Sanjay Mohindroo

Sanjay Mohindroo

What can Hindu philosophy teach us about AI ethics, consciousness, accountability, and building a sustainable Human-AI equilibrium?

What Ancient Indian Philosophy Can Teach Us About AI

A few years ago, an AI conversation would have been about algorithms, data, chips, and computing power.

Today, one of the world's leading AI companies is reportedly flying a Hindu monk to San Francisco to discuss morality, consciousness, and how an AI system should behave.

That should make every CEO and board member pause.

Anthropic, the company behind Claude, invited Swami Sarvapriyananda, a Vedanta monk, to its headquarters for private discussions on AI ethics, morality, consciousness, and Claude's training. According to NDTV, the discussions also brought together thinkers from other religious traditions and specialists in areas such as mental health.

The obvious reaction is to ask: Why Hindu philosophy?

I think that is the wrong first question.

The better question is: 

What can a 2,000-year-old philosophical tradition contribute to a problem that Silicon Valley cannot solve with engineering alone?

And that question takes us somewhere much more interesting.

The conventional wisdom is wrong

The conventional wisdom says AI alignment is primarily a technology problem.

Make the model safer. Add guardrails. Improve testing. Control the agents. Monitor the outputs. Keep a human in the loop.

All of that matters.

But none of it answers the most fundamental question:

Aligned to what?

Humanity doesn't have one universally accepted definition of morality, purpose, responsibility, or even consciousness.

AI is therefore exposing a problem that technology has largely avoided: we have become very good at building powerful tools without reaching agreement about what those tools should ultimately serve.

That is why philosophy has suddenly become relevant to the AI conversation.

And this is where Indian philosophical traditions deserve a closer look.

Not because they are "more advanced" than other religions.

Not because ancient India secretly possessed modern science.

And certainly not because Hindu philosophy has all the answers.

The reason is simpler.

Some Indian philosophical traditions spent extraordinary intellectual effort examining questions about consciousness, selfhood, knowledge, perception, action, responsibility, and the purpose of human existence.

Those happen to be precisely the questions AI is now forcing back onto the boardroom table.

Don't confuse philosophy with technology

There is a dangerous leap that should be avoided.

Ancient India made important contributions to mathematics, astronomy, linguistics, medicine, and other fields. Classical Indian philosophical traditions also developed sophisticated theories of knowledge and cognition. The Stanford Encyclopedia of Philosophy describes Indian epistemology as a rich tradition spanning almost twenty centuries, with competing schools developing detailed theories of how knowledge is generated and justified.

That is historical fact.

It does not mean ancient India had artificial intelligence, modern physics, aviation, or nuclear technology.

Claims like that require evidence, not cultural pride.

The more interesting proposition is different:

A civilization does not need to have invented modern technology for its philosophy to become relevant to modern technology.

That distinction matters.

Why consciousness changes the conversation

Consider the question Anthropic reportedly explored with religious and philosophical thinkers:

Could increasingly sophisticated AI systems become conscious?

We do not have a settled scientific answer.

And intelligence is not the same thing as consciousness.

A system can produce extraordinary language, reason through complex problems, recognize patterns, and behave as though it understands something without establishing that it possesses subjective experience.

Advaita Vedanta offers a radically different starting point.

In Advaita, consciousness is not simply a property attached to a physical organism. The tradition identifies the deepest self, Atman, with Brahman, ultimate reality, and treats consciousness as fundamental to what a person is.

You don't have to accept that metaphysics to see its relevance.

It forces a question modern AI discussions sometimes skip:

Are we assuming that intelligence and consciousness are the same thing?

If they are different, then building increasingly intelligent machines does not automatically bring us closer to building conscious machines.

That distinction could become enormously important.

The first principle: Dharma, not delegation

For business leaders, one of the most useful concepts is dharma, broadly understood as duty, responsibility, and right conduct, although its meaning varies considerably across Indian traditions and contexts.

AI is rapidly changing who performs work.

That does not answer who owns the outcome.

Imagine an enterprise workflow in which:

1.   One AI agent researches.

2.   Another analyzes.

3.   A third recommends.

4.   A fourth executes.

The workflow can be almost completely autonomous.

But accountability cannot be autonomous.

Someone still has to own the consequence.

That gives boards a simple principle:

AI can inherit execution. It cannot inherit accountability.

This becomes especially important as agentic AI moves into underwriting, finance, healthcare, security, procurement, compliance, and customer decisions.

The question for the board should not simply be, "Can the agent perform the task?"

It should be:

"Who is accountable when the agent is wrong?"

That is a dharma question expressed in modern corporate language.

The second principle: Karma means looking beyond the KPI

Karma is frequently reduced to the phrase "what goes around comes around."

The deeper idea is about action and consequence.

That has enormous relevance to AI economics.

A company deploys AI and cuts operating costs by 20 percent.

The spreadsheet looks excellent.

But what happened underneath?

Did the company also lose experienced employees?

Did institutional knowledge walk out the door?

Did junior employees lose opportunities to learn?

Did managers become dependent on systems they don't understand?

Did customers receive faster service but less human judgment?

Did the organization become more efficient and less resilient at the same time?

The immediate financial result is one part of the story.

The second-order consequences are the other part.

So, a board should add another question to every major AI investment:

What consequences are we creating that aren't appearing in the ROI calculation?

That is a very different way of measuring transformation.

The third principle: Ahimsa and the hidden cost of efficiency

Ahimsa, commonly associated with non-harming, provides another useful lens.

The obvious interpretation is physical harm.

For AI governance, the broader interpretation is more interesting.

What if harm is cognitive?

What happens when people stop remembering because AI remembers for them?

What happens when employees stop learning because AI produces the answer?

What happens when executives stop challenging assumptions because the dashboard looks authoritative?

What happens when organizations eliminate entry-level roles and then discover five years later that they no longer have a pipeline of experienced professionals?

These are not necessarily AI failures.

They are AI deployment failures.

Technology can make a process more efficient while making the organization weaker.

That is why the right objective isn't maximum automation.

It is optimal augmentation.

The fourth principle: Aparigraha and the question of "enough"

This may be the most uncomfortable idea for the technology industry.

Aparigraha, often translated as non-possessiveness or restraint in accumulation, challenges the assumption that more is always better.

AI strategy currently rewards more:

More automation.

More agents.

More data.

More compute.

More intelligence.

More processes automated.

But should a company automate everything it can?

No.

The relevant question is:

What should we automate?

Some activities are repetitive and should probably disappear.

Others develop judgment, institutional knowledge, relationships, creativity, and leadership.

Those may need to remain human even when AI could technically perform them.

The objective should therefore not be:

"How many humans can we remove?"

It should be:

"How much human capability can we create?"

That is a radically different AI strategy.

The fifth principle: Viveka, the discipline of discernment

There may be no more important human capability in the AI era than discernment.

AI can produce an answer in seconds.

That does not make the answer true.

Classical Indian philosophy developed extensive debates around pramāṇa, or legitimate means of knowledge, asking how cognition becomes justified and how genuine knowledge can be distinguished from error. Nyāya, in particular, developed detailed theories concerning perception, inference, testimony, and false cognition.

That is remarkably relevant to generative AI.

We are entering a world where machines can produce plausible explanations faster than humans can verify them.

The scarce resource will not be information.

It will be judgment about information.

Boards should therefore ask a different question when evaluating AI adoption:

Are we increasing the organization's intelligence, or merely increasing the volume of machine-generated information?

Those are not the same thing.

The sixth principle: Loka Sangraha and the purpose of automation

There is another idea worth bringing into the conversation: loka sangraha, broadly associated with sustaining social order and acting for the welfare of the larger community.

This creates a useful challenge to purely cost-driven AI transformation.

Suppose two companies achieve the same financial benefit.

Company A uses AI primarily to eliminate labor costs.

Company B uses AI to increase employee productivity, retrain its workforce, and redirect people toward higher-value activities.

The financial spreadsheets may look similar.

The societal consequences are not.

For boards, the question becomes:

Does our AI strategy strengthen the system in which our company operates?

Employees are part of that system.

Customers are part of it.

Communities are part of it.

Suppliers are part of it.

The workforce we are training today may be the leadership pipeline we need tomorrow.

A practical framework for boards

None of this requires a board to become a scholar of Vedanta.

The ideas can be converted into six practical questions.

1. Dharma: Who owns the outcome?

For every autonomous AI process, identify the human executive accountable for the result.

2. Karma: What are the second-order consequences?

Measure not only cost savings, but workforce, customer, knowledge, resilience, and reputational effects.

3. Ahimsa: What capabilities could we unintentionally damage?

Identify cognitive, organizational, and human harms created by excessive automation.

4. Aparigraha: What is enough automation?

Do not automate simply because technology makes automation possible.

5. Viveka: How will humans challenge machine-generated knowledge?

Build verification and judgment into critical decisions rather than treating AI output as authority.

6. Loka Sangraha: Does AI strengthen the larger ecosystem?

Evaluate whether the transformation builds sustainable organizational and societal capability.

This is not a religious governance framework.

It is a human governance framework inspired by philosophical ideas.

Does this prove ancient India was scientifically more advanced?

No.

And we should resist that temptation.

Ancient India undeniably produced major intellectual achievements. Indian traditions contributed to mathematics, astronomy, linguistics, medicine, and philosophy, and the historical development of science on the subcontinent is well documented.

But saying that ancient Indian thinkers developed sophisticated theories of consciousness does not prove that they anticipated modern neuroscience.

Likewise, recognizing the sophistication of Indian mathematics does not mean ancient India possessed modern physics.

The stronger argument is more subtle.

Ancient Indian intellectual traditions may have explored some dimensions of the human mind and human conduct with a depth that modern technological civilization has not always prioritized.

And AI is forcing those dimensions back into the conversation.

The real opportunity

The future of AI will not be decided only by engineers.

It will be shaped by philosophers, psychologists, economists, ethicists, religious thinkers, policymakers, executives, and ultimately society.

That is why Anthropic's reported engagement with religious and philosophical thinkers matters. It suggests that the AI industry is beginning to recognize a fundamental limitation of pure engineering: you can build an extraordinarily capable machine without answering what that capability should ultimately serve.

Hindu philosophy does not have the answer.

Neither does Christianity.

Neither does Buddhism.

Neither does Western philosophy.

No single tradition does.

But perhaps the future requires us to stop looking for one answer.

Instead, we should build a global conversation around the oldest questions humanity has asked about consciousness, responsibility, knowledge, restraint, purpose, and the good society.

That is the real opportunity.

The objective should not be to make machines more human.

Nor should it be to make humans more machine-like.

The objective should be to build a society in which machines amplify human capability without eroding human agency, judgment, responsibility, and purpose.

That is what I would call Human-AI equilibrium.

And perhaps the irony of the AI revolution is this:

The more intelligent our machines become, the more seriously we may have to ask what it actually means to be human.

Do you believe ancient philosophical traditions have a meaningful role to play in shaping the future of AI, or should we leave questions of AI ethics entirely to modern science and technology?

If this perspective resonates, subscribe to TechnologyTrends or share your experience in the comments.

Sanjay Mohindroo

AI ROI is not just headcount reduction. Learn how boards can measure the capacity, growth, resilience, and capability AI creates across the enterprise.

The AI Cost Paradox.

 Why Cutting Headcount Is the Wrong Measure of AI Success

A company invests $100 million in AI and cuts $150 million from annual workforce costs.

On paper, that looks like a triumph. In the boardroom, it may be the beginning of a much more expensive mistake.

The conventional wisdom I want to challenge is simple: if AI can do the work, the quickest way to prove ROI is to remove the people who used to do it.

That logic is financially neat, operationally dangerous, and strategically incomplete.

The real question is not how much work AI can eliminate. It is how much additional economic capacity AI can create, and whether the enterprise can convert that capacity into growth, resilience, better decisions, faster execution, and stronger competitive advantage.

That is the AI cost paradox.

The Easiest AI ROI Is Often the Narrowest

Boards are right to ask where the return is.

AI investment is no longer experimental. Companies are spending heavily on infrastructure, models, cloud capacity, data platforms, applications, talent, and increasingly autonomous agents. At some point, the expenditure has to translate into measurable business value.

The problem is that the easiest value to measure is usually labor cost.

If AI allows a process to be run by six people instead of ten, the arithmetic is immediate. Four salaries disappear. The savings drop into the cost base. The business case becomes visible.

But the balance sheet does not capture what may have disappeared with those four people.

Experienced employees carry accumulated knowledge about customers, products, exceptions, workarounds, supplier behavior, regulatory interpretation, system limitations, and the history behind decisions that make little sense when viewed only through current documentation.

That knowledge is rarely stored cleanly in a database.

So, the board should ask a second question before celebrating the savings: what capability did we remove along with the cost?

AI Creates a Capacity Dividend

This is where the economics of AI need to be reframed.

AI value should not be treated as:

AI value = labor cost eliminated

A better boardroom equation is:

AI value = cost eliminated + cost avoided + productivity capacity created + revenue enabled + quality improvement + risk reduction + innovation value, less AI investment, transformation cost, and AI-related risk.

That is a much harder equation to manage.

It is also much closer to the truth.

Suppose AI releases 500,000 hours of human capacity across an enterprise. Those hours are not worth zero because payroll has not fallen.

They may represent additional customer capacity, shorter product cycles, more sales activity, faster research, stronger risk management, better service, and projects that were previously uneconomic.

I call this the capacity dividend.

The strategic question is not whether the dividend exists. It is what management chooses to do with it.

Redeployment Before Reduction

A board should insist that every significant AI automation program answer five questions in sequence.

1. What work is being eliminated?

Be specific. Reporting, reconciliation, data entry, document processing, routine analysis, customer administration, or some other defined activity.

2. How much capacity is being released?

Do not stop at FTEs. Measure hours, cases, transactions, customer interactions, engineering output, projects, product cycles, or sales capacity.

3. Where can that capacity create more value?

This is the critical step. Can the released capacity be directed into growth, innovation, customer acquisition, product development, operational improvement, or risk management?

4. What skills are required for that move?

This turns AI from a technology program into a workforce transformation program.

5. What capacity remains genuinely surplus?

Only now should structural workforce reduction become the default discussion.

The principle is simple: redeployment before reduction.

This is not an argument against reducing headcount. Some roles will become economically unnecessary. Some activities will disappear. Some organizations will genuinely need fewer people.

The point is sequencing.

Workforce reduction should be an outcome of transformation, not the definition of transformation.

The Biggest Hidden Risk Is Capability Debt

There is another reason boards should be cautious about aggressive AI-led downsizing.

AI is entering knowledge work directly.

Entry-level analysts, researchers, coordinators, administrators, and junior professionals spend much of their time gathering information, building spreadsheets, preparing presentations, performing basic analysis, documenting processes, and responding to routine requests.

AI can increasingly perform a significant share of that work.

The productivity upside is obvious.

The less obvious problem is that these tasks have historically been part of the apprenticeship system through which judgment develops.

If AI does the junior work, where do tomorrow's experienced professionals come from?

The same issue appears in middle management.

A large portion of traditional management consists of collecting information, consolidating reports, tracking performance, coordinating teams, preparing presentations, monitoring projects, communicating priorities, and escalating exceptions.

AI and agentic systems can automate more of that coordination layer.

That may allow organizations to operate with fewer managerial layers, which could be a major advantage.

But if companies remove junior roles and middle-management layers without redesigning how expertise is developed, they may optimize today's cost structure at the expense of tomorrow's operating capability.

That is capability debt.

Like technical debt, it is easy to create and difficult to see until the organization needs the capability it has quietly removed.

AI May Make Expertise More Valuable, Not Less

There is a popular assumption that more capable AI automatically reduces the value of experienced people.

I think the opposite may often be true.

The more powerful the machine becomes, the more valuable the people may become who understand when not to trust it.

Experienced employees know which customers matter disproportionately, which process exceptions are normal, which supplier promises should be discounted, which system documentation is wrong, which regulatory interpretations have practical consequences, and which apparently logical recommendation will fail in the real world.

AI can process enormous volumes of information.

Information is not the same as institutional understanding.

This changes the role of expertise.

The future expert may do less manual analysis and more supervision, exception handling, judgment, orchestration, and decision-making. AI does not necessarily remove expertise. It can increase the leverage of expertise.

For senior leaders, that is a far more interesting economic proposition than replacing skilled people with cheaper automation.

The Incentive Problem Most AI Strategies Ignore

There is also a behavioral issue that receives too little board attention.

Imagine an employee discovers that AI can eliminate 40 percent of their workload.

If the implied corporate message is, "Thank you, your job is now 40 percent redundant," what incentive does that employee have to identify the next automation opportunity?

Very little.

The rational behavior is to protect the work.

That is exactly the opposite of what an AI transformation requires.

Organizations need a different implicit contract: if you automate low-value work, we will help you move toward higher-value work.

Employees should be rewarded for finding automation opportunities, improving processes, adopting AI, increasing productivity, developing new skills, and creating new business value.

That turns the workforce from a target of automation into a participant in transformation.

It also makes adoption far more likely.

The Board Needs a Different AI Scorecard

If headcount reduction is the main measure of AI success, management will naturally optimize toward headcount reduction.

Boards need a broader scorecard.

I would group it into five categories.

First, financial value: AI investment, run-rate savings, avoided hiring, revenue enabled, margin improvement, and payback period.

Second, productivity: output per employee, cycle time, throughput, hours released, customer capacity, and engineering capacity.

Third, capability: AI proficiency, critical-skill coverage, internal mobility, expert retention, knowledge capture, and succession depth.

Fourth, transformation: processes redesigned, agent autonomy, new products, new services, and new revenue streams.

Fifth, risk: AI incidents, model failures, compliance issues, human override, AI-off resilience, and concentration risk.

This gives the board a more complete view of whether AI is making the enterprise cheaper, more capable, or both.

Those are not the same thing.

The New Executive Equation

A useful way to think about the quality of AI transformation is:

AI Autonomy × Business Value × Human Capability Retention

If autonomy increases but business value does not, the organization has created technological complexity.

If business value rises but human capability collapses, the organization may be consuming its future.

If human capability is preserved but AI adoption remains weak, the organization risks protecting yesterday's operating model.

The goal is the intersection.

This is why the strongest AI strategy is not simply a technology strategy.

It is a capital allocation strategy, an operating model strategy, and a workforce strategy at the same time.

The Counter-Argument.

Surely Some Companies Simply Need Fewer People?

Yes.

Any serious discussion of AI has to acknowledge that some work will disappear and some organizations will require fewer employees.

AI will reduce repetitive administrative work. Some coordination layers will become thinner. Some knowledge-work roles will be redesigned beyond recognition.

There is no strategic virtue in preserving work that no longer creates value.

But that is not the same as assuming that the first visible productivity gain should be converted into layoffs.

The better sequence is:

Automate. Measure. Redeploy. Upskill. Transform. Grow. Then ask what capacity is genuinely surplus.

This sequence forces management to test whether released capacity can produce a higher return somewhere else in the enterprise before treating it as waste.

That is simply disciplined capital allocation.

The Real Competitive Advantage Is Organizational Multiplication

The next phase of AI will separate companies into two broad models.

The first will use AI primarily as labor substitution.

They will automate work, remove roles, reduce cost, and improve short-term operating efficiency.

That can be a valid model.

The second will use AI as capability multiplication.

They will automate low-value work, release human capacity, redeploy people toward higher-value activity, redesign roles, increase productivity, and make new products, services, or markets economically viable.

They may still end up with fewer people.

But the reduction will be a consequence of a more capable operating model, not the objective itself.

That distinction matters because the most powerful enterprise may not be the one with the smallest workforce.

It may be the one where every person has dramatically more capability because of the intelligence working alongside them.

The Board Question That Matters

Perhaps the most useful question a board can ask is this:

If AI gave us 30 percent more organizational capacity without increasing our workforce, what would we do with it?

If the answer is, "reduce the workforce by 30 percent," then AI is being treated primarily as a cost program.

If the answer is, "serve more customers, develop new products, improve quality, accelerate innovation, strengthen resilience, enter new markets, and raise the productivity of our people," then AI is being treated as a transformation platform.

That is the choice.

AI will eliminate work. Some jobs will disappear. Some layers will thin. Some traditional career paths will change permanently.

But eliminating work is not the same as eliminating value.

The people inside an enterprise are not just a cost line. They are accumulated capability.

The objective of AI should not be to make the organization smaller.

It should be to make the organization more capable.

What is your board measuring today: headcount removed, or capability created?

If this is a debate happening inside your organization, subscribe to TechnologyTrends or add your perspective in the comments.

Sanjay Mohindroo

AI is forcing humanity to reconsider consciousness, responsibility, purpose, and human agency. What can ancient philosophical traditions teach us about living with intelligent machines?

Artificial intelligence is one of the most consequential technologies humanity has ever created.

Yet much of the conversation around AI remains remarkably technical.

How intelligent will machines become?

How quickly will AI agents improve?

Which jobs will disappear?

How much productivity can we gain?

How do we make AI safer?

These are important questions. But they may not be the most important ones.

There is another set of questions that predates computers, algorithms, and even modern science:

What does it mean to be human?

What is consciousness?

What is knowledge?

What makes an action right or wrong?

What responsibilities come with power?

What should humans choose not to do, even when they can?

And perhaps most importantly:

What kind of society do we want to build?

These questions have occupied philosophers, theologians, scholars, and thinkers for thousands of years.

AI has simply given them a new urgency.

AI Is More Than a Technology Question

We often approach technological problems with technological solutions.

If AI makes mistakes, improve the model.

If AI produces bias, improve the training data.

If AI creates security risks, build better controls.

If AI takes jobs, redesign the workforce.

All of these responses have value.

But there is a deeper layer.

Suppose an AI system can perform a task better, faster, and more cheaply than a human.

Does that automatically mean we should automate it?

Not necessarily.

There is a difference between capability and wisdom.

Technology answers the question:

Can we do this?

Philosophy asks:

Should we do this?

And civilization depends on the second question as much as the first.

Why Look Backward?

Ancient philosophical traditions were developed long before artificial intelligence, but many of them spent extraordinary amounts of time examining questions that AI is now forcing us to confront.

Indian philosophical traditions, for example, explored consciousness, selfhood, perception, knowledge, action, consequence, duty, and liberation.

Buddhist traditions developed sophisticated thinking around perception, suffering, attachment, and the constructed nature of the self.

Greek philosophy examined reason, knowledge, virtue, ethics, and the good life.

Christian, Jewish, and Islamic traditions developed extensive bodies of thought around human dignity, moral responsibility, knowledge, stewardship, law, and the relationship between human beings and the divine.

Jain philosophy placed extraordinary emphasis on non-harm and intellectual pluralism.

Confucian thought explored relationships, social responsibility, duty, order, and human flourishing.

These traditions are profoundly different.

They should not be collapsed into one generic concept called "ancient wisdom."

But they share something important.

They took seriously the question of how human beings should live.

That is precisely the question AI is beginning to complicate.

What Indian Philosophy Can Contribute

Consider several ideas from Indian philosophical traditions.

Dharma: Capability Does Not Eliminate Responsibility

Dharma is a complex concept with different meanings across Indian traditions and contexts, but broadly includes ideas of duty, conduct, responsibility, and living appropriately within a moral and social order.

Applied cautiously to AI, it raises an important question:

If a machine performs an action, who remains responsible for the consequences?

An autonomous agent may research.

Another may analyze.

Another may recommend.

Another may execute.

But distributing execution across machines does not necessarily distribute accountability.

Someone still owns the outcome.

The more autonomous our systems become, the more important human responsibility becomes.

Karma: Look Beyond the Immediate Result

Karma is often reduced in popular culture to "what goes around comes around."

The philosophical concept is considerably richer.

One useful way to bring the idea into an AI discussion is to focus on action and consequence.

An AI deployment may produce an immediate benefit.

It may reduce costs.

Increase speed.

Improve efficiency.

But what happens next?

Does it weaken human expertise?

Create organizational dependency?

Remove opportunities for people to develop judgment?

Concentrate decision-making power?

Create second- or third-order risks that were invisible in the original business case?

AI strategy therefore needs something more than an ROI calculation.

It needs consequence thinking.

Ahimsa: What Should We Protect?

Ahimsa is commonly associated with non-harm.

In an AI context, that idea can be extended beyond physical harm.

What about cognitive harm?

What happens if people gradually stop exercising skills because machines always perform them?

What happens when people stop remembering because search is always available?

What happens when writing, reasoning, navigation, research, and decision-making become increasingly outsourced?

The question becomes:

What human capabilities should we deliberately preserve even when AI can perform them for us?

That is a very different question from simply asking what AI can automate.

Aparigraha: Just Because We Can Automate It...

Aparigraha is often associated with non-possessiveness and restraint.

It offers another useful thought experiment.

Technology companies and organizations naturally pursue capability.

More automation.

More intelligence.

More autonomy.

More scale.

But perhaps the objective should not always be maximum automation.

It might instead be optimal augmentation.

Just because something can be automated does not mean it should be.

Sometimes keeping a human deliberately in the loop is not inefficiency.

It is resilience.

It is accountability.

It is institutional memory.

It is judgment.

Viveka: The Importance of Discernment

Perhaps one of the most important ideas for the AI age is discernment.

AI can produce an answer that sounds authoritative without necessarily being correct.

It can generate convincing images, persuasive arguments, realistic voices, fabricated evidence, and plausible explanations.

The challenge is no longer simply finding information.

It is determining what deserves to be believed.

AI therefore increases the value of human judgment rather than eliminating it.

The more synthetic information we encounter, the more important discernment becomes.

But Is Hindu Philosophy "More Advanced"?

This is where intellectual discipline matters.

The existence of sophisticated ideas within Indian philosophy does not demonstrate that Hinduism is inherently more advanced than Christianity, Buddhism, Judaism, Islam, Jainism, Greek philosophy, Chinese philosophy, or other intellectual traditions.

Nor does the existence of ancient Indian achievements in mathematics, astronomy, linguistics, medicine, or philosophy prove claims that ancient India possessed modern technologies such as computers, aircraft, nuclear weapons, or artificial intelligence.

Those claims require evidence.

There is no need to make them.

The more interesting proposition is much more defensible:

Different civilizations developed sophisticated ways of thinking about questions that AI is now making unavoidable.

And perhaps we should listen to all of them.

Intelligence Is Not the Same as Consciousness

There is another question where ancient philosophy becomes particularly interesting.

AI systems can increasingly demonstrate sophisticated forms of reasoning, language generation, planning, and problem-solving.

But does intelligence mean consciousness?

Not necessarily.

An AI can manipulate symbols, generate language, recognize patterns, and perform complex tasks without that establishing that it has subjective experience.

Indian philosophical traditions, particularly Vedanta, contain extensive debates about consciousness and the nature of the self.

These traditions do not provide a scientific answer to whether an AI system is conscious.

But they provide something equally valuable for this conversation:

a reminder that intelligence, selfhood, awareness, and consciousness are not necessarily the same thing.

That distinction could become increasingly important as machines become more capable.

The Bigger Question

Perhaps we are approaching AI from the wrong direction.

Instead of asking only:

How do we make AI more intelligent?

We should also ask:

How do we make humans wiser about intelligence?

Instead of asking:

How much work can machines perform?

Ask:

What human capabilities should remain active and meaningful?

Instead of asking:

How autonomous can AI become?

Ask:

Where should humans retain responsibility and judgment?

And instead of asking only:

How do we control AI?

Ask:

What kind of relationship should humans have with increasingly capable machines?

That is a much larger conversation.

And it cannot belong exclusively to technologists.

It belongs to philosophers, scientists, policymakers, educators, business leaders, religious thinkers, artists, psychologists, and ordinary people.

Because AI is not merely changing what machines can do.

It is forcing humanity to reconsider what humans are for.

Perhaps the greatest contribution ancient philosophy can make to the AI age is not teaching machines how to think.

It is reminding humans to think about what thinking is for.

And that raises an even bigger question:

What could we learn if we placed the world's great philosophical traditions into conversation with each other about the future of AI?

That is where the next conversation begins.

What do you think?

Should the AI conversation be primarily a technological one, or do we need a much broader philosophical conversation about what it means to be human in an age of intelligent machines?

#ArtificialIntelligence #AI #AIethics #HumanAI #Technology #FutureOfWork #Philosophy #AIgovernance

Sanjay Mohindfroo 

AI should multiply workforce capability, not just cut headcount. A board-level framework for automation, cognitive capital, governance, risk, and growth.

In three decades of enterprise technology, I have seen one pattern repeat: when a new technology arrives, management first tries to force it into the economics of the old operating model.

With AI, that usually becomes one question: “How many FTEs can we remove?”

That question is measurable, board-friendly, and dangerously incomplete.

The conventional wisdom says AI transformation should convert automation directly into labor savings. My view is the opposite: if your first AI KPI is headcount reduction, you may destroy the very business capability that makes the technology valuable.

AI Transformation Is Not Workforce Reduction

There is no point pretending AI will not eliminate work. It will.

Some tasks will disappear. Some roles will shrink. Some positions will ultimately become unnecessary. Boards should expect that.

But workforce reduction should be an outcome of transformation, not the definition of transformation.

Consider a team of 100 people. AI removes 30 percent of repetitive work. The traditional response is simple: reduce the team by 30.

The better question is harder: what could the same 100 people achieve with 30 percent more capacity?

Could they serve more customers, reduce risk, accelerate product launches, improve quality, enter new markets, or solve problems the organization has been postponing for years?

That released capacity is not waste. It is an asset.

If management turns every productivity gain immediately into a cost reduction, it captures only one form of value, and often the least strategic one.

The Hidden Asset on the Balance Sheet

Experienced employees carry something most automation business cases fail to price: institutional knowledge.

They know which customer exception matters, which report is unreliable, which supplier requires escalation, which control cannot be bypassed, and why a process that looks inefficient was designed that way in the first place.

AI can process the documented procedure. It does not automatically inherit the undocumented judgment around it.

This matters because a company can save salary costs while simultaneously weakening its operating system.

I call that cognitive capital: the accumulated domain knowledge, judgment, exception-handling ability, historical context, and decision experience that allows an organization to function when the standard process breaks.

Boards track financial capital, physical capital, technology assets and intellectual property. They should start asking whether AI transformation is increasing or decreasing cognitive capital.

The real risk is cognitive debt.

Technical debt appears when shortcuts create future engineering cost. Cognitive debt appears when an organization outsources too much thinking to machines and no longer retains enough human capability to challenge, explain, or recover the process.

You see it when teams cannot operate without AI, experts disappear, exception handling deteriorates, or nobody can reconstruct why a decision was made.

That is not efficiency. It is dependency.

The Four-Layer AI Operating Model

A mature AI strategy needs four systems working together.

First, AI automation: what can machines execute reliably?

Second, human capability: what must people continue to understand and be able to do?

Third, workforce transformation: how do today’s roles evolve into AI-enabled roles?

Fourth, AI governance: what is AI allowed to do, with what data, under what conditions, and with what degree of autonomy?

Most companies over-invest in the first layer because it produces the easiest business case.

That is precisely the mistake.

Higher automation combined with lower human capability and weaker institutional knowledge creates a more fragile enterprise, not a more advanced one.

The Board Must Govern Autonomy, Not Just Adoption

The next phase of enterprise AI is not simply copilots. It is agents that can analyze, decide and execute across workflows.

That changes the governance question.

The issue is no longer only, “Can AI do this?”

It becomes, “Should AI do this, and what happens to human capability if it does?”

Autonomy should rise with evidence, predictability, reversibility and control, not merely because the model is technically capable.

A low-risk reconciliation process may justify bounded autonomy. A regulatory interpretation, strategic pricing decision or crisis response may not.

And “human in the loop” is not a sufficient control if the human merely clicks approve repeatedly.

For important decisions, management needs to define where humans intervene, what exceptions trigger escalation, who remains accountable, and whether the organization can still operate if the AI is unavailable.

One simple test is an AI-off exercise.

Can the team still diagnose the problem, perform the critical process, handle exceptions, and explain the reasoning behind a decision?

If the answer is no, the organization may have automated faster than it has learned.

A Six-Part Board Framework for AI Capability

I would ask boards and CEOs to apply six principles to every major AI transformation.

1. Automate work, not capability

Automate repetitive, low-value work aggressively. But identify the human expertise embedded in the process before it disappears.

The right question is not whether AI can perform the task. It is whether the organization can afford to lose the capability associated with it.

2. Upskill before you replace

Give experienced employees AI tools before concluding that their role is redundant.

The people who know the process are often the best people to teach the organization how to redesign it. They know the exceptions, dependencies, and workarounds that formal documentation misses.

3. Redeploy capacity before reducing capacity

Every successful automation creates an AI dividend.

If a process once consumed one million hours and AI reduces that to 600,000, management has created 400,000 hours of capacity.

Some of that capacity may eventually become cost reduction. But first ask whether it can generate revenue, improve customer experience, accelerate innovation, strengthen controls, or open new markets.

4. Preserve human judgment

Not every decision deserves the same degree of automation.

Routine, reversible work can move toward autonomy faster. High-impact decisions need stronger human judgment, escalation, and accountability.

The objective is not to keep humans clicking approval buttons. It is to keep human judgment where it creates economic and risk value.

5. Increase autonomy with evidence

AI agents should earn greater authority through demonstrated reliability.

Track success rates, escalation rates, intervention rates, policy violations, reversals, and business outcomes.

Build what I would call an agent flight recorder: an auditable history of what the agent was asked to do, what data and tools it used, what actions it took, where humans intervened, and what outcome resulted.

Autonomy without evidence is not innovation. It is unmanaged operational risk.

6. Measure capability, not just cost

If executives are rewarded mainly for FTE reduction, AI will become a headcount program.

The scorecard must be broader: hours saved, cycle-time improvement, quality, exception rates, revenue enabled, new capacity, critical-skill retention, institutional knowledge captured, internal mobility and AI-off resilience.

The objective is not maximum automation.

It is maximum business value per unit of human capability.

The Counter-Argument: What If Cost Reduction Is the Goal?

There are situations where headcount reduction is economically rational.

If work is repetitive, low-risk, well-documented, easily reversible, and carries little strategic knowledge, aggressive automation may be exactly the right decision.

The mistake is not reducing cost.

The mistake is treating all work as if it has the same capability value.

Boards should distinguish between low-value labor that can be removed and high-value expertise that should be amplified.

That distinction is where strategy begins.

The CEO Question Has to Change

The old question is easy:

“How many people can AI replace?”

The better question is more demanding:

“If AI gave us 30 percent more organizational capacity without increasing the workforce, what could we accomplish that we cannot accomplish today?”

That question changes the investment case.

It connects AI to growth, customer acquisition, product expansion, resilience, quality and competitive advantage.

It also changes accountability. Management can no longer declare victory because a cost line fell. It has to show that the enterprise became more capable.

The strongest AI operating model is therefore not:

Automate → Eliminate

It is:

Automate → Augment → Upskill → Redeploy → Transform → Grow

That sequence does not reject efficiency. It captures efficiency without automatically sacrificing capability.

AI will remove work. It should.

But if the people who understand the business become the first casualties of automation, the company may discover too late that it automated away the knowledge required to run the business well.

The board-level measure of AI success should not be, “How many people did we replace?”

It should be, “How much more capable did the organization become?”

Where do you draw the line between legitimate automation-driven cost reduction and dangerous loss of human capability?

Subscribe to TechnologyTrends, or share your view in the comments.

Sanjay Mohindroo

One question reveals whether leaders truly agree on outcomes, value, and accountability before a transformation consumes more capital and credibility.

The One Question That Exposes Misalignment in 60 Seconds

Give a leadership team 60 seconds and one sheet of paper.

Ask each person the same question separately, and you may learn more about the state of a major transformation than you will from a 60-page steering committee deck.

The question is:

“Twelve months from now, what single business outcome will prove this initiative was worth the capital, how will we measure it, and who is accountable for delivering it?”

Over three decades around enterprise technology, I have learned that the most dangerous form of misalignment is rarely open disagreement.

Open disagreement is visible. It can be debated.

The expensive kind is false alignment. Everyone approves the same programme, uses the same vocabulary, attends the same steering committee, and leaves the room believing they agreed on something they actually defined very differently.

That is why this question matters.

It turns alignment from an impression into something you can test.

Most leadership teams confuse consensus with alignment

The conventional wisdom is that large transformation programmes need stakeholder buy-in.

They do.

But buy-in is not alignment.

A board can unanimously approve a technology investment while individual executives expect entirely different returns from it.

The CEO may believe the programme is about improving customer experience.

The CFO may believe it is about reducing operating cost.

The CIO may believe it is about replacing an ageing technology estate.

The business unit leader may expect faster growth.

The risk function may primarily want stronger controls.

Everyone can support the programme.

Everyone can also be pulling it in a different direction.

That distinction matters because large initiatives rarely fail from a complete absence of intelligent people, project plans or governance meetings.

They often fail because important decisions are made against different definitions of success.

When budgets tighten, which capability survives?

When implementation creates disruption, which benefit justifies continuing?

When two business units compete for priority, which objective wins?

When the programme is six months late, what gets protected and what gets cut?

If the leadership team has never agreed on the primary business outcome, those questions get answered tactically.

That is when transformation becomes a collection of compromises rather than an instrument of strategy.

The 60-second business alignment test

The test is deliberately simple.

Before reviewing roadmaps, architecture, vendors, or programme status, ask the leaders responsible for the initiative:

Twelve months from now, what single business outcome will prove this initiative was worth the capital, how will we measure it, and who is accountable for delivering it?

Ideally, ask them individually before discussing the answers as a group.

You are looking for four things.

1. Outcome: Are we solving the same business problem?

The first test is whether people describe the same result.

Not the technology.

Not the project.

Not the activity.

The result.

“Implementing a new platform” is not an outcome.

“Completing the cloud migration” is not an outcome.

“Deploying AI across customer service” is not an outcome.

Those describe work being performed.

A business outcome sounds different:

• Reduce customer onboarding time from ten days to two.
• Increase revenue per sales employee by 15 percent.
• Reduce inventory locked in the supply chain by 20 percent.
• Cut the cost of servicing a customer transaction by 25 percent.
• Reduce the financial exposure created by a critical operational risk.

The distinction looks obvious on paper.

It becomes surprisingly difficult in a boardroom.

If one executive describes the outcome as growth, another as cost reduction, and another as technology modernisation, the initiative is not yet aligned. It may simply have accumulated several rationales to secure approval.

That is an uncomfortable conclusion.

It is also far cheaper to discover before the capital is spent.

2. Measure: Would we recognise success if we saw it?

The second test is measurement.

I have seen many initiatives with extensive programme dashboards and surprisingly weak measures of business value.

Green milestones do not necessarily mean a successful investment.

A programme can be on budget, complete every technical milestone and still fail commercially.

Boards should therefore distinguish between delivery metrics and outcome metrics.

Delivery metrics tell you whether the programme is progressing.

Outcome metrics tell you whether the company is becoming better because of it.

Both matter, but only one answers the investment question.

Consider a company funding a major digital customer programme.

A delivery dashboard might report:

• percentage of functionality completed,
• number of users migrated,
• system availability,
• implementation milestones achieved.

Useful information.

But none of those numbers tells the board whether customers are buying more, staying longer, receiving faster service or costing less to serve.

The board should be able to identify one primary business measure that determines whether the investment created value.

If success cannot be measured, almost any result can later be presented as success.

That is not governance.

It is retrospective storytelling.

3. Time: When exactly should value become visible?

Transformation language often becomes vague around time.

We talk about strategic value, future capability, long-term competitiveness, and foundations for growth.

Some investments genuinely require patience.

But “long term” can also become a convenient hiding place for weak accountability.

A board allocating capital should know when evidence of value is expected to appear.

Not necessarily the full return.

Evidence.

If the programme is expected to take three years, what should be materially different after twelve months?

If nothing measurable is expected for thirty-six months, the board should understand why.

Time creates discipline because it converts ambition into a commitment.

Without a time horizon, programmes can remain strategically important almost indefinitely.

4. Accountability: Which executive owns the outcome?

This is where the question becomes uncomfortable.

Ask who owns programme delivery, and the answer is usually easy.

There is a programme director.

There may be a CIO, transformation office, implementation partner, and steering committee.

Ask who owns the business outcome, and the answer is often less clear.

That is a problem.

Technology can enable a reduction in working capital.

It cannot own working capital.

Technology can enable sales productivity.

It cannot own revenue.

Technology can provide customer data.

It cannot own customer retention.

If a transformation promises a business result, a business executive must ultimately own that result.

This does not reduce the CIO’s accountability. It makes accountability more accurate.

The CIO remains accountable for technology capability, reliability, security, execution and the integrity of the investment.

But if a programme claims it will increase revenue, improve margins or change customer behaviour, the executive responsible for that business outcome must be visibly committed to delivering it.

A steering committee is not an accountable owner.

Neither is “the organisation”.

When everyone owns the outcome, nobody truly does.

What misalignment sounds like in the boardroom

Imagine asking six executives the 60-second question before approving the next phase of a major programme.

You receive these answers:

CEO: “It should materially improve customer retention.”

CFO: “It needs to take at least 10 percent out of our cost base.”

CIO: “We need to retire our legacy environment and reduce operational risk.”

COO: “It should simplify processes across the organisation.”

Business leader: “We need faster product launches.”

Programme sponsor: “We need to deliver the transformation roadmap.”

None of these objectives is irrational.

That is precisely the problem.

They are all plausible enough to coexist without anyone noticing that the company has not made a choice.

A major programme can support several benefits, but it still needs a dominant economic or strategic logic.

Why?

Because eventually those benefits will compete.

A decision that optimises customer experience may increase operating cost.

A decision that accelerates implementation may delay legacy retirement.

A decision that standardises processes may reduce flexibility for a high-growth business unit.

Without an agreed hierarchy of outcomes, every trade-off becomes political.

The programme does not lack governance.

It lacks a governing objective.

A simple board framework: O-M-T-A

I use a very simple way of thinking about the answer.

Call it O-M-T-A:

Outcome. Measure. Time. Accountability.

Before committing significant capital, the board should be able to complete one sentence:

We are investing in this initiative to achieve [OUTCOME], evidenced by [MEASURE], by [TIME], with [EXECUTIVE] accountable for delivering the business result.

If that sentence cannot be completed without twenty minutes of debate, the organisation is not ready to debate technology choices.

That debate comes later.

First agree on what the money is supposed to accomplish.

The board should test five things

Once the sentence is written, ask:

1.   Is there one primary outcome?

Secondary benefits are fine, but the organisation should know which result wins when trade-offs appear.

2.   Is the measure economic or strategically meaningful?

Avoid confusing implementation progress with enterprise value.

3.   Is the time horizon explicit?

Define when evidence of value should become visible.

4.   Does one executive own the result?

Committees can govern. Individuals remain accountable.

5.   Would the same answer survive outside the meeting?

Ask leaders independently. Alignment produced only after group negotiation may be compliance, not conviction.

That fifth test is particularly useful.

Senior leadership teams are very good at creating consensus in meetings.

The stronger test is whether they remain aligned when they are no longer sitting around the same table.

Misalignment is a capital allocation problem

It is tempting to classify this as a communications issue.

I think that seriously understates the risk.

Misalignment is a capital allocation problem.

If a company commits $50 million to a transformation and its executives disagree about what the investment is fundamentally intended to achieve, the organisation has effectively approved multiple competing investment theses under one budget.

That affects far more than project execution.

It changes vendor choices.

It changes sequencing.

It changes organisational design.

It changes which capabilities receive funding.

It changes which compromises are acceptable.

And, eventually, it changes how success or failure is reported to the board.

The cost of misalignment therefore does not appear as a single line item.

It appears as rework, delayed benefits, scope expansion, political escalation, underused capabilities and investments that technically finish but never produce the return originally expected.

“But large transformations have multiple objectives”

This is the most reasonable objection to the one-question test.

Of course they do.

A major ERP transformation, for example, may improve control, lower cost, standardise processes, reduce technology risk and create better management information.

The answer is not to pretend those secondary outcomes do not exist.

The answer is to establish hierarchy.

Every serious strategic investment needs a primary reason for existing.

If the company had only 60 percent of the available capital, which benefit would it protect?

If the programme had to sacrifice one objective to secure another, which one wins?

If the board had to judge the investment five years later using only one measure, what would it choose?

Those questions expose priority.

And priority is what makes strategy executable.

A strategy that treats every objective as equally important has avoided the hardest part of strategy: choosing.

Use the question before the programme is in trouble

Most organisations perform alignment exercises after warning signs appear.

Budgets increase.

Timelines move.

Benefits become uncertain.

Business sponsors disengage.

Then everyone asks whether the programme is aligned with strategy.

That is too late.

The 60-second question belongs much earlier.

Use it when approving the business case.

Use it before selecting major partners.

Use it at the start of each major investment phase.

Use it when leadership changes.

Use it when a programme requests substantial additional funding.

Most importantly, use it before discussing the technology itself.

Boards do not need to become technology committees.

They need to become much harder to satisfy on the connection between technology and enterprise value.

The real test of alignment

A perfectly aligned leadership team does not need identical language.

It needs a common economic logic.

Ask five leaders why the organisation is spending the money.

If one talks about growth, another about efficiency, another about risk, another about modernisation and another about completing the programme, do not congratulate yourself on having a broad transformation agenda.

You have probably discovered five different investment theses.

Resolve that before approving the next tranche of capital.

The best governance question is often not the most sophisticated one.

Sometimes it is simply the question nobody has forced the room to answer precisely.

Twelve months from now, what single business outcome will prove this initiative was worth the capital, how will we measure it, and who is accountable for delivering it?

Ask it separately.

Give people 60 seconds.

Then compare the answers.

You may discover that the transformation problem you thought you had is actually an alignment problem.

And discovering that early can save far more than another round of programme optimisation ever will.

What is the one question you use to determine whether a leadership team is genuinely aligned, rather than simply in agreement?

Subscribe to TechnologyTrends, or add your perspective in the comments, especially if your experience leads you to a different conclusion.