Careers and the future of work, five notes on what AI changes for people
The debate about AI and work tends to get stuck on one question, whether a profession will disappear. In the notes below, the author looks at changes that arrive before that. The technical barrier that falls, the skill of defining what to build, jobs that shrink without the profession ending, juniors without a learning path and the value of judgment and accountability. Each note carries its original date and a link to the post. The data is valid as of the date of each note.
The technical barrier fell, the cognitive one became more valuable
January 19, 2026
We are watching the quiet collapse of the divide between technical and non-technical. According to OpenAI data, 75% of enterprise users say they use AI to do tasks they simply could not do before. It is not just speed. It is new capability.
AI has become a skill equalizer and helps most those who had less technical ability. Those whose value was purely technical see their edge shrink. Those who master context, customers and business gain superpowers.
The technical barrier fell. The cognitive barrier became more valuable. 2026 will not reward those who know how to use tools. It will reward those who know what needs to be done.
Source, OpenAI report The state of enterprise AI, December 8, 2025 · Original post
Knowing what should be built
August 18, 2026
Perhaps the most important skill in the AI era is not knowing how to build. It is knowing what should be built. Andrew Ng published his AI Engineering Skills Map, based on an analysis of more than 10,000 job postings, and one of its four competencies is Shaping the Build.
As agents get better at turning a specification into working software, part of human value shifts from execution to defining the problem. That goes beyond engineers.
In the enterprise, I would add one more layer. It is not enough to evaluate whether the agent executed correctly. We need to evaluate whether that execution produced a result for the business.
Source, Andrew Ng's The AI Engineering Skills Map, The Batch, August 14, 2026 · Original post
AI can cut jobs without eliminating the profession
October 5, 2026
AI does not need to eliminate a profession to reduce its jobs. When a company delivers the same result with fewer hours of human work, hiring needs can change, even if the profession still exists.
The International Labour Organization estimated in 2025 that one in four workers is in an occupation with some exposure to generative AI, and considers transformation of roles more likely than elimination. But transforming a role can also change team size.
You cannot assume that all additional productivity turns into layoffs. Nor can you guarantee it will create jobs. For those who lead, responsibility starts before the cut, by preparing people for the work that is changing.
Source, ILO, Generative AI and Jobs, A Refined Global Index of Occupational Exposure, May 20, 2025 · Original post
How we will train the next seniors
October 7, 2026
If AI does the juniors' work, how will we train the next seniors? Researching, preparing a first analysis and solving simple problems are productive activities and also ways of learning.
A Stanford update from August 2026 showed that, in US payroll data, employment of workers aged 22 to 25 in occupations highly exposed to AI was about 19% below the pace of young workers in less exposed occupations. The study does not establish causality, and the adjustment shows up mainly in reduced hiring.
A professional can deliver something excellent with AI and still not know how to recognize when it got something wrong. That is why productivity and development need to be tracked separately.
Source, Stanford Digital Economy Lab, No Widespread Displacement, but the AI Employment Gap for Young Workers Has Widened to 19%, August 12, 2026 · Original post
Judgment, context and accountability
October 8, 2026
Saying that humans will always be indispensable is not a career strategy. The common answer is to develop creativity, critical thinking and empathy. I agree, but I find that answer incomplete. We need to understand where those capabilities create economic value.
In my view, three dimensions deserve attention. Judgment, to decide when there is no clearly right answer. Context, to understand consequences beyond the available information. Accountability, to own what happens after a decision.
There is an important difference between producing an answer and taking responsibility for what happens after it. If AI is changing the cost of executing work, it is also changing the way that work will be valued.
Further reading
- Copilot in SharePoint can now evaluate and improve skills with tests
- Vertical software in the AI era, where the moat is process engineering
- Productivity, knowledge and democratization
Notes originally published on LinkedIn between January 19, 2026 and October 8, 2026, with each post's link next to its note. This collection has been edited and organized with context and sources. The views are those of the author, a co-founder of OperOS AI.