, the mistakes you were sure about, and finding someone a few steps ahead.
Know someone over 40 who is worried AI will make them irrelevant? Send them this, and ask them what the tool would get wrong in their job.
Disclaimer: This article is general information about work and careers. It is not legal or financial advice for your own situation.
Affiliate Disclosure: This article contains affiliate links. As an Amazon Associate, All For Life earns from qualifying purchases, at no extra cost to you. We only link to products we would recommend without the commission.
Sources
Brynjolfsson E, Li D, Raymond L (2025). Generative AI at Work. The Quarterly Journal of Economics 140(2):889–942. doi.org/10.1093/qje/qjae044
Noy S, Zhang W (2023). Experimental evidence on the productivity effects of generative artificial intelligence. Science 381(6654):187–192. doi.org/10.1126/science.adh2586
Dell'Acqua F, McFowland III E, Mollick ER, et al. (2023). Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of Artificial Intelligence on Knowledge Worker Productivity and Quality. Harvard Business School Working Paper 24-013. doi.org/10.2139/ssrn.4573321
Wikipedia. Mata v. Avianca, Inc. en.wikipedia.org
Wikipedia. Crystallized intelligence. en.wikipedia.org
Deming DJ (2017). The Growing Importance of Social Skills in the Labor Market. The Quarterly Journal of Economics 132(4):1593–1640. doi.org/10.1093/qje/qjx022
Hartshorne JK, Germine LT (2015). When Does Cognitive Functioning Peak? The Asynchronous Rise and Fall of Different Cognitive Abilities Across the Life Span. Psychological Science 26(4):433–443. doi.org/10.1177/0956797614567339
Neumark D, Burn I, Button P (2019). Is It Harder for Older Workers to Find Jobs? New and Improved Evidence from a Field Experiment. Journal of Political Economy 127(2):922–970. doi.org/10.1086/701029
Azoulay P, Jones BF, Kim JD, Miranda J (2020). Age and High-Growth Entrepreneurship. American Economic Review: Insights 2(1):65–82. doi.org/10.1257/aeri.20180582
Rohwedder S, Willis RJ (2010). Mental Retirement. Journal of Economic Perspectives 24(1):119–138. doi.org/10.1257/jep.24.1.119
Kim TJ, von dem Knesebeck O (2015). Is an insecure job better for health than having no job at all? A systematic review of studies investigating the health-related risks of both job insecurity and unemployment. BMC Public Health 15:985. doi.org/10.1186/s12889-015-2313-1
Park DC, Lodi-Smith J, Drew L, et al. (2014). The impact of sustained engagement on cognitive function in older adults: the Synapse Project. Psychological Science 25(1):103–112. doi.org/10.1177/0956797613499592