Three years ago, this was still a debate

In April 2023, I wrote an article asking whether artificial intelligence would change the course of history.

At the time, the question was far from settled. ChatGPT had only recently entered the mainstream. Most companies were still experimenting. The dominant conversation revolved around prompts, hallucinations, and whether AI would become useful enough to justify the attention it was receiving.

My view was different.

I believed that AI was not simply another technological trend. I saw it as the convergence of several decades of progress in computing, connectivity, data availability, and machine learning. More importantly, I believed it would eventually become one of the most consequential technologies humanity had ever developed.

Three years later, I believe that more strongly than I did then.

What has surprised me is not that AI has progressed, it is known that innovation always progresses. What has surprised me is the speed at which the conversation has evolved. We have moved from asking whether AI matters to reorganizing companies, investment strategies, and national priorities around it. That transition usually takes decades but in the case of AI, it has taken just a few years.

The mistake is thinking AI is another automation wave

When people compare AI to previous technology revolutions, they often focus on productivity. They assume AI is primarily about doing existing tasks faster or more cheaply.

From my perspective, and based on what previous technological revolutions have taught us, that interpretation is too narrow.

• The printing press amplified the distribution of knowledge.

• The steam engine amplified physical work.

• Computers amplified calculation.

• The internet amplified communication.

• Artificial intelligence amplifies reasoning.

For the first time, we have created systems capable of participating in activities that were previously considered exclusively human: writing, analyzing, designing, researching, coding, and increasingly making recommendations that influence decision making.

This is why AI feels different from previous waves of innovation. It is not simply improving a process. It is beginning to influence the way humans think, create, and solve problems. Sometimes even too much…

The implications extend far beyond productivity metrics. They affect education, management, organizational design, scientific discovery, healthcare, and virtually every field that depends on human knowledge.

We are still underestimating the scale of change

If there is one prediction I feel comfortable making today, it is that we are still underestimating what comes next.

Much of the current discussion focuses on replacing tasks. Will AI replace developers? Will it replace analysts? Will it replace marketers? Will it replace customer support teams?

These questions are understandable, but they miss the larger transformation.

The real impact of AI will not come from replacing individual activities: It will come from changing how organizations are structured, how decisions are made, and how value is created.

The companies that emerge stronger from this transition will not simply be the ones using more AI tools. They will be the ones willing to rethink assumptions that have remained largely unchanged for decades. How many people are needed to perform a function? What should managers actually manage? How should knowledge flow through an organization? What work creates value and what work simply creates activity?

These are not technology questions. They are organizational questions. And AI is forcing us to revisit all of them simultaneously.

The new challenge is understanding what AI is actually doing

One of the most interesting consequences of AI is that it is making performance harder to understand.

Historically, organizations could observe activity and use it as a rough proxy for contribution. The relationship was imperfect, but generally understandable.

That relationship is breaking down: When one person can accomplish the work of many, when AI generates code, documents, analyses, and recommendations and when humans and machines collaborate continuously, traditional ways of evaluating performance become less useful.

The challenge is no longer determining whether AI is being used. Most organizations have already crossed that initial threshold. The challenge is understanding whether AI is creating value, where it is creating value, and what unintended consequences it may be introducing.

The organizations that succeed over the next decade will not necessarily be those that adopt AI the fastest. They will be those that develop the deepest understanding of its impact.

The question has changed

Three years ago, I asked whether artificial intelligence would change the course of history.

At the time, reasonable people could disagree. Today, the evidence is overwhelming.

AI is already changing how software is built, how companies operate, how knowledge is created, and how decisions are made. The effects are visible across industries, geographies, and functions.

The question is no longer whether AI will change history.

The question is whether we will be able to understand the changes it creates.

As AI becomes embedded into every workflow, every decision, and every organization, many of the assumptions we have relied on for decades start to break down. Activity no longer equals contribution. Effort no longer equals output. Traditional ways of measuring performance become less useful every day.

That realization is one of the reasons we started Pensero.