When COVID hit, we rewrote the rules of the workplace almost overnight. Offices closed and kitchen tables became conference rooms. Leaders who had insisted that work could only happen in one place suddenly discovered that their organizations could operate from anywhere. We changed where we worked, when we worked, and how we connected, because circumstances gave us no choice. In the process, we proved that many of the rules we had treated as permanent were really just habits.

Now AI is transforming work again. So why are we trying to fit it into the same workplace we inherited from the past? This time, we do not have to wait for a crisis to force change. We have the opportunity to redesign work intentionally, not simply improve the old model.

For more than a century, the workplace has been shaped by assumptions created for another era: fixed hours, fixed locations, fixed job descriptions and fixed career ladders. Then we added new technology to those old structures and called it transformation. But putting AI into an outdated system will not create the future of work. It will only make the past move faster.

Instead, we should ask what work would look like if we were building it today from scratch, with no legacy systems holding us back. We would not begin with offices, organizational charts or a 9-to-5 schedule. We would begin with people. What can technology do best? What can people do best? How can the combination make every person more capable? And how do we ensure that everyone has access to the tools, training, and opportunities being created? The future of work needs a new operating system.

The traditional workday was built around visibility. People arrived at the same time, worked in the same place, and were often rewarded for being seen. But presence is not the same as performance. If we designed work today, we would organize it around outcomes. Teams would agree on what needs to be accomplished, what quality looks like, and when collaboration is necessary. People would have far more agency over when and how they do the rest.

This could include block scheduling, where employees choose working periods that fit their lives, while meeting clear responsibilities and shared collaboration times. Someone caring for a parent, raising a child, returning to school, or even playing golf on a Tuesday afternoon would operate within the same system. Flexibility would no longer be an accommodation granted to some people; it would be part of how work works for everyone.

From Job Titles to Human Potential

Much of today’s debate asks whether AI will replace jobs. But jobs are collections of tasks, and tasks are not people. If we started from scratch, we would separate work into what AI can do, what people and AI can do together, and what depends on distinctly human strengths. AI can synthesize information, automate repetitive processes and accelerate first drafts. People bring judgment, imagination, empathy, accountability, relationships and the ability to understand what matters.

The goal should not be to eliminate a person when technology can perform part of a role; it should be to redesign that role around where the person can contribute more value. AI should remove the work that limits human potential, not remove the human.

That also means replacing rigid job descriptions with living maps of skills, capabilities, and aspirations. People should be able to move across projects, functions, and opportunities based on what they can contribute and what they are ready to learn, not just the title they already hold.

From Privileged Access to Universal Capability

The most important question about AI may not be what it can do; it may be who gets to use it. If the best tools, training, and opportunities are concentrated among executives, technical teams, and people who already have access, AI will accelerate existing advantage. The new divide will not simply be between people who have AI and people who do not; it will be between people who know how to use it with confidence and purpose, and people who were never given the opportunity to learn.

In a workplace built for today, every employee would have an AI partner. Everyone would receive role-specific training during paid working hours, practical support, and the freedom to experiment without being penalized for learning. Access to AI would be treated as workplace infrastructure, not an executive perk.

AI could also make access to opportunity more transparent. Instead of relying on who is visible, who knows whom, or who feels confident enough to raise a hand, organizations could continuously match people with projects, mentors, sponsors, learning experiences, and leadership roles based on their skills, interests, and potential.

But technology cannot become an invisible gatekeeper. If AI influences hiring, compensation, scheduling, performance, and/or promotion, people should know how it is being used. Decisions must be explainable, auditable, and open to human review. AI should widen the door, not quietly decide who gets through it.

From Career Ladders to Opportunity Marketplaces

The career ladder assumes that progress happens in one direction. But careers today are not linear , and people should not have to wait for the position above them to become available before they can grow. An AI-native workplace would operate more like an opportunity marketplace. Employees could discover short-term projects, build new capabilities, contribute across teams, and explore different paths without leaving the organization. Growth could come through expertise, innovation, client impact, mentorship, and/or leadership, not only through managing more people.

Learning would no longer sit outside the job. Every role would include protected time to develop new skills. Each employee could have a personalized growth map showing what they know, where they want to go, and which experiences could help them get there. Companies would be measured not only by how much productivity they extract from people, but by how much potential they develop in them.

From Productivity to Possibility

AI will undoubtedly make work faster. But speed is not the highest ambition. Before companies decide how to measure the return on AI, they should decide what they want AI to create. Is it reducing costs? Improving decisions? Expanding access? Increasing earning power? Helping people move into better roles? Creating more time for innovation, connection, and care?

The answer should include both performance and possibility. This is Double ROI: return on investment and return on impact. Organizations should measure productivity and financial value alongside who has access to AI, who is advancing, whose skills are growing, and whether people have more agency over their work and their lives.

AI should not determine who matters; it should help us recognize how much more every person can contribute. We cannot build the workplace of the future by adding AI to the power structures of the past. We have a once-in-a-generation chance to build work again, around outcomes instead of hours, capabilities instead of titles, opportunity instead of proximity and access for everyone.

If we build it intentionally, AI can make every person more capable and every kind of talent more visible. That is how we create a future of work where possibilities become probabilities.