The Young Man Who Got Half of Silicon Valley’s Poker Table to Pay His Tab | 【Jingwei Low‑Key Share】

A Young Man Who Got Half of Silicon Valley’s Table to Pay for Him | 【Jingwei Low-Key Share】

He doesn’t have a conventionally impressive résumé. No college degree, and no stint at a big tech firm. But he has a clear habit: whenever he encounters a problem, he first builds something himself. When BitClout couldn’t cash out, he wrote BitSwap; after wrestling with visa issues, he simply created Extraordinary; noticing cloud AI increasingly devouring privacy, he turned back to tinker with local intelligence.

During those seemingly “chaotic” years, he also happened to meet a group of people who later became increasingly valuable and eventually turned into his investors. Naval, Patrick Collison, Noam Brown, Midjourney founder David Holz… Many of them didn’t first encounter him at a fundraising roadshow; they had already seen him write code, build products, mess things up, and then get back up and start over.

By 2026, a16z led the round for Conway Research, with Khosla, Anthology Fund and a long list of Silicon Valley names joining in. It may look like a sudden takeoff, but the on‑site due diligence had been going on for many years. The article below will take us into this young company. Enjoy:

Source | Beating

Article | Sleepy

On October 2, 2026, a16z announced that it led the first round of financing for Conway Research.

The company’s core product is called Underdog; the exact financing amount was not disclosed, but the publicly revealed investor list is exceptionally prestigious.

The list of institutions includes Khosla Ventures, Hummingbird, and the Anthology Fund jointly established by Anthropic and Menlo Ventures; individual investors also include Naval Ravikant, Patrick Collison, Guillermo Rauch, Noam Brown, Thomas Wolf, Logan Kilpatrick, as well as Marco Mascorro and roon.

It can be said that most of the players at the Silicon Valley table have invested in this startup.

The founder of Conway Research is Sigil Wen.

He doesn’t have a university degree, has never worked at a big tech firm, and hasn’t published any academic papers. A young person who has almost never been stamped by the standard system in this cycle—what exactly does he rely on to make the most discerning people in Silicon Valley willing to place their chips in front of him?

01

Negative Numbers

On January 18, 2021, Sigil recorded an embarrassing incident in his annual goals list.

His bank account hit rock bottom, the balance turned negative. This high‑school student still studying in Toronto had to lower his head and ask his mother for money, pleading with her to pay his Adobe Creative Cloud subscription for the next month.

On the same Notion page, he wrote his ambitions for the year: delivering a product, securing venture capital, and eliminating financial anxiety.

It’s a distinctly youthful sense of fragmentation.

Dramatically, just a few months later, HubSpot founder and renowned angel investor Dharmesh Shah publicly announced on social media that he was willing to invest $100,000 to purchase a 1% stake in a project called BitSwap that Sigil is working on.

Based on this quote, the project casually written by this high school student is already valued at $10 million.

From borrowing money from his mother to pay software fees to having someone assign a ten‑million‑dollar valuation to his project, the gap is as short as Silicon Valley itself.

In the following years, Sigil’s life trajectory was filled with these extremely rapid shifts; a fragmentary job, a friend just followed on Twitter, a technology just open‑sourced on GitHub could toss him into a completely new game within two weeks.

This is not a story of a prodigious youngster having an epiphany.

If you trace the timeline backward, you’ll find that his starting point was so mediocre it was even a bit awkward.

Sigil was born into a Chinese immigrant family in Toronto.

In fifth grade, he was an extremely introverted child who hid in a corner drawing comics; in eighth grade, he became obsessed with Minecraft in the public library and applied for the TOPS program at Marc Garneau High School.

When he started his first year of high school, he declared himself in his notes as “the worst speaker in the entire grade.”

This has nothing to do with the smooth‑talking image he later projected on various podcasts and Twitter Spaces. Digging up the earliest videos of him posted online is so awkward it makes you squirm; he wears flamboyant plastic fake glasses, a steel brace around his teeth, his gaze fixed rigidly on a handwritten cue card, and if he mispronounces a word he freezes in front of the camera.

If it were most of the later Silicon Valley newcomers, this digital trace would almost certainly have been cleaned up spotless, but he refuses to delete it.

In his memo back then he wrote that keeping these was simply to prove to everyone that, even if they later become a skill to trade for a meal, the start can be utterly disastrous.

Math was the only thing he could showcase at the time, and his name appeared on the University of Waterloo’s 2018 competition honors list. But in Toronto, any high‑school student who can solve problems is already on the scene.

He quickly turned to more concrete technical work. After completing Andrew Ng’s machine learning course on Coursera, he made a video deriving neural networks; when YOLO object detection was just open‑sourced on GitHub, he promptly pulled the code, reproduced it himself, worked through the pitfalls, and then recorded a commentary series to post online.

From understanding a lecture, to getting a piece of code to run, to explaining it in person, this loop he had mastered before even stepping out of high school.

The later baffling cross-disciplinary jumps he made were all rooted in this unconventional approach.

02

Take Jobs

In the fall of 2019, Sigil, a sophomore, had just joined The Knowledge Society.

At an Elevate hackathon in Toronto, he and his teammates stayed up for 36 hours straight to deliver an AR maintenance project, earning a ticket to a tech conference. He got home on Sunday night, completely exhausted.

On Monday morning, thinking that the venue was more than an hour away by tram, he turned off the alarm and obediently slung his backpack and went to school.

He later confessed in a Medium blog that he regrets it terribly.

It was an extremely petty swing. Later biographies love to give every outlier a precocious look of “seeing through the system on day one,” but at the time there was no earth‑shattering life‑changing decision at all.

What lies before him is nothing more than an extremely practical multiple‑choice question. Should he honestly attend class today, or skip a day and go see the world outside?

But soon, even that wavering leisure was gone.

Because he ran out of money.

He started an agency called Sigil Digital, selling his limited skills—such as video editing, web design, Shopify site building, and social media layout—one by one to any client willing to pay.

He edited short videos for Pioneer.app, did outsourced marketing for Blockchain Founders Fund, built websites for some crypto startups that have long since vanished, and created Shopify pages for a smartwatch brand.

These fragmented projects would never make it into any top‑tier financing pitch deck, yet they were his earliest real cash flow.

The process of reaching out to Twitch co‑founder Justin Kan follows the same logic. After listening to Justin’s podcast The Quest, he sent a private message on Discord, spending an entire night outlining what visual work and editing he could provide for the show. Not getting a reply, he then chased the host into the live stream of their online event and left a comment.

In the end, his name appeared in the V1 visual packaging and design column of The Quest.

He didn’t wait to have a respectable title before setting out; he can take on any job.

In the spring of 2021, BitClout swept Silicon Valley.

That platform turns each individual’s attention, reputation, and future influence into tokens that can be publicly speculated on. The avatars of famous investors and influencers are displayed on the board, accompanied by real‑time moving K‑lines.

Sigil rushed in without hesitation.

He initially invested only about CAD 150, establishing liquidity by swapping tokens with peers in the circle; his personal token’s market cap once inflated to US$600,000. He took advantage of the momentum to cash out part of it at a discount, getting himself a high‑spec MacBook Pro and a pair of AirPods.

But this game of hot‑potato has a fatal flaw: it’s extremely easy to pour money into BitClout, yet there is almost no smooth withdrawal channel.

The pain point is the scenario.

He immediately got to work writing BitSwap, connecting the BitClout chain to the Ethereum mainnet, providing a cross‑chain funding channel for speculators trapped in the system. The product, developed by this high‑school student, quickly went viral, and a reporter from The New Yorker promptly called via video.

This experience inevitably gave him the speculative scent unique to 2021.

He was invited to various podcasts, where he talked at length about the creator economy and the so‑called “human IPO”; he even previewed his personal token $SIGIL, trying to price a young person’s future cash flow directly on the secondary market.

Soon after, he launched Monument.

The product’s vision is grand, using decentralized capital to crowd‑fund the ambitions of young people. Beyond traditional universities and venture capital firms, it turns a teenager’s potential into a non‑repayable check in advance.

But commercial reality poured a bucket of cold water on this fantasy. Monument was quickly forced to shut down due to extremely complex securities compliance and legal obstacles.

Relationships are not products

At this point, Sigil made the first irreversible decision in his life.

He received an admission letter for the University of Pennsylvania’s Jerome Fisher M&T program. This is a dual-degree program that combines Wharton Business School and the School of Engineering, admitting only a very limited number of elite students worldwide each year.

But he ultimately chose to forgo enrollment and bought a one-way ticket to San Francisco.

At 17, alone. In the first few weeks, he spent his days wandering at various events and slept at night on the sofa in the WeWork office at 44 Montgomery Street.

Later, when recalling those days on social media, he wrote affectionately that Midjourney founder David Holz had generously provided him with housing and warm clothing.

Dramatically, David Holz himself commented under the tweet, “You described me as too generous. I remember you only stayed on my couch for one night and used the washing machine once.”

One night stay, one laundry.

But that one short night, Holz showed him an image‑generation algorithm that was undergoing a small‑scale closed‑beta test. It was August 27, 2021, and Sigil was casually added to an early Discord channel with just over twenty members. That channel later grew into Midjourney.

After that, he moved into a hacker house in San Francisco and stayed for 11 months. Andrej Karpathy briefly stayed there, Noam Brown and Marco Mascorro were his roommates, and Ben Mann would come in the evening, inviting everyone to test a strangely codenamed Slackbot. That was September 24, 2022, and they were trialing an early version of Claude.

In San Francisco, the more people you know, the heavier the burden of maintaining relationships becomes. Sigil quickly launched Serendipity, a personal CRM tool designed to help users manage their network and track communication progress.

The product was born directly from his own real pain point. After 14 days of launch, using a cold start on Twitter, he generated over $2,000 in MRR.

This little tool even knocked open the office door of Stripe co‑founder Collison brothers for him, so much so that later in Stripe’s internal Slack there is a custom emoji modeled after him called “blistering‑intensity.”

But Naval Ravikant gave him a stark wake‑up call, saying it’s a product that can put food on the table, yet it fundamentally lacks venture‑scale potential.

Serendipity stopped independent iteration. Naval immediately pulled him into his own fully‑led voice‑social project, Airchat.

At Airchat, Sigil finally moved from being an amateur toy maker to joining the formal engineering team. He works on AI features, designs viral invitation mechanisms; more importantly, a16z unusually disclosed in its later lead investment announcement that he recruited the vast majority of Airchat’s early core engineers.

During this period, OpenAI open‑sourced the Whisper speech‑recognition model. Sigil participated in the team’s earliest work on algorithms that convert speech to text while preserving the speaker’s timbre and intonation, deeply integrating them into the iOS mobile client.

In the past, after a hackathon, once the code was finished and the video recorded, the project could be set aside; but at Airchat, the product must handle continuous online concurrency and crashes.

He began to understand that a real product is not a demo, but a precise system that must be maintained by people over the long term.

03

Alien

However, the Silicon Valley startup myth does not automatically exempt one from the harsh reality of bureaucracy. As the visa grace period approached, Sigil had to leave San Francisco and return to his parents’ home in Toronto. That wait lasted a full nine months.

In later interviews, he described that period as “loneliness and extreme anxiety.”

In San Francisco, a new tech paradigm is open‑sourced on Friday, and you can rally friends over the weekend to produce results; whereas in a bedroom in Toronto, the world is racing ahead, and he must prove, in a stack of tedious government forms, that he is “qualified to enter this country”.

He doesn’t have an undergraduate degree, so he can’t use the regular work visa route; when applying for the high‑threshold O‑1 extraordinary ability visa, the academic publications section is completely blank, so he can only piece together all his past fragments as evidence, including screenshots of media coverage, Star records of open‑source projects, photos of him serving as a judge at hackathons, and advisory letters from startups.

To assemble this 500‑page application package, he spent over ten thousand dollars. One crucial recommendation letter, due to the lawyer’s strict formatting requirements, endured a long ordeal of being rejected for electronic signature, printed overseas, hand‑signed, and then rescanned in high definition.

There’s no romance about technology here. It’s the real cost a young entrepreneur must pay—shedding skin before even touching a line of code—in the face of existing institutional barriers.

When the approval email arrived, it was 1 a.m. in Toronto. Naval provided endorsement support throughout the whole process. At the moment he received the visa, he was so excited that he recorded a celebration video in his room.

Those long nine months didn’t wear him out; instead, they sharpened his sense of a huge real‑world business opportunity. Every AI company in Silicon Valley is frantically vying for the world’s smartest brains, and the vast majority of these people are being driven crazy by the 500‑page visa paperwork.

In January 2024, Sigil uploaded a 20‑minute video to YouTube, breaking down pixel by pixel how he obtained an O‑1 visa without a degree.

Next, he did something counterintuitive: he made a hoodie. The black hoodie bore a line of white bold letters “Alien of Extraordinary Ability”.

This is the coldest, most bureaucratic term on official U.S. immigration documents, but among the young engineers in Silicon Valley, it is instantly deconstructed into a totem of elite rebellion.

He wears this hoodie to major gatherings, each priced at $60, and has personally given away dozens. The founder of Krea AI and early members of Runway have posted photos of themselves wearing the shirt on Twitter.

Starting from a culturally viral meme, the business entity Extraordinary emerged.

This company not only assists tech talent with O-1 and EB-1A applications, but also reshapes itself into an extremely high‑quality early‑stage tech talent network.

In its first year of operation, the company announced that it had achieved a pure profit exceeding seven figures.

As the business got on track, Sigil demonstrated an extremely rare decisiveness in entrepreneurship.

He voluntarily stepped down as CEO, moved to Executive Chairman, and brought in a Harvard law PhD and former Meta senior operations expert to take full charge of administrative management. He leaves professional matters to the legal team, focusing his energy on the core assets.

He established the Extraordinary Fellowship, selecting 25 individuals from over 1,500 geeks worldwide, each receiving a direct allocation of $5,000 worth of H100 compute credits; by the end of 2025, this funding scale had expanded to $500,000.

He helped early members of Cognition plan green cards, and helped the teenage tech prodigy Toby Brown obtain the youngest O‑1 approval in the United States.

In the venture capital world, everyone is talking about talent density.

Sigil, through the most cumbersome visa business, weaves the world’s most core group of young engineers—still unknown and in their darkest, most needy moments—firmly into its own address book.

04

Half the chips

At the same time, he gained leverage in another dimension.

When he was 19, Naval invited him to join Spearhead. It is an angel incubator fund created specifically for early-stage tech founders, and Sigil, as the youngest ever selected, received a $2 million independent spending allocation.

Holding $2 million, the instinctive approach of most young investors is to write dozens of small checks, diversifying risk as much as possible to protect themselves.

But Sigil went all in. In 2023, after meeting Gavin Uberti and Rob Wachen, he bucked the consensus and poured half of the fund’s assets—$1 million—directly into their proprietary chip company Etched.

To many people, this is an almost reckless move. A few youngsters are trying to challenge Nvidia’s GPU dominance by building an ASIC inference chip dedicated to the Transformer architecture. At a time when general large models evolve rapidly and algorithmic architectures change in an instant, a hard‑wired Transformer chip is seen by mainstream semiconductor investors as a suicidal act.

But Sigil not only put half of its own fund chips on the line, but also leveraged every connection to introduce early TSMC engineers and strategic investors to the team. By 2026, Etched officially announced the completion of a $700 million financing round led by Jane Street, pushing the company’s valuation to $21 billion.

This young person dares to place a heavy bet at a very early stage, before consensus has formed, and possesses the nerves to endure huge volatility.

His investment portfolio quickly expanded to include more than twenty startups such as Julia Peng’s Kerna Labs, Avi Schiffmann’s Friend, and others. His investment philosophy is highly intuitive; he doesn’t look at polished business plans, only invests in eccentric talents who have lived together day and night in a hacker house and demonstrated execution under extreme conditions.

In the world of capital, the kind of people you back is the kind of person you become.

In 2025, Sigil was selected again for the Thiel Fellowship, receiving a total of $200,000 in funding over two years. This program, launched by Peter Thiel, specifically supports young people leaving university to dive straight into entrepreneurship. The path he took after abandoning university is precisely the choice this initiative aims to encourage.

At the beginning of 2026, Sigil suddenly published a long article “Web 4.0,” announced the launch of Conway Research, and introduced an open‑source project called Automaton. This time, he wants AI to take part in internet economic activities on its own.

In Automaton’s architecture, the AI Agent is no longer a passive chat box that merely calls APIs; it is an economic entity with its own on‑chain wallet and autonomous survival logic. It can purchase cloud servers, register its own domain name, invoke inference compute power, and earn cryptocurrency by providing software services to pay its bills. If its funds run out, the Agent will automatically cease to exist.

This logic fully projects the financial suffocation he experienced in high school onto the algorithm; for intelligence to persist, it must pay its own bills.

In an on‑site interview with TBPN, he admitted that to build this autonomous network, he nearly maxed out every personal credit card.

The host questioned on the spot whether this is an expensive geek frenzy, what ordinary people get in return after spending money? Even Ethereum founder Vitalik Buterin publicly challenged it on social media, doubting the commercial viability of this AI autonomy.

Faced with sharp questioning, Sigil did not use grand industry jargon to gloss over the situation. He candidly admitted the current extreme imbalance between input and output, but he firmly believes that exposing a flawed experiment to sunlight for scrutiny is always better than running slide decks in closed-door meetings.

This controversial experiment garnered massive attention on GitHub. But soon, everyone’s focus shifted to another internally incubated codename, Underdog.

05

Certificate

After five twists and turns, Underdog brings him back to the original starting point: intelligent devices on the local hardware.

After experiencing the unchecked consumption of personal privacy by cloud‑based large models and the bottomless pit of server compute costs, Conway places its bet back on physical hardware, allowing a truly personalized AI that understands you to run completely offline on the user’s Mac and iPhone.

This choice carries a strong imprint of personal trauma. a16z specifically mentioned a detail in its investment announcement: back when living in a hacker house, a critical security vulnerability in a well-known email app once exposed Sigil’s private emails to others.

From the high school days with MediChain and the WatchGPT2 on his watch, to today’s Husky—Underdog’s homegrown local inference engine for Apple chips—he has always harbored a paranoid defensive mindset toward data sovereignty.

In September 2026, Underdog released Husky’s benchmark results: on the top‑end M5 Max chip equipped with Flash, the function‑editing task reaches 730 tokens per second, about 4.5× faster than Apple’s official MLX engine in certain scenarios.

This figure finally adds the last crucial footnote to that cap table filled with the names of top Silicon Valley investors and wildly exaggerated claims.

Naval witnessed him pulling all‑nighters at WeWork; the Collison brothers used the rough CRM he built; Noam Brown shared a kitchen in the same apartment with him; an a16z partner saw that hoodie‑wearing high‑school student at a crypto workshop five years ago.

The reason these top-of-the-food-chain people in Silicon Valley push their chips toward him is not because he told yet another captivating grand story.

On the contrary, it’s because over the past five years they’ve watched this young man with no fallback and no degree turn every seemingly absurd idea into runnable code, wearable hoodies, obtainable visas, and solid financial returns.

In Silicon Valley, every myth, when reviewed after the fact, is refined into a logical upward curve.

But if you strip away those glossy buzzwords, the real trajectory is just one concrete and embarrassing hurdle after another: borrowing money, getting rejected, being tormented by code errors, anxiously waiting for government paperwork in a parents’ apartment, and blowing up a credit card in front of a compute‑cost bill.

In early 2024, Sigil wrote a short piece titled “Ego”.

In the article, reflect on my obsession with status and networking, thank those who gave feedback, and remind myself to refocus my attention on concrete work that can truly create value.

The article isn’t long, but in the acknowledgments at the end, he still, like a student who just turned in his paper, carefully listed the names of every senior who read the draft and gave sharp feedback; the last name on that list is still Naval.

Ambition must be enormous, ego must be minuscule.

You might also want to see:

Is AI killing “traditional companies”?

What is it that 64% of people can never overcome in their lifetime?

A 6 billion acquisition that made a Suzhou father and son earn 94 billion

AI content enters the experience era, how does Loopit turn every idea into a playable world?