HARIO SETO S

FROM LOSS TO DRIVE

BORN FROM INSTABILITY

Dara’s life was shaped by political upheaval.

His family lost safety, status, and everything they had built in Iran.

That experience did not just create fear.
It created a permanent drive to rebuild and never take stability for granted.

WHEN THE FLOOR CAN DISAPPEAR

He described a feeling that never leaves:

The rug can be pulled from under you.

That mindset can create anxiety.
But in business it also creates vigilance, ambition, and a refusal to become complacent when things look safe on the surface.

HIS FATHER’S IMPACT

Watching his father lose not just wealth but also his sense of value left a deep mark.

The lesson is massive:
Work is not only about income.
It is tied to dignity, worth, contribution, and identity.
That matters in every discussion about AI and unemployment.

WANTED TO MAKE FAMILY PROUD

AMBITION BEGAN AS DUTY

He did not begin with a master plan to become CEO.

His early motivation was simpler:
Make his family proud.

That matters because many careers are built first on duty, responsibility, and hunger long before they are shaped by a polished career narrative.

IMPACT MATTERS MORE OVER TIME

After losing everything, making money mattered.
That was real.

But as safety improved, what mattered more was building something important.

This is a useful founder lesson:
Money can motivate the start.
Impact often sustains the long game.

LEARN FROM THE FATHER YOU SAW

One memory stayed with him:
Visiting his father’s factory and seeing workers respect him.

Not fear.
Respect.

That shaped a core idea:
Real leadership is not just scale or authority.
It is building something big while treating people with dignity.

ENGINEERING AS A CEO MINDSET

WHY ENGINEERING MATTERS

He loved engineering because equations map to reality.

That same logic shaped his view of business:
A company is an organism and a machine.

The CEO’s job is to engineer goals, structures, incentives, teams, and systems that make the machine produce the intended result.

PICK THE RIGHT GOALS

Problem solving is not enough.

You also have to choose the right problem.

That is a strong operator lesson:
A well-run team can still lose if it is optimizing the wrong goal.
Execution excellence cannot save a company from strategic misalignment.

GREAT CEOS THINK LIKE BUILDERS

His framing is sharp:
The CEO is not just a spokesperson or capital allocator.
The CEO is designing a living system.

That means structure, pacing, accountability, information flow, and talent design all become part of the product of leadership itself.

BET ON PEOPLE

GREAT COMPANIES START WITH GREAT PEOPLE

One lesson he learned early:
Always bet on people.

Companies rise and fall.
Markets change.
But great people remain unusually valuable over long periods.

This matters for founders, investors, and hiring managers making long-term decisions.

WHAT MAKES SOMEONE WORTH BETTING ON

He pointed to traits like success, honor, loyalty, and follow-through.

Not charisma alone.
Not hype alone.

A great person says what they will do and then does it.
That consistency compounds trust over time and becomes a strategic asset in business.

RELATIONSHIPS OVER TRANSACTIONS

The idea is not just to find talent once.
It is to stay with good people through their careers.

That is how strong networks and enduring businesses are built.
You do not just collect deals.
You build long arcs of trust with exceptional people.

SPOTTING OPPORTUNITY IN TRANSITIONS

FIND THE LEADERS IN THE SHIFT

When big transitions happen, exact outcomes are unclear.

So instead of predicting every detail, identify who is already emerging as the leader in the shift.

That is how he and his team approached online commerce categories like travel, ticketing, and personals.

OVERPAYING FOR GREATNESS

He said they often overpaid for great companies based on what the market thought at the time.

But those prices looked cheap later because the future was not linear.

Strong transitions often create exponential outcomes that ordinary valuation logic fails to fully price.

HUMANS THINK LINEARLY

People project the future in straight lines because everyday life feels linear.

But new technologies can create hockey-stick outcomes.

The opportunity often lives in the gap between what people assume will happen and how fast a superior technology actually scales.

TURNAROUNDS REQUIRE FORCE

A BROKEN ENGINE IS AN EMERGENCY

At Expedia he saw something dangerous:
The technology engine was broken.

Old codebase.
Weak reinvestment.
Coasting leadership.

His lesson was decisive:
When you see the bell ring, act.
Do not wait for more proof while momentum turns against you.

TECHNOLOGY DECAY BECOMES EXPONENTIAL

He explained that just as growth can become exponential, decline can too.

A bad technology trajectory may not look catastrophic in year one.
But if left untouched it can turn into a long-term disaster.
Leaders must act before the full damage becomes visible.

SOMETIMES YOU REPLACE THE TEAM

He was blunt:
Sometimes the shortcut to changing culture is changing people.

Values posters are not enough.
If the company is coasting, you may need hungry people with the right operating instincts before culture can actually reset in real life.

TRANSPARENCY AS SELF-DEFENSE

TELL THE TRUTH FIRST

His view is powerful:
As a leader, if you hide reality from your team, they will hide reality from you.

Transparency is not just ethics.
It is information strategy.
You tell the truth because it is the only way to increase the odds of receiving truth in return.

BAD DECISIONS OFTEN START WITH BAD DATA

He said many CEO failures are not from low intelligence.
They come from getting the wrong information.

This is crucial:
A leader must design channels that surface uncomfortable truth quickly.
Otherwise polished summaries slowly separate leadership from reality.

GO TO THE SOURCE

He learned to bypass filtered reporting and hear directly from the source.

The farther information travels up a hierarchy, the more its sharp edges get rounded off.
That is why direct channels with people deep in the system can be so valuable for CEOs.

CULTURE OF HARD WORK

HARD WORK IS A SKILL

One of his strongest ideas:
Working hard is a skill.

It includes discipline, focus, repetition, emotional endurance, and the ability to keep going after losses.

He sees it as one of the most important advantages a person can build over time.

RELENTLESSNESS COMPOUNDS

His logic is simple:
If you can take two shots while others take one, you compress time.

You get more data.
You learn faster.
You increase your odds of success.

Relentless execution is not just effort.
It is a compounding system for learning and winning.

BE HONEST ABOUT THE STANDARD

He believes a company should be clear:
If you come here, you will work hard.
You will be stretched.
You will be held accountable.

That honesty lets the right people opt in.
It also reduces the damage caused by hidden expectations and cultural mismatch.

FLEXIBILITY IS NOT LAZINESS

He makes an important distinction:
You can work very hard and still have flexibility.

Dinner with family.
Emails at night.
Early morning check-ins.

The issue is not rigid office theater.
The issue is whether the person and the company are truly committed to output.

RISK, LOSS, AND LEARNING

SAY THE LOSS OUT LOUD

He admired leaders who can say:
They won.
We lost.
Next.

That matters.
Do not deny the loss.
Do not drown in it.
Study it enough to learn.
Then move.
A company or person that cannot metabolize failure gets stuck in fear and self-protection.

TOO DEFENSIVE IS DANGEROUS

As companies become successful they often become risk-averse.
They have more to protect.

He argues the opposite should happen.
A stronger company should be able to take smarter risks because it has more resources to absorb mistakes while pushing for bigger upside.

SET THE EXAMPLE ON RISK

He pushes teams to take smart risks through language, example, and personal decisions.

People do not learn courage from slogans.
They learn it when leaders take visible risks, survive misses, and show that intelligent experimentation is expected rather than punished.

VALUES THAT ACTUALLY MEAN SOMETHING

GENERIC VALUES ARE FORGETTABLE

He criticized generic value lists that sound nice but describe every company on earth.

If your values could fit any brand, they do not really guide behavior.
Strong values should reflect how your company is different and how people are expected to act under pressure.

DO THE RIGHT THING. PERIOD.

His favorite value was simple:
Do the right thing. Period.

No long paragraph.
No corporate poetry.

The power is in the burden it places on the employee:
Use judgment.
That pushes responsibility downward instead of hiding behind process or technical excuse.

GO GET IT

One value he highlighted was 'go get it.'

It fits Uber’s product and its operating posture.
It signals movement, aggression, initiative, and winning intent.

Good values are memorable because they capture actual behavior in language the company can live with daily.

AI IS ALREADY INSIDE UBER

UBER WAS BUILT ON APPLIED AI

He explained that Uber already runs on AI:
pricing, routing, matching, batching, and orchestration across tens of millions of daily trips.

This matters because it shows AI transformation often begins long before the public narrative catches up to it.

COMFORT WITH IMPERFECTION

Uber learned to live with AI systems that work most of the time but still create edge-case failures.

That is a mature lesson:
AI adoption is not about waiting for perfection.
It is about building organizations that can operate with probabilistic systems responsibly.

APPLIED AI WILL RESHAPE EVERY TEAM

He does not frame Uber as a frontier research lab.
He frames it as a company moving hard on applied AI.

That distinction matters for builders:
You do not need to invent foundation models to gain advantage.
You need to apply models deeply inside real workflows.

THE BIG JOB QUESTION

AI CAN REPLACE MUCH HUMAN WORK

He said the capability may exist within about 10 years for many intellectual jobs and longer for physical jobs.

He did not give false comfort.
The point was clear:
The scale of disruption may be enormous and society may not be ready for the retraining challenge.

THE SPEED IS THE PROBLEM

Society has adjusted to past technological shifts.

But his concern is timing.
If the capability arrives too fast, the social systems around retraining, identity, income, and meaning may not adapt quickly enough.
That is where the real shock could happen.

JOBS ARE ALSO ABOUT WORTH

He connected work to value and self-worth.

That is why job loss is not only an economic issue.
It is psychological, social, and existential.

Any serious AI strategy that ignores human meaning will misread the scale of the coming societal challenge.

CODING IS CHANGING FAST

MOST ENGINEERS ALREADY USE AI

He shared that most of Uber’s coders already use AI tools.

But the biggest gains came from a smaller group of power users.

That is a useful signal:
Adoption alone is not the advantage.
Depth of usage and workflow redesign are where real productivity jumps begin.

FROM WRITING CODE TO ORCHESTRATING AGENTS

He sees coding moving from manual construction toward orchestration.

The engineer still matters.
But the job increasingly becomes directing systems, reviewing outputs, shaping architecture, and managing agent-based production rather than typing every piece by hand.

PRODUCTIVITY GAINS CHANGE HIRING LOGIC

His view is pragmatic:
If engineers become more productive, maybe you hire more of them to move faster.
Later, maybe you add fewer humans and more agents and GPUs.

That is how AI shifts strategy:
Not only what teams do but what management decides to buy.

AUTONOMY AND THE FUTURE OF UBER

AUTONOMOUS VEHICLES WILL LIKELY WIN

He was direct:
Autonomous systems are becoming safer than human drivers in key contexts.

That creates a massive social upside through fewer deaths and cheaper transport.
But it also means large numbers of driving-related jobs may eventually disappear.

NOT TOMORROW BUT NOT FAR AWAY

He does not expect full replacement overnight.
Regulation, hardware, manufacturing, sensors, and real-world deployment take time.

Still, the direction is clear:
Over 15 to 20 years, a growing share of trips could be fulfilled by non-human systems.

WHAT DO THE DRIVERS DO?

When asked what millions of displaced drivers will do, he did not pretend to have a clean answer.

That honesty matters.
He pointed to new forms of platform work and AI-related tasks.
But the balance between new opportunity and automation remains uncertain.

ADVICE FOR PEOPLE IN THE AI ERA

WORK HARD FIRST

His default advice to young people stayed the same:
Work hard.

Not because it solves everything.
But because it remains one of the few durable edges that compounds across industries, roles, and changing technology cycles.
A strong work ethic travels well across uncertainty.

DO NOT OVER-PLAN YOUR CAREER

He warned against overly rigid career plans.

Why?
Because they reduce curiosity.
People start filtering reality for evidence that supports their existing plan instead of staying open to life-changing signals coming from the real world around them.

LET THE WORLD CHANGE YOU FIRST

Before trying to change the world, let the world change you.

That is one of the strongest lines in the interview.
It argues for curiosity over ego, signal over rigid identity, and responsiveness over self-constructed certainty in rapidly shifting environments.

FINAL LESSONS

WHAT FOUNDERS CAN LEARN

Tell the truth early.
Act fast when the engine is broken.
Bet on exceptional people.
Build direct information channels.
Treat hard work as culture.
Use values that actually guide behavior.
Take smart risks.
Operate with urgency before the market forces urgency on you.

WHAT PROFESSIONALS CAN LEARN

Do not depend only on job titles for security.
Build resilience.
Learn to work hard.
Stay useful.
Stay curious.
Adapt faster.
Treat AI as a force you must understand, not a topic you postpone.
And remember that meaning and contribution matter as much as income.

THE DEEPER TENSION

The interview sits inside one central tension:
Technology can improve life and still destabilize millions.

That means serious leaders must do two things at once:
Build the future aggressively and speak honestly about the cost of getting there.

START SMALL INSIDE A BIG MARKET

Start with a large market like finance, healthcare, or marketing. Then narrow down into a specific sub-niche where competition is smaller and users have clearer pain points.

MAP THE FULL WORKFLOW

UNDERSTAND THE WORK END TO END

Study how the niche works from start to finish. Document every step people go through to complete their work.

IDENTIFY WHERE MONEY CHANGES HANDS

FIND THE REVENUE MOMENTS

Look for the exact moments where payment happens. These are usually the highest-value places for software to insert itself.

SPOT REPETITIVE MECHANICAL TASKS

REPETITION IS AUTOMATION OPPORTUNITY

Find tasks that happen repeatedly every day. Repetitive actions are the easiest targets for automation.

QUANTIFY THE COST OF THE WORKFLOW

PUT A NUMBER ON THE PAIN

Calculate how much time or money the workflow consumes. Once you quantify the cost, the value of your SaaS becomes much easier to sell.

CREATE SCROLL-STOPPING CONTENT

BUILD ATTENTION BEFORE THE PRODUCT WINS

Build media around the niche. Share insights, tips, and workflow pain points to attract an audience that already cares about the problem.

STUDY CONTENT ENGAGEMENT

ENGAGEMENT REVEALS DEMAND

Analyze posts that get saves, replies, and direct messages. These signals show where real interest exists.

DOUBLE DOWN ON ORGANIC WINNERS

REPEAT WHAT ALREADY WORKS

When a piece of content performs well organically, create more content around that same angle, pain point, or format.

RUN PAID ADS ON PROVEN CONTENT

PROMOTE WHAT HAS ALREADY PROVEN ITSELF

Turn successful organic posts into ads. If they work organically in the niche, there is a strong chance they can work as paid campaigns too.

CAPTURE EMAILS FROM DAY ONE

OWN THE AUDIENCE RELATIONSHIP

Build an email list early. Social platforms can change anytime, but email gives you a direct channel to your audience.

PERFORM THE WORKFLOW MANUALLY

DO THE WORK BEFORE YOU AUTOMATE IT

Start by doing the service manually. This gives you a practical understanding of the actual problem and outcome.

DOCUMENT EVERY STEP PRECISELY

PRECISION CREATES SYSTEMS

Write down each part of the process in detail so it can later be standardized, delegated, or automated.

SEPARATE JUDGMENT FROM MECHANICAL TASKS

KNOW WHAT HUMANS SHOULD KEEP

Figure out which tasks require human judgment and which are purely repeatable mechanical actions.

CONVERT MECHANICAL TASKS INTO AGENT WORKFLOWS

TURN REPETITION INTO AUTOMATION

Mechanical tasks are ideal candidates for AI agents and structured workflow automation.

DESIGN AI AGENTS TO COMPLETE TASKS

BUILD AGENTS AROUND SPECIFIC OUTCOMES

Each AI agent should be responsible for clear tasks inside the workflow.

CONNECT AGENTS TO REAL TOOLS

AGENTS NEED ACCESS TO SYSTEMS

Agents become useful when they connect to email, CRM, payments, Slack, and other real tools.

ADD ORCHESTRATION AND VERIFICATION

COORDINATION IS THE NEW INTERFACE

Future software will coordinate multiple agents, manage retries, and verify outputs before delivering results.

STORE USER PREFERENCES AND MEMORY

MEMORY BECOMES YOUR MOAT

Track user behavior and preferences so the product becomes smarter over time.

LAUNCH NARROW WITH HIGH-TOUCH ONBOARDING

GO DEEP BEFORE YOU GO BROAD

Serve a small group extremely well first and learn quickly.

PUBLISH MEASURABLE PROOF

PROOF SELLS BETTER THAN CLAIMS

Show measurable results such as hours saved or revenue increased.

SHIFT FROM PER-SEAT PRICING

TRADITIONAL SAAS PRICING IS WEAKENING

Charging per user seat is becoming less attractive in the AI era.

ADOPT PER-TASK PRICING

CHARGE FOR WORK DONE

Customers increasingly prefer paying for tasks completed.

MOVE TOWARD OUTCOME PRICING

PRICE BASED ON VALUE DELIVERED

The best pricing aligns with results delivered to the customer.

EXPAND INTO ADJACENT WORKFLOWS

GROW SIDEWAYS IN THE WORKFLOW

After solving one workflow, expand into nearby ones.

ORCHESTRATE MULTIPLE AGENTS

BUILD AN AGENT ECOSYSTEM

Multiple agents should collaborate across the workflow lifecycle.

BUILD SWITCHING COSTS

DATA AND MEMORY CREATE LOCK-IN

The more data your product stores, the harder it is to replace.

TURN POWER USERS INTO CASE STUDIES

SUCCESS STORIES DRIVE GROWTH

Real user stories build trust and attract more customers.

HIRE OPERATORS FROM THE NICHE

DOMAIN EXPERTS ACCELERATE PROGRESS

Industry insiders understand problems better than outsiders.

REINVEST PROFITS INTO GROWTH

REINVEST IN PRODUCT AND DISTRIBUTION

Use profits to improve features and expand reach.

BECOME THE DEFAULT EXECUTION LAYER

OWN THE WORKFLOW OF THE NICHE

The goal is to become the software layer that runs the core workflow for that niche.

THE CORE SHIFT

AI MOVES INTO EXECUTION

AI will increasingly run the job end to end: research, draft, build, deploy, report, revise. It doesn’t get tired, it scales instantly, and it improves via iteration.

When execution is a repeatable chain, AI becomes the default worker that keeps shipping outcomes, not just suggestions.

HUMANS MOVE INTO DIRECTION

Humans become the mastermind that gives direction. Not “more ideas,” but better decisions: what matters, what to ignore or prioritize, what risks.

When outputs are infinite, the scarce skill is choosing the right target, defining “good,” & steering the system as reality changes.

WHY WORK BECOMES COMPOSABLE

JOBS ARE CHAINS

Most jobs are chains, not a single skill: research > plan > draft > build > distribute > measure > iterate.

AI is strongest on chained work because it can execute each step fast and consistently, then loop back with new inputs. The more modular the work, the easier it is to automate.

EXECUTION BECOMES CHEAP

Writing, coding, design, ad iterations, replies, reporting, QA, summarization.

The marginal cost of “one more version” collapses.

Output becomes abundant.

The bottleneck shifts away from production speed to selection quality: who knows what to make, what to ship, and what to stop.

AI EDGE : ORCHESTRATION

DEFAULT INTERFACE: ORCHESTRATION

The winners won’t be single tools.
They’ll be doers that coordinate tools, data, and actions. AI agents will plan tasks, call APIs, operate apps, trigger workflows, recover from errors, & continue.

The interface: “Goal + constraints” > autonomous execution > reporting > iteration.

AI LAYER: OPERATOR

Runs tasks end to end. Produces outputs in volume. Monitors metrics. Executes revisions. Maintains consistency.

It’s built for repeatability. Operator AI is not just generating text. It is managing steps, state, tools, and retries until the deliverable matches the brief and constraints.

HUMAN EDGE : MASTERMIND

SPOT THE VALUABLE PROBLEM

While many fail when they solve a visible problem, not a valuable one.

Human value rises in identifying leverage, where effort creates disproportionate returns. In an abundance world, production is not the advantage.
Choosing the right problem, market, pain, timing > the real competitive edge.

ASK THE RIGHT QUESTION

AI answers questions.
Humans decide which question worth asking. The question defines the search space, constraints, what “good” means.

Bad questions produce busywork at scale.
Good questions produce clarity: what matters, what proof, what trade-offs, & what outcome.

CHOOSE THE RIGHT HYPOTHESIS

A hypothesis is : “If we do X for Y, we expect Z because…”

AI generate many of these.
Humans choose the one aligned with reality, constraints, risk tolerance, and brand.

The mastermind picks the best bet, defines the test, sets the success metric, & decides whether to double down or pivot.

WHAT STAYS HUMAN

GOAL SELECTION

AI optimizes inside a goal.
Humans choose which goal matters.

This is strategy:
what game you are playing, what you will not do, and what “winning” looks like.

Human sets direction :
Goals, priorities, constraints, success metrics, and timeline.

TASTE AND JUDGMENT

When execution is abundant, taste becomes the filter.

The mastermind decides what is right for this audience, this moment, & this positioning. It includes: tone, trust, cultural signals, long-term brand memory, & the difference between “works now” & “works sus

TRADE-OFFS UNDER UNCERTAINTY

Real work is constraint navigation: time, budget, brand risk, legal risk, team capacity, politics, and timing.

AI can propose options.
but dont own consequences.

The mastermind chooses the trade-off, accepts risk, & makes the call with incomplete information.

TRUST AND ACCOUNTABILITY

People don’t blame the tool. They blame the owner.

Humans approve, validate, sign & carry responsibility.
That’s why direction stays human: ethics, safety, legality & reputational impact.

Human defines guardrails, verification steps & escalation rules before delegating execution to AI.

ETHICS

AI can execute anything asked to do.
Humans decide what should never be done.

Ethics is not optimization.
It is boundary-setting: what is acceptable, what is harmful, what crosses trust, law, or long-term damage.

The mastermind defines moral limits before speed and scale amplify mistakes.

MULTI-LAYER KNOWLEDGE

Real direction requires combining multiple layers: domain expertise, context, human behavior, timing, culture, constraints, etc.

AI knows fragments.
Humans integrate them into one coherent outcome.

Human connects signals across those layers, decides how shape a single, intentional outcome.

THE WINNING LOOP

THE LOOP

Spot leverage (problem/opportunity)
Ask the right question
Choose the hypothesis
Set goals and constraints
Ask AI to executes
Ask AI reports
Ask AI iterates
Updates direction.

The loop speed is the advantage.

Who runs this loop best will out-ship larger teams.

WHAT'S NEXT

HOW LONG THIS WILL LAST

At least the next 5 years.

Not because AI will stop improving.
But because direction compounds slower than execution.

Execution scales instantly.
Direction requires experience, context, judgment, and accountability.
Those do not compress easily.

WHY EXECUTION MOVES FAST

AI improves on speed, cost & autonomy every cycle.

Execution becomes faster, cheaper, and more automated.
Orchestration becomes normal.
Output becomes infinite.

The ability to “do” stops being rare.
It becomes baseline.

WHY DIRECTION MOVES SLOW

Direction requires experience, context, and consequence.
It depends on taste, ethics, responsibility, and long-term thinking.
These do not compress into compute.

This asymmetry keeps the mastermind role human.

THE REAL BOTTLENECK

For the next 5 years, the hardest work stays upstream :
Choosing the right problem.
Asking the right question.
Making the right hypothesis.
Defining what “good” actually means.

AI can accelerate answers.
It cannot decide meaning.

THE STRUCTURAL SHIFT

This is not a temporary advantage.
It is a structural reallocation of value.

Doers are automated.
Operators are AI.
Direction belongs to humans.

WHO WINS

Those who train execution will be replaced.
Those who train direction will compound and win

The next 5 years reward:
THE MASTERMIND, not makers.
Decision-makers, not producers
System runners, not task doers.

FINAL WORDS

This door is open now.
It won't last forever.

But for at least the next 5 years,
human leverage lives in direction.

Train yourself, your kids that skill.