Dylan Field 的 Thiel 奖学金申请书

During my Flipboard internship I was frustrated using Fireworks every day. Why couldn’t design tools be more like Google Docs? So Evan and I decided to use WebGL to build a cloud first design tool in the browser.

While this story is technically true, it’s just one small part of a much longer saga. On the path to the Figma of today, there were so many twists and turns — we did not wake up one morning with a fully formed idea.

Because of today’s attention economy, most of the “founding stories” you hear favor narrative simplicity over a deep dive into the messy, complex and often existential process of starting a company. As a result, first time founders have unrealistic expectations for how long it takes to build something of value.

In the rare occasions where founders do share the messy details, their memory is not always reliable. As a species, entrepreneurs are optimists with very high pain tolerance. We forget the painful parts and skip over key details that no longer serve us.

Given these challenges, how can we accurately tell any origin story? My point of view is that the only “True” way to recount a company's founding is through sharing artifacts from the past. In this post, I’ll share my Thiel Fellowship application, one of the first steps on Figma’s journey. And in future posts I’ll follow Reid Hoffman and Mathilde Collin's lead and share annotated versions of Figma’s Seed, Series A and Series B pitch decks.

Startups are creative endeavors and they usually don’t follow clean, predictable trajectories, especially in the early days. The “perfect idea” rarely comes from fancy spreadsheets and market analysis. My original pitch in the Thiel Fellowship application (below) was around drones — quite far from creative tools! While I didn’t have perfect confidence in what we were going to build, I did know that I wanted to work with Evan Wallace — the TA for several of my classes at Brown and also the smartest, kindest, most humble person I had ever met. A two year program, the Thiel Fellowship offered time and financial resources that enabled us to play and explore. This eventually led to the Figma of today.

The application I submitted in 2011 is below. It’s unmodified except for the removal of private data, even in places that make me cringe. 2020 commentary in italics.

My hope is that by sharing these artifacts, others will feel permission to embrace ambiguity and take a similar path. Despite the conventional wisdom, don’t let the lack of the “perfect idea” stop you. Have faith in your abilities to iterate towards a meaningful product / business even if you don’t yet know exactly what you are going to build.

Submission Timestamp: Sat, Dec 31, 2011 at 10:07 PM

Legal Name: Dylan Field

I am applying as: an individual

Optional - Personal or Company Websites: dylanfield.com

Optional - LinkedIn Profile: http://www.linkedin.com/in/dylanfield 

Optional - Publications about you:

Optional - Publications written by you:

Currently I am: stopped out of school

Most recent High School, College, or Uncollege: Brown University

I'm currently a... Junior

Please list all the colleges to which you applied for admission: Brown, Olin, UC Berkeley, UC San Diego, UC Irvine, UC Davis, UC Santa Cruz, Cal Poly San Luis Obispo

Please list any colleges to which youʼve been accepted for a degree program: Brown, UC San Diego, UC Irvine, UC Davis, UC Santa Cruz, Cal Poly San Luis Obispo

Were you homeschooled? No

High School GPA and maximum possible GPA: 4.23 / 4

College GPA and maximum possible GPA: 3.75 / 4 (Calculated)

Please describe your activities and clubs, your role in them, and what you accomplished. This needn’t be an exhaustive list. (200 words maximum)

Over the past year I’ve been very involved with Brown’s Computer Science Department Undergraduate Group – the “CS DUG” for short. When I first arrived at Brown, I was amazed by the knowledge and creativity my classmates possessed. However, I sadly realized that while brilliant, most of the students ended up working for big companies and didn’t do many independent projects outside their classes. Hoping to change this culture, I joined the DUG in the fall of my sophomore year and organized the New England College Hackathon (nech2011.com) in partnership with Microsoft and Zynga. There were over 120 attendees and 20 hacks were submitted.

This fall I was elected President of the DUG and worked with my classmates to further change the department’s culture. We hosted events around entrepreneurship, offered Friday “Hack Nights” as an alternative to frat parties, enabled students to easily advertise their independent projects, and worked on a department remodel that will take place this spring. Lastly, we raised awareness of the DUG inside the department and doubled our membership. It’s difficult to quantify culture, but we’ve already noticed a large uptick in the number of students creating apps outside the classroom and considering entrepreneurship.

2020 commentary: it’s worth noting that the hackathon mentioned above was a complete disaster. (They say comedy is tragedy plus time, and nine years later it’s still not funny…) Thankfully, the culture of entrepreneurship at Brown seems to be getting stronger! I’m especially impressed with campus groups like Hack@Brown.

Please list or describe any achievements and prizes. (200 words maximum)

2020 commentary: To be clear, 2nd place for Putnam was out of the people who took the exam at my local university. I’m not actually that good at math... my biggest regret from my time at Brown is not taking more math classes. (My second biggest regret is not taking RISD classes.)

If you feel any of your answers require context (such as a country-specific issue or an extenuating circumstance), please explain here (200 words maximum).

I am intentionally choosing not to submit my SAT scores as part of this application. It is my belief that the SAT is a poor reflection of aptitude and can easily be gamed. The concept of standardized testing is contradictory to two values the Thiel Fellowship supports: lifelong learning and independent thought. I hope this choice does not disqualify my application.

Have you ever started a business, led an organization, or dedicated yourself to a major project?

Yes.

If yes, in 500 words or fewer, please describe the organization or project you started, your reason for starting it, and your role.

I interned for LinkedIn’s Data Analytics team the summer after my freshman year. My mentor, Pete Skomoroch, discovered an innovative way to extract and disambiguate skills from LinkedIn profiles. While the algorithm had room for improvement, there was already a treasure trove of data. For any profile, we could identify the skills each user had. We could also return an ordered list of the most connected users for each skill. To make the data available, Pete and I developed informational pages for each skill. However, we had a gut feeling that there were more ways the data could be used.

One of my tasks for the summer was to explore new ways to use the data. I designed visualizations, discovered trends and created games. Then, at the end of June, Pete and I saw a presentation by a computational linguist named Robert Munro. Munro had just finished a project where he crowdsourced the translation of Creole text messages after the Haiti earthquake. In the presentation, Munro described how the main bottleneck was not a lack of ground responders in Haiti. Rather, the main challenge he faced was finding people that could translate Creole to English.

This sparked an idea: perhaps the Skills data could be used to match volunteers and NPOs. I started to research the non-profit space and realized this was a real problem that LinkedIn had the solution to. With backing from the data team’s leadership, I presented the idea – “LinkedIn for Good” – to product review. However, one of LinkedIn’s top executives pointed out that my solution could not scale to every non-profit. I was told to return to my normal work.

I followed their instructions, but each day I returned to the project after completing my normal duties. I worked out some of the kinks in the idea and partnered with LinkedIn’s VP of Marketing, Patrick Crane, to make LinkedIn for Good a track that employees could work for on LinkedIn’s first “InDay.” InDay, a Hackathon for all members of the company, was a huge success. Over forty people (at the time 6.5% of the company) decided to work on LinkedIn for Good. In just one day, we finished a product spec, created mock-ups, developed a basic prototype, designed a QA process to validate NPOs and defined a go-to-market strategy. Just one month after I was told to stop working on LinkedIn for Good, I found myself pitching the project to the entire company during an all hands meeting. Jeff Weiner, LinkedIn’s CEO, hailed the project as a success and suddenly the entire company was buzzing about the idea.

Shortly afterwards I finished my internship and returned to school. Initially I was worried that the project would die out. However, LinkedIn hired Meg Garlinghouse, the former director of Yahoo’s “for good” effort to spearhead the initiative. The project has changed significantly over the last year and a half. Although my name is no longer attached, it’s exciting to realize that the seed I planted as an intern has grown into a major initiative for LinkedIn.

Tell us about a risk youʼve taken or a challenge you've faced. Tell us whether you failed or succeeded, how you behaved, and how you think this reflects your character.

One week after I graduated from high school, I moved to Berkeley and worked with Jessica Mah and Andy Su on a startup called Indinero.com: the “Mint.com for small businesses.” We coded all summer and built a (very) minimal viable product, but in the end we had trouble raising money. Jessica and Andy offered to make me a cofounder and give me a sizeable amount of equity if I stayed onboard and deferred Brown.

This might sound odd, but I viewed enrolling at Brown as a risk. I truly believed Indinero had a lot of promise. However, at the end of the day I just wasn’t very interested in accounting. My gut told me I should go to Brown so I declined their offer. Today Indinero is by some measures quite successful – they got into YC, received $1.5 million in angel funding and have paying customers. Despite their success, I believe I made the right choice by following my instincts.

What expertise do you have to execute on the work that you want to do?

When I started middle school, I was lucky enough to participate in an annual LEGO Mindstorms robotics challenge. I discovered my love of code through some of the harder tasks like a game called “Woots and Snarks.” After graduating, I helped run the competition as a volunteer and judge.

I chose to attend Technology High School for two reasons: the school’s FIRST robotics team and the opportunity to take math classes at a local state university. If I wasn’t so engaged with FIRST and higher math, I probably would have dropped out of high school. Instead, I formed compromises with my teachers where, provided I kept my grades up, I could skip class and play with robots or work through math problem sets.

At Brown, I’ve taken classes in Robotics and Computer Vision. Outside the classroom, I’ve learned how to process large data sets – a skill I think will be increasingly important as we move into the world of cloud robotics. Unfortunately, I have very limited electrical engineering background and almost no mechanical engineering background. However, I don’t think my lack of engineering background will block progress given the wide range of low cost drones available in today’s market. If it is a problem, I think I can work with others to fill this gap in expertise.

Who is your favorite philosopher, writer, entrepreneur, scientist, poet, economist, or historical figure? In six words or less, not counting his or her name, explain why.

Leonardo da Vinci: curious, imaginative polymath who understood beauty

2020 commentary: OK, so this answer was really cliché

Did you apply for a Thiel Fellowship last year? No.

What do you want to accomplish in the next 10 years?

Ten years from now I hope to be growing a disruptive, profit generating company that has made a significant contribution to the world.

What do you want to accomplish in the next 2 years?

Two years from now I hope to be growing a disruptive, profit generating company.

What do you plan to accomplish in the next 3 months?

During the next three months I will travel to Israel, spend two weeks with the Obama campaign and work in a product role at Flipboard. (2020 commentary: I didn’t end up working for the Obama campaign)

Tell us one thing about the world that you strongly believe is true, but that most people think is not true. If this belief shapes the way you live, tell us how.

Chocolate is repulsive. Even the smell of it makes me want to vomit. Although I have other beliefs that distance me from the majority, no conviction elicits a stronger reaction than admitting I detest chocolate. As a young child, I quickly realized I was not normal. One of my first memories is of a little girl frankly asking me on the playground if I was an alien. Others simply labeled me as a freak.

Even today, my peers are shocked when I reveal this preference. Just a few weeks ago a girl I was interested in dating offered to share her dessert with me. “Too full,” I told her. “Come on,” she implored. “It’s soooo good. Just take a bite.” Finally, I admitted the truth. At first she just stared at me. Then, slowly, she turned away. “We can’t be friends anymore.” She told me.

According to research into hours of baby videos, I have never enjoyed chocolate. My mother tells me the first time I was allowed to try any sweet was on my first birthday. It was supposed to be a joyous occasion. I don’t have any recollection of the incident, but footage shows me happily sitting in a high chair surrounded by family friends. A guest blows out the candles on a large, chocolate cake and I am fed the first slice. My face changes from joy to confusion to visible disgust – then the video ends. I like to imagine that I spat it back out.

Over the years, I’ve been offered chocolate countless times. At first I refuse. Some people, like the girl I mentioned earlier, are persistent. After I tell people that I don’t like chocolate, they react in one of two ways: either they immediately classify me as a mutant or inquire about the condition. Many ask if I’m allergic. I’m often tempted to lie and end the discussion, but the truth is that I’m not allergic. I’ve forced myself to swallow the black goo before and – besides having the impulse to make myself throw up – nothing bad has happened. I just abhor the taste.

While my hatred is rare, I’m not alone. A fellow chocolate hater and writer for Gilt Groupe found a study by Hofstra University that determined an entire 2% of the population doesn’t like chocolate. My personal theory is that hating chocolate is a recessive trait – my grandma also disliked it.

One day I intend to research the phenomena and find out the truth. Scientists at Nestle wrote in a paper from 2007 that my belief is due to an imbalance of gut bacteria, but I’m not sure if I believe the results given their sample size of 11. Perhaps the researchers are correct and one day my belief will be changed through science. In the meantime, I’m perfectly happy with a simple sugar cookie or a slice of fresh blackberry pie à la mode – vanilla, of course.

2020 commentary: Silicon Valley is obsessed with “contrarian” ideas. (Recently people have started calling them “narrative violations” — same difference.) The obsession stems from venture capital firms… it’s really hard to be a good investor if you don’t have unique insights about the future. In any case, the whole “contrarian” thing has always bugged me for some reason, even back in 2011, so I thought it would be fun to be “meta contrarian” in this answer.

How are you going to change the world?

Over the past few months, Unmanned Aerial Vehicles, also known as UAVs or drones, have received significant media attention for their use in the war on terror. However, the most interesting application of drone technology flew under the radar. On June 23rd, three men with rifles chased a sheriff off a farm in North Dakota. The sheriff responded by calling in a large team of reinforcements and a Predator drone. While the reinforcements waited around the farm’s perimeter, the drone identified the suspects’ location and watched from overhead. Once the suspects were unarmed, the officers were sent onto the property.

This was the first known civilian arrest conducted with the assistance of a UAV. It has inspired discussion about the way drones may be used in the future to aid police in manhunts or hostage situations. Personally, I am more excited about low hanging fruit that the media is ignoring: monitoring traffic and catching reckless drivers.

Our current method for catching reckless drivers is inefficient. Police officers are distributed to strategic locations and wait inside their car to catch speeding or inebriated motorists. Because the risk of getting caught is low, drivers will often ignore speed limits and simply slow down when they see a police car. If a speed trap is well known it can easily be avoided.

In contrast, UAVs can monitor a much larger area than any single officer. Even with the price of maintenance and gas, drones can reduce law enforcement operating costs and save taxpayer money. The technology already exists to directly replace police helicopters with UAVs. Eventually an officer will be able to control multiple drones and dispatch officers to intercept reckless drivers. It’s not difficult to image a future where traffic control is completely autonomous with drones automatically identifying reckless drivers, recording license plate numbers and notifying drivers when they receive a ticket. Officers would have to be dispatched only in the most drastic of situations.

The prospect of using UAVs in civilian settings faces three limiting factors: software, battery life and the FAA. Battery life is gradually improving and the FAA just approved pilot programs for civilian drone use. However, software needs to be developed so that fewer operators are needed. Among other features, the software also needs the capability to determine how fast a vehicle is traveling.

I am going to change the world by creating better software for UAVs. After I finish at Flipboard, I will cofound a company with the smartest programmer I know and work on this problem. In the meantime I am learning as much as I can about the UAV space and adapting ROS (http://www.ros.org/wiki/) to low-end drones.

This is a crazy idea. I’m looking forward to working with the Thiel Foundation to make it a reality.

2020 commentary: when I was writing this, I was totally obsessed with drones. Evan (who had actually spent a lot of time building / programming drones) was already pushing for us to work on something related to WebGL instead. A few weeks after I submitted this application, Evan convinced me that we shouldn’t be in the drone space because (a) the run / debug cycle for hardware is annoying (b) regulation was a wild card (c) we couldn’t come up with a drone idea that didn’t hurt people or violate their privacy. Looking back, I regret writing this essay — as someone who cares a lot about fighting for privacy, this is not a technology I want to exist in the world.

What questions do you spend the most time thinking about?

1. What is the definition of privacy? Is it possible to account for privacy while developing an application?

I thought at length about the nature of privacy when working as a research assistant to danah boyd, a sociologist studying how teens use technology, during my freshman year at Brown. At the time, Facebook was being accused of exposing user information and obfuscating privacy controls. It’s easy to criticize others mistakes, but when I interned at LinkedIn and Flipboard I realized that baking privacy into a social product is difficult.

In addition, my first question contains flawed assumptions; there is no static definition of privacy. The norms around what is considered private have changed dramatically in just the last few years. While it’s impossible to define privacy, it is possible to define a violation of privacy. A violation of privacy is the delta between a user’s expectations and reality. Therefore, the best answer I have found to this question is to understand and respect user expectations throughout the development cycle. However, I return to this question often in search of a better solution.

2. How can content be distributed better?

For the past century, mainstream media has had a monopoly on the creation and curation of content. Now anyone can post a tweet, start a blog or upload a video. Flipboard has improved the experience of digesting content, but distribution of content is still an unsolved problem. I spent a lot of time asking myself this question and look forward to implementing the answers I’ve come up with when I return to Flipboard. (Unfortunately I can't write about specifics.)

3. Who owns data?

Big data applications are fueled by the aggregation of users’ individual data. Often, the sum is greater than the parts; any individual’s data is almost worthless, but all the data combined is valuable. Should users be notified if data is being collected about them? Do they have the right to get their data back out of the system? I asked these questions and more on a panel at Strata 2011 (http://strataconf.com/strata2011/public/schedule/detail/17602) but we didn’t come up with many answers.

4. What will we trust when anyone can manipulate visual data?

The most mind-blowing paper I read this year was “Rendering Synthetic Objects into Legacy Photographs.” (Karsch, et al, 2011) In the paper, Karsch describes how to realistically insert models of objects into photographs or videos of scenes with minimal human annotation. The results are so good that I can’t tell they are doctored. Words can’t do this paper justice – if you have the time, it’s worth watching a video describing the work: http://kevinkarsch.com/publications/sa11.html.

This paper has huge implications. Currently we regard visual information as universally true – we don’t worry that photographs or videos have been altered. When every person can manipulate photographic reality, we will no longer be able to trust visual data. In the future, how will we navigate trust? Will we develop verification schemes? Rely on centralized authorities? Create new content types?

2020 commentary: not going to lie, I’m kinda proud of calling deep fakes back in 2011.

Tell us, in just six words, why you should be a fellow.

I'm a hardworking, innovative risk taker.

That's it! In the coming weeks, I'll also share our Seed / A / B decks to show how Figma developed over time. If you have any questions about this post, please leave a comment below. Thanks for reading and stay safe!