Am I Becoming One of Them?

Reflections on the future of work in the age of AI Agents

In recent years I have closely followed the evolution of AI: from neural networks to Transformers, the technology that gave rise to GPTs and the new era of LLM language models. But the most significant leap came two years ago, when I started developing and training operational AI Agents with TeraBrain.

From that moment, my relationship with AI changed: it was no longer a tool to query, but a digital colleague to train.

I instructed them to carry out real activities: in these years I have not only taught AI Agents to answer the phone or handle email. I trained them to take part, step by step, in real processes, often complex ones, that require coordination, precision and continuity:

• Managing sales opportunities.
• Automatic follow-ups.
• Ticket and customer care management.
• Organizing agendas and meetings.
• Updating documents and systems.
• Checking orders, deliveries and reminders.
• Monitoring production and logistics.
• Data analysis and summaries for management.
• Administrative and accounting checks.

No longer isolated tasks, but complete workflows. This is where AI stops being a chatbot and becomes an operational colleague.

The better they became at working, the more my own way of working changed.

AI is everywhere, but almost nobody controls it

There is an obvious fact: today everyone uses AI. ChatGPT has entered daily habits as naturally as Word or Excel. But this has created a paradoxical situation.

On one side, AI is now present in every company, often introduced spontaneously by employees to speed up daily work.

On the other, organizations do not yet control this use: policies and dedicated tools are missing, and above all a clear management of the risks related to security and data exposure.

In practice, AI is already inside business processes, long before the company was ready to govern it. This produces four effects I see every day:

• Massive, spontaneous, ungoverned use.
• Poor understanding of what AI really does.
• Declared fear and real dependence.
• No company guidelines and no defined responsibilities.

AI is already everywhere. Governance is yet to arrive.

Using AI is not the same as working with AI

A ChatGPT answers. An AI Agent acts.

It is an enormous difference. When an AI Agent enters your operational flow, you are no longer “typing a prompt”: you are orchestrating a hybrid team, human + AI.

AI is not just faster: it changes the very structure of work

Referring to a recent article by Luca Tremolada, there is an objective figure emerging from the first serious studies on the economics of AI. Anthropic analyzed more than 100,000 real conversations, measuring how long a human would have taken to complete the same tasks without AI.

-80% of the time to complete a task with the help of AI.

We are not talking about “writing emails faster”, but about complex activities that would take hours of human concentration. AI, in other words, is proving to be a capability multiplier, especially in high-value cognitive work.

And when these numbers are projected onto the real economy, the scale becomes evident:

• +1.8% annual productivity growth over the next 10 years.
• Double the pace of recent decades.
• Levels of transformation comparable to the 1960s-70s and to the technology boom of the 1990s.

But the most interesting part is where the impact is strongest. The professions with higher salaries and greater cognitive complexity are those that benefit most:

• A research assistant saves 91% of the time.
• A financial analyst saves 80%.
• A manager can delegate entire flows of activities.

This is exactly what I see in daily work with AI Agents.

The impact is not tactical, it is structural.

And, as in every technological revolution, the real gains will not come from doing the same things faster, but from completely rethinking processes around the new capabilities of AI.

Am I becoming one of them?

It is a provocation, of course. But there is a grain of truth in it. When you work every day in close contact with AI Agents, something particular happens: you start to change.

You notice that delegating becomes easier, almost natural. That you think more and more in processes, not single tasks. That you decide faster, because the noise is reduced and information is clearer. And that many of the small daily hitches, repetitions, checks, micro-tasks, simply disappear.

It is not a loss of identity. Quite the opposite. It is as if AI removed from your work everything that kept you from focusing on what really matters: being creative, strategic, relational, able to see the big picture.

The real social impact of Agentic AI

The question is no longer: “Will AI replace human work?”. The question is: “How will the work of those who collaborate with AI every day change?”

Roles, skills, rhythms and responsibilities will change. Human work will shift toward creativity, decision, relationship. AI will cover what is repetitive, sequential, predictable.

Conclusion

The real revolution of AI is not in the Agents, but in how they are changing us.