AI, Work, and the Question After “Now What?”
I recently had the opportunity to participate in a panel discussion at Post University with a group of graduate students visiting from Germany. The topic was AI and the changing business environment, but the conversation quickly became much more practical than theoretical.
That was the most encouraging part.
This was not a room asking whether AI mattered. That question is already settled. The students were asking the better questions: What does AI mean for entry-level workers? How should organizations help experienced employees adapt? When should AI not be used? How do we balance innovation with privacy, governance, and trust?
Those questions say a lot about where the conversation is heading.
One of the themes I tried to emphasize is that AI should not be viewed only as a tool for doing the same work faster. That is the obvious use case, and it is where many organizations will start. But the more interesting question is what becomes possible once some of that work becomes faster, cheaper, or easier.
The internet is a useful comparison. In the early days, many organizations thought the web was mainly a better way to publish information or communicate electronically. Then came e-commerce, cloud computing, streaming, digital marketing, remote work, software-as-a-service, and entire business models that were not obvious at the beginning. There was hype. There was overinvestment. There was the dot-com crash. But the crash did not mean the internet was unimportant — it meant many people had misunderstood where the lasting value would come from.
AI may follow a similar pattern.
There will be hype. There will be expensive failed experiments. There will be companies that slap AI onto everything because investors, competitors, or executives expect them to. But the durable value will likely come from organizations that move beyond “How can AI make this faster?” and start asking “Now that this is possible, what should we do differently?”
That is a very different leadership question.
We also discussed the fear that AI will eliminate jobs. That fear is not irrational. Some roles will change. Some tasks will disappear. Some people will be displaced. Pretending otherwise is not useful.
But I think the most common framing is too narrow. It assumes that the current set of human work is fixed, and that AI will simply take portions of it away. History suggests something more complicated. Major technological shifts reduce the labor required for existing work, but they also create new categories of work, new services, new industries, and new ways of organizing society.
The real question is not only what AI will replace. It is what AI will enable.
One example we discussed was Salesforce reportedly redeploying thousands of employees from repetitive internal work into sales roles after expanding its use of AI. Whether every company follows that pattern is another matter, but the management question is the important part. Once AI creates additional capacity, what does leadership do with it? Cut costs? Improve quality? Serve more customers? Build new capabilities? Enter new markets?
That is where strategy begins.
The students also raised very practical concerns about adoption. One student described working with older colleagues who had deep technical knowledge but struggled to identify where AI could help them. That is a real organizational issue.
Often the problem is not that experienced employees cannot use the tools. The tools are not especially hard to get started with — I suspect most employees can write a prompt. The issue is that they may not yet see the connection between the tool and their own work.
That creates a new kind of opportunity for younger professionals. They may become AI translators inside their organizations: people who understand the tools well enough to help subject-matter experts identify useful, safe, and practical applications. That does not mean replacing the experienced worker. It means pairing technical fluency with institutional knowledge.
Another student asked when not to use AI. That may have been the most insightful question of the session.
Much of today's conversation about AI assumes that every problem should eventually have an AI solution. History suggests otherwise. The most successful organizations are rarely the ones that adopt every new technology indiscriminately; they are the ones that understand where technology creates value and where it creates unnecessary complexity. AI is no exception.
If a simple spreadsheet, database query, checklist, or manual process produces a better-controlled and more explainable result, use that. AI is not automatically better because it is newer. In fact, AI can easily introduce false precision, unnecessary complexity, or inconsistent answers. Good judgment includes knowing when the old tool is still the right tool.
That point matters especially in regulated or high-trust environments.
The German and European context added another important layer to the discussion. In the United States, organizations often ask, “What can we do with AI?” In Europe, the question is more likely to begin with, “Are we allowed to do this, and under what conditions?” That is not just bureaucracy. It reflects a different relationship with privacy, data rights, governance, and institutional trust.
Both approaches have tradeoffs. The U.S. may move faster and innovate more aggressively, but with greater risk of missteps, breaches, and public backlash. Europe may move more deliberately, but with stronger expectations around accountability and trust. My guess is that both paths eventually converge, but they may pay different prices along the way.
The strongest takeaway for me was that AI literacy is no longer optional — but AI literacy is not the same thing as prompt tricks.
It includes knowing how to question outputs. It includes understanding where data comes from. It includes recognizing hallucinations. It includes knowing when privacy or confidentiality should stop you from putting information into a tool. It includes being able to explain, defend, and validate the work AI helped produce.
In other words, AI does not eliminate the need to understand your work. It raises the cost of not understanding it.
I shared an example from my own work where I used AI to help write Python scripts for a data analysis project. It worked — until I realized I could no longer fully explain how I had arrived at the result. That was a useful warning. AI had helped me move quickly, but speed had outrun understanding. I had to stop, back up, and have the system walk me through the logic step by step.
That experience changed how I use AI. I no longer think of it as a replacement for expertise. I think of it as a collaborator, research assistant, accelerator, and sometimes a useful skeptic. But I still own the answer.
That may be the most important leadership lesson. AI can draft, summarize, analyze, generate, and recommend. But leaders are still accountable for judgment. They are still accountable for trust. They are still accountable for deciding what should happen next.
It was also fascinating to hear how this generation is thinking about AI and its impact on their careers. Their questions were as insightful as they were grounded in their current experience: practical, skeptical, curious, and very aware that AI is not an abstract future issue. It is already part of the workplace they are entering.
For these AI natives, I believe the opportunities ahead are incredible. They will not just be asked to adapt to AI. They will help define how it is used, where it belongs, where it does not, and what new kinds of work it makes possible.
That may be the most encouraging takeaway of all. The future of AI at work will not be shaped only by the technology. It will be shaped by the people thoughtful enough to keep asking better questions.
Five years from now, I suspect many businesses will realize they got the first phase of AI wrong. They will have focused too much on adopting tools and not enough on redesigning work. They will have focused too much on speed and not enough on purpose. They will have asked “How do we use AI?” before asking “Why are we using it?”
Everyone will be “doing AI.”
The successful leaders will be the ones who answer the better question:
Now what?