When Empathy Becomes a Feature: Reflections from Dreamforce

It is difficult to explain Dreamforce to anyone who has not stood in the middle of it. For one week each year, San Francisco becomes a stage. Streets close to traffic, buildings light up in Salesforce blue, and the city fills with tens of thousands of delegates moving in purposeful streams between keynotes, cafés and concerts. It feels less like a conference and more like a festival of belief, a gathering built around the promise that technology will make us more connected, more efficient, and somehow more human.

This year, that promise was written in two letters: A I. The message was everywhere. Every presentation, panel and demonstration returned to it. The language was rapturous. We were told that artificial intelligence would empower us to serve better, sell better, and care better. It would allow people to focus on what really matters by handling everything that does not.

Yet the more I heard about empathy, the less of it I could feel. Demonstrations showed systems that could analyse speech, gauge sentiment, and compose the perfect reply before the human in the conversation had even drawn breath. There were charts of emotional accuracy and algorithms that claimed to know what a client was likely to feel before the client had felt it. It was technically brilliant, but strangely hollow.

Dreamforce is extraordinary in its polish. The atmosphere has the energy of a concert and the coordination of a military exercise. The presentations are rehearsed with theatre-level precision. The optimism is infectious. But beneath the surface runs a quiet dissonance. What once relied on instinct and attention has been turned into a function. Selling, which was once a conversation, is now a process of prediction. The people involved are still there, but the exchange itself has lost its texture.

The irony is that the whole event speaks in the language of care. Executives describe their systems as if they were partners in empathy, companions that understand us, anticipate our needs, and help us to be our best selves. Yet this is not understanding in any true sense. It is recognition without thought, a mirror rather than a mind. The system does not know the customer; it knows the pattern the customer fits.

Once, salesmanship was a human art. The best salespeople possessed a kind of quiet intuition. They could sense when a pause mattered, when a tone softened, when a question concealed uncertainty. They listened not just to words but to the rhythm of speech and silence. Their craft lay in attention, and attention is never formulaic.

Now, attention itself is being outsourced. AI copilots whisper advice, suggest phrasing, record sentiment, and calculate the “next best action.” The salesperson becomes an operator following instructions from a machine that claims to understand emotion better than anyone in the room. It is efficient, but it is also bloodless. The human presence remains, but the judgment has gone.

This shift is not confined to sales. Across almost every industry, companies have begun to speak of empathy as something that can be programmed. Hotels and airlines talk about “AI-driven personalisation.” Banks send letters that sound warm but are written by software. Messages arrive that seem thoughtful until you receive the same one twice. The gesture is there, but the grace has gone.

Care cannot be automated because care is not a process. It is an attitude. It begins with noticing, with choosing to see another person as more than a variable in a model. Automation can only simulate that choice. It can reproduce the appearance of sincerity, but not its substance.

Dreamforce embodies this paradox perfectly. It preaches connection while quietly replacing it with correlation. The organisers speak of a future where machines will handle the burdens of work so that humans can focus on meaning. Yet the act of delegation changes the habit itself. The more we let algorithms manage empathy, the less we remember how to practise it.

The danger is not that machines will replace people, but that people will start behaving like machines. We are already rewarded for it. In many workplaces, speed and consistency count for more than discernment. Mistakes are unforgivable, but blandness is safe. The ideal employee is efficient, reliable and interchangeable. It is a short step from that ideal to the kind of service that is technically flawless and emotionally vacant.

Dreamforce does not see this as loss. To its devotees, it is progress. Every advance is framed as liberation. The language is always the same: freeing people from friction, giving them back time, helping them to focus on what matters. Yet something vital disappears in that removal of friction. Perfection is sterile. It is the minor struggle, the small delay, the uneven rhythm of real conversation that gives service its warmth.

What unsettled me most was how easily the vocabulary of feeling has been absorbed into corporate speech. Words such as “kindness,” “understanding” and “trust” now appear in product brochures. There are metrics for empathy and dashboards for compassion. Somewhere along the way, we have begun to confuse the language of care with the presence of it.

Real service is not seamless. It is attentive. The moments we remember are rarely the smooth ones. They are the ones in which someone took the trouble to see, to think, to adapt. Those moments depend on human judgment, and judgment cannot be written into an algorithm because it involves the one thing machines cannot replicate: moral choice.

If there is a lesson from Dreamforce, it is that progress without restraint easily becomes parody. The future will not belong to those who automate empathy, but to those who use technology without surrendering thought. The institutions that endure will be the ones that treat machines as tools, not as masks. They will remember that efficiency is not the highest form of excellence, and that the appearance of care is not the same as the reality of it.

Dreamforce left me with admiration as well as unease. The technology was dazzling, the ambition sincere. Yet walking back through the city that night, past banners still glowing with corporate blue, I kept thinking of something no speaker had said. The purpose of selling, or of any service, has never been prediction. It has been presence: the act of giving one’s full attention to another person.

When that attention is gone, no algorithm, however advanced, will bring it back.