A personal story · YC Startup School 2026

YC flew me to San Francisco to sit in a room with Jensen Huang and Sam Altman.

From a kid in a small town in Maharashtra dreaming of Silicon Valley, to a seat a few rows from the people shaping the future of AI. The best two days of my life.

Scroll into the painting
Jaideep sitting in front of the giant red Y sculpture outside Chase Center
I.

The pilgrimage

I grew up in a town in Maharashtra, India, with about 3,000 people. Today I study at Duke, in North Carolina. San Francisco was always the dream.

In July, Y Combinator paid for my round-trip flights from North Carolina to San Francisco.

N Homea town of ~3,000, Maharashtra DukeDurham, North Carolina San FranciscoChase Center flight paid by YC
II.
0

petitions. One seat for me.

Builders from over 120 countries applied to Startup School 2026. Keep scrolling: the sky empties until one light is left. That one is mine.

III.

A constellation of the people running the AI race

The Steve Jobs of my generation, on one stage. Chips, models, the tools people build with, and AI that has to survive the real world.

Jensen HuangSam AltmanJeff Dean Alexandr WangBoris ChernyPeter Steinberger Patrick CollisonDmitri DolgovChelsea Finn NVIDIAOPENAIGOOGLEMETA CLAUDE CODEOPENCLAWSTRIPEWAYMOPHYSICAL INTELLIGENCE
From my seat

The people I grew up reading about, a few rows away

Every photo below is from my phone, taken from where I sat at Chase Center.

Jensen Huang in his leather jacket in the armchair on the Center Court stage
Jensen Huang, Center Court
Waymo talk on the big screen: The Waymo Foundation Model
Waymo's "7 Lessons" talk
Sam Altman's closing fireside chat, the crowd behind filming on phones
Sam Altman's closing session, Day 2
Big screen with the caption Master speed above the Startup School 2026 stage and the arena
"Master speed." The whole arena
A speaker in a blue hoodie laughing in the armchair
Center Court, Day 2
Jumbotron caption asking for a message in a bottle to an 18 year old self, the full arena below
"A message in a bottle to the 18 year old version of yourself"
IV.

The sermons, and what I kept

The single greatest time to start a company.
Jensen Huang · NVIDIA · Day 1

NVIDIA's founding bet was wrong. Facing that fast is what saved it. Agents need control: change one word in a plan, get one word of change back.

Your model is table stakes, eval is your strategic moat.
Dmitri Dolgov · Waymo · Day 1

In physical AI the demo is 1% of the work. The rest is a game of nines.

I can go start the world's most ambitious crazy company.
Sam Altman · OpenAI · Day 2 closing

Be right while everyone thinks you are wrong, and accept that it takes a long time.

Hype is weather. Adoption is climate.
Peter Steinberger · OpenClaw · Day 2

A weekend project that became the fastest-growing repo on GitHub in 60 days.

If you have to fine-tune, you don't actually get a single general purpose model!
Chelsea Finn · Physical Intelligence · Day 2

Robots keep the last seconds as video and compress everything older into text.

That's a hell of a lot slower than knowing it in cognitive L1 cache
Patrick Collison · Stripe · Day 2

Learn things yourself, even when an agent can look them up.

Delete the system prompt. Add back only what breaks.
Boris Cherny · Claude Code · Day 1 (paraphrase)

With the newest model, his team cut more than 80% of it.

Ignore LinkedIn.
Alexandr Wang · Meta · Day 2

When intelligence gets cheap, vision and ambition are what's scarce.

The most important thing is to pick something you're super excited about.
Jeff Dean · Google · Day 1, with Diana Hu

Pick problems today's models fail at almost completely, around 1%, not 20%. Those are still open. And moving data costs about a thousand times more than computing on it, which is why AI hardware looks the way it does.

Jaideep courtside with a thumbs up, the stage and jumbotron behind
V. · Courtside · keep scrolling
The closing fireside · on the jumbotron
"We were like hiding out in this little building in Cambridge. PG was making us dinner."

From my seat, at the rail

YC's origin story, told a few rows from me. Twelve seconds I filmed during the closing session.

VI.

The offerings: about $35K in credits and tools

$0Compute creditsOpenAI, Anthropic, Cursor, xAI, AWS, Azure
Claude Max20x plan, Claude Code includedFrom Anthropic, to every attendee
$0Meta API creditsFrom Alexandr Wang, to the whole room

All of it, by my count, about $35K.

VII.

The revels

More parties than nights, and a room full of the most cracked people of my generation, from Taiwan to MIT, Stanford to China to France. Somewhere in there is the next Jensen Huang.

A packed founder meetup at an arcade bar
AWS photo booth prints on a purple table
Group selfie at a house party, badges on
Group selfie with Jaideep in a Spartan helmet
Crowd on the lawn outside Chase Center under the orange Y
Coda

Best two days of my life. I still can't fully believe they happened.

A kid from a town of 3,000 sat a few rows from the people building the future, on flights YC paid for.

Jaideep in front of a wall of framed photos from YC's history
Later, a second YC badge: the Startup Internship Expo at YC's office.