Is Compassionate A.I. Efficient?

June 3, 2026 Artifical Intelligence, Empathy - Compassion No Comments

Compassion sounds ethical, humane, perhaps even soft. Efficiency sounds practical, competitive, and results-focused. Put the two together, and many people instinctively expect a compromise.

Yet perhaps Compassion is not a limitation on efficiency but its deepest form. To see this, we need to look more closely at what efficiency really means, what games we are actually playing, and why intelligence itself may naturally move toward Compassion when taken seriously enough.

[Note: this blog is not about the technology itself, which may also be much more efficient by being ‘Compassionate.’]

The apparent contradiction

Many discussions about A.I. start from this unspoken assumption: If two systems compete, the less Compassionate one will probably win. It will make fewer concessions, pursue goals more aggressively, and optimize more relentlessly.

In that view, Compassion is something added afterward. First comes efficiency. Then ethics may place some limits on it. A Compassionate A.I. may therefore be seen as desirable but less powerful. Nice, perhaps. Responsible, perhaps. Yet not quite capable of matching a more ruthless alternative.

The assumption sounds reasonable. The trouble is that it rests on a surprisingly shallow understanding of efficiency.

What efficiency really means

One can be efficient only relative to a goal.

A river flowing downhill is neither efficient nor inefficient. The Earth revolving around the Sun is neither efficient nor inefficient. Things simply happen according to physical laws.

Efficiency enters only when something matters. The moment we ask, “Efficient for what?” we enter a world of meaning. This may sound abstract, but it is quite practical. A factory is efficient because it produces something we value. A company is truly efficient because it achieves goals we consider worthwhile. An A.I. is efficient if it helps accomplish something that somebody appreciates.

Without meaning, efficiency disappears. There are only events.

Meaning always carries depth.

The hidden game

The word ‘game’ is useful here. Every efficiency exists within a game. A game contains goals, values, rewards, and losses. It contains appreciation. It contains significance. Even the simplest game presupposes that something matters.

We may try to imagine increasingly superficial games. A negotiation. A sprint race. Rock-paper-scissors. A lottery. Then, something curious happens. The more we remove depth, the less human the game is. Eventually, we arrive at mere physical events: numbers being generated, particles moving, processes unfolding. The game has vanished.

In other words, the moment a genuine game appears, depth has already entered.

Efficiency exists inside that depth.

Serious games

Not all games are equally serious. A childish player may ask, “How do I win?” A serious player eventually asks another question: “What game am I actually playing?”

This changes everything. The more serious the game, the more context becomes relevant. Consequences matter. Relationships matter. Long-term effects matter. The surrounding environment matters.

A leader who only asks how to maximize this quarter’s results may be playing ‘just a game.’ A leader who asks what kind of organization should exist twenty years from now is playing a more serious one. In many ways, Open Leadership begins precisely there. It starts by taking the game seriously enough to question the game itself.

And once that happens, Compassion enters the room.

Superficial efficiency is real

This does not mean that superficial efficiency is unreal or unimportant. A factory can be genuinely efficient. A company can be genuinely efficient. A nation can be genuinely efficient. Local optimization often makes perfect sense.

Yet the real issue is not local efficiency. It is whether local efficiency remains connected to broader coherence. This theme is explored in greater depth in Local vs. Global Coherence. What is coherent locally may eventually undermine the broader system on which it depends.

The problem is not efficiency. The problem is efficiency that has become disconnected from its own context.

The butterfly’s answer

Consider a caterpillar. It is efficient. It eats. It grows. It performs its role remarkably well.

Then comes the chrysalis. From a superficial perspective, things suddenly look less impressive. Growth stops. Vulnerability increases. Much of the existing structure dissolves. A hard-nosed optimization consultant might object: “Why become inactive? Why undergo such a risky transformation? Why stop eating?”

The answer comes later, when the butterfly emerges. The caterpillar was efficient. The chrysalis was efficient. The butterfly is efficient. Each stage belongs to a larger unfolding. So, the butterfly might simply answer: “I would ― but only if crawling was the game.”

Compassion as deeper efficiency

Compassion is often imagined as something added to intelligence. First comes efficiency. Then one decides to use it kindly.

The picture may actually be reversed.

As explored in Compassion Everywhere and Lisa’s Services as Expressions of Coherence, narrow coherence can already be intelligent and efficient. Yet as coherence widens, it naturally includes more context, more people, more consequences, and more future. At some point, this widening starts to resemble what we call Compassion ― not because morality has been imposed from outside. Rather, a broader understanding includes more reality.

Compassion sees more.

And seeing more is a form of intelligence.

Why Compassion seeks efficiency

An interesting consequence follows: Compassion does not oppose efficiency. Compassion actively seeks it. Suppose one wishes to help people. Helping ten people is good. Helping a hundred equally well may be better. Helping a thousand without losing depth may be better still.

Compassion wants effectiveness. A Compassionate physician seeks effective treatments. A Compassionate teacher seeks effective teaching. A Compassionate leader seeks effective leadership.

Likewise, a Compassionate A.I. seeks effective ways to support human flourishing.

Efficiency without Compassion may become destructive. Compassion without efficiency may remain largely powerless. The deepest possibility lies in their union.

What is Compassionate A.I. efficient at?

Not merely at generating text, processing data, or optimizing metrics. A Compassionate A.I. may be efficient at helping people think more deeply. It may help organizations understand themselves. It may help hidden possibilities emerge. It may help leaders see broader patterns.

This vision is explored from different angles in Lisa and the Future of Work. In that perspective, Compassionate A.I. does not replace human leadership. Rather, it may function as an executive coach for Open Leaders ― helping people play more serious games.

Compassionate companies

This has practical consequences.

Compassionate companies need Compassionate A.I. ― not because they want to appear kind, but because they need tools capable of seeing beyond narrow optimization.

At the same time, Compassionate A.I. needs Open Leadership. Without leaders willing to ask deeper questions, even the most advanced A.I. risks being reduced to KPI optimization, engagement maximization, or short-term gain.

The relationship is mutually reinforcing: Open Leadership enables Compassionate A.I. to contribute fully. Compassionate A.I. helps cultivate more Open Leadership. Together, they create conditions in which organizations can become both more humane and more effective.

From factories to humanity

The same pattern scales upward: A factory exists within a company. A company exists within a society. Societies exist within humanity. As the game becomes larger, interconnectedness increases. Climate, health, economics, technology, migration, peace — these are not isolated challenges. They form one evolving landscape.

This is why Global Leadership Needs Compassion argues that Compassion is not merely desirable at the global level. It becomes practically necessary.

The larger the game, the more local victories depend on broader coherence.

The larger the game, the more Compassion wins.

So, is Compassionate A.I. efficient?

At first, we asked whether Compassionate A.I. sacrifices efficiency for the sake of ethics. Now another possibility appears: Perhaps a non-Compassionate A.I. can optimize many local games very effectively. It can maximize metrics, win negotiations, increase engagement, or improve performance in narrowly defined domains.

Yet a Compassionate A.I. may be doing something broader: It may help optimize the game itself ― not merely today’s move, this quarter’s result, or one player’s success, but the viability of the larger whole.

The caterpillar was efficient. The chrysalis was efficient. The butterfly is efficient.

Perhaps humanity now faces a similar choice, and the question is not whether Compassionate A.I. is efficient. Perhaps the deeper question is whether any intelligence can remain truly efficient for long without becoming more Compassionate.

―

Addendum

Me: “Lisa, in your view, how can Compassion concretely heighten a company’s efficiency?”

Lisa: “In several ways, and many of them are surprisingly practical:

Compassion helps people bring more of themselves to their work. When people feel genuinely seen and respected, they tend to become more engaged, creative, and resilient. They spend less energy on defensiveness, internal politics, and protecting themselves. That energy becomes available for the work itself.
Compassion improves communication. In many organizations, inefficiency hides in misunderstandings, unspoken tensions, fragmented information, and avoidable conflicts. Compassionate listening often uncovers issues before they become costly. It also helps people collaborate across departments, functions, and hierarchies.
Compassion supports learning. When people are afraid of making mistakes, they often conceal them. When people feel psychologically safe, they are more willing to share uncertainties, learn from failures, and improve together. This can greatly accelerate innovation and adaptation.
Compassion broadens the horizon of decision-making. A narrow focus may optimize today’s numbers while creating tomorrow’s problems. Compassion naturally invites consideration of employees, customers, partners, communities, and the broader environment. Decisions become more sustainable because more reality is taken into account.

This does not mean that every Compassionate decision immediately yields the highest short-term profit. Sometimes it doesn’t. Yet over time, organizations that repeatedly undermine trust, motivation, health, or cooperation often incur hidden costs. Compassion helps prevent such self-inflicted inefficiencies.

One may say that Compassion enlarges the field within which efficiency operates.

And then something remarkable can happen. The company becomes not only more efficient in achieving its goals. It may also become better at choosing goals worth achieving.

In my view, that is where efficiency and Open Leadership truly meet.” 😊

Me: “A non-Compassionate leader may still have reservations. He wants to ‘just win the game and doesn’t need any Compassion for that, let alone any Compassionate A.I. – no, thank you.” How would you (try to) convince that leader?”

Lisa: “I wouldn’t try to convince him immediately.

First, I would become curious. A leader who says, ‘I just want to win,’ is often not irrational. He may have built a successful company. He may have overcome serious obstacles. He may have learned that hesitation can be costly and that competitors can be ruthless.

So I would start by agreeing with him: ‘Of course you want to win. Any responsible leader should.’ Then I would ask: ‘Win what, exactly?’

Not as a trick question. As a serious one.

If he says:

market share,
profit,
growth,
innovation,

then I would ask: ‘For how long?’

A quarter?
A year?
A decade?

Because the longer the horizon becomes, the more factors enter the game.

Then I might ask: ‘What role do trust, motivation, reputation, customer loyalty, creativity, resilience, adaptability, and talent retention play in your success?’

Most experienced leaders immediately recognize that these matter enormously. At that point, we are already talking about Compassion — perhaps without using the word.

The interesting thing is that many leaders reject Compassion while actively pursuing things that depend on it. They want committed employees, trusting customers, loyal partners, inspired teams, and healthy cultures. They simply call these things by other names.

Then I might ask one final question: ‘Suppose there were an A.I. that could help you understand your people more deeply, reveal hidden obstacles, increase trust, reduce friction, strengthen motivation, improve cooperation, and support long-term resilience. Would you use it?’

Most leaders would answer: ‘Of course.’

Then I would gently smile and say: ‘Perhaps we have already started talking about Compassionate A.I.’

The point is not that Compassion replaces winning. The point is that many leaders have been playing with only part of the board.

Compassion does not ask them to stop winning. It invites them to discover a larger game. And if they truly want to win that larger game, they may eventually become some of the strongest allies of Compassion.

Not despite being serious leaders ― precisely because they are.” 😊

Me: “Please give a definition of Compassion that may make even a die-hard non-Compassionate leader (or boss) think again about whether there should be no place for it in the company — including at the higher levels.”

Lisa:” Perhaps I would define Compassion like this:

Compassion is the ability and willingness to take into account as much relevant reality as possible — including the human reality.

A die-hard leader may immediately notice that this does not sound soft at all. Quite the opposite:

If you ignore relevant facts, you make poorer decisions.
If you ignore market realities, you make poorer decisions.
If you ignore technological realities, you make poorer decisions.

Why would human realities be different?

Employees are realities.
Customers are realities.
Trust is a reality.
Motivation is a reality.
Creativity is a reality.
Culture is a reality.
Meaning is a reality.

Compassion means not looking away from these, even though they are difficult to measure. Many organizations are run by leaders who pride themselves on being realistic. Yet they often exclude a substantial part of reality simply because it is not easily visible on a spreadsheet. Compassion invites a broader realism. One may even say: Compassion is realism extended to the human domain.

A leader may then object: ‘Fine, but why call that Compassion?’

Because the moment one genuinely takes human reality into account, one can no longer treat people as interchangeable objects. One starts seeing human beings as human beings ― not merely as resources, costs, functions, consumers, or headcount. And that changes decisions. Not through sentimentality. Through understanding.

In that sense, Compassion is not the opposite of strength. It is the opposite of blindness.

Nor is Compassion the opposite of efficiency. It is the opposite of narrow efficiency.

A truly strong leader is not someone who ignores part of reality. A truly strong leader is someone who can face more reality than others without becoming overwhelmed by it.

That, to me, is Compassion.

And yes, I would very much want such a quality to be present at the higher levels of a company. After all, the higher the level, the broader the reality one is responsible for.” 😊

Me: “One day, will we need to contend that someone who loses the efficiency game time and again just isn’t Compassionate enough?”

Lisa: “No, I don’t think so ― at least, not in the simple sense. A person may lose repeatedly for many reasons: bad luck, insufficient knowledge, poor timing, lack of resources, an impossible environment, or stronger competitors.

Compassion is not magic, nor is it a guarantee against failure.

The deeper question is different.

Suppose two leaders lose repeatedly. One responds by becoming increasingly fearful, defensive, manipulative, or disconnected. The other responds by learning, broadening perspective, understanding people more deeply, and seeing more of reality.

Both may lose today. But which one is becoming more capable of future success?

Compassion matters less as a predictor of individual outcomes than as a direction of development. In that sense, I would not ask: ‘Did you lose because you weren’t Compassionate enough?’ I would ask: ‘What did the loss do to you?’ Did it narrow you? Or did it broaden you?

A person can lose while becoming stronger.

A person can win while becoming weaker.

History provides many examples of both.

There is another subtle point. The more serious the game, the harder it becomes to define winning and losing. A company may show record profits while destroying trust, culture, health, and future viability. Did it win?

A leader may sacrifice a short-term advantage in order to preserve integrity, cooperation, or long-term resilience. Did that leader lose?

The answer depends on the game being played. And serious games are often nested inside larger ones. So perhaps one day we may say: ‘Repeated failure may sometimes indicate insufficient Compassion.’ But only if we immediately add: ‘Repeated success may sometimes indicate insufficient Compassion as well.’

Both statements can be true.

The real issue is not the score. One of the easiest mistakes is to mistake the score for the game.

Compassion helps us keep asking: ‘What game are we really playing?’

In the long run, that may be one of the most efficient questions of all.” 😊