apply for Y Combinator W2017

[chadwhitacre]

Working Drafts

• Gratipay
• Chad
• Kwesi
• video

Previous

• 2017, pre-Walker
• 2014
• I don't find a ticket for 2012.
• 2005(?) is lost to the sands of time. :)

Next

Todo

• percentage growth numbers by month since Gratipocalypse
• surprising or amusing discovery
• video
• bios
• LOC

[chadwhitacre]

Due date is Oct 4 ... 6 days!

@kaguillera is tentatively willing to apply together, he and I, as co-founders. We're thinking in terms of a non-profit application.

Need a one-minute video. We'll have to do that tomorrow afternoon.

[chadwhitacre]

We also need LOC for gratipay/gratipay.com. Anyone want to compute that? :-)

https://www.google.com/search?q=github+loc

[chadwhitacre]

I reached out to the Teespring founders, after seeing their video as one of the examples. They are one of the few corporate givers extant on Gratipay 2.0.

[chadwhitacre]

The founder bio forms are almost as extensive as the main application.

[chadwhitacre]

I drafted some answers on the YC form. I don't like them. Gonna sleep on it and see what happens tomorrow ...

[chadwhitacre]

Google Docs for Gratipay, Chad, and Kwesi.

[chadwhitacre]

Heard back from Walker from Teespring and moved the conversation to private email. I will follow up when our application is ready for review.

[chadwhitacre]

https://www.ycombinator.com/video/

[chadwhitacre]

Gotta work on numbers. We need percentage growth, not absolute growth.

[chadwhitacre]

Video! Takes one through four:

1. https://www.youtube.com/watch?v=DU5G1-5KTPQ
2. https://www.youtube.com/watch?v=FDVb7zLyPbg
3. https://www.youtube.com/watch?v=ys2wh_XFvAE
4. https://www.youtube.com/watch?v=6TAZdt-9s8U

[chadwhitacre]

Planning to go with 4, the rest are kinda fun though. :)

[chadwhitacre]

Okay! We have first drafts of the application and our bios. Outta time for today. We're planning to go over those again tomorrow and get something to Walker, and then to cowork again on Monday so we can incorporate feedback and get something submitted.

[techtonik]

If you had any other ideas you considered applying with, please list them. One may be something we've been waiting for.

1. On-demand allocation. Because we're radically transparent, we can publish requirements for every person on the "payroll", which is living_cost + amortization + savings. We can start from living_cost. Then we can show it the minimal living_cost criteria is met and let people fix that by changing their allocation scheme.
2. 100 GP. Every week, every person gets 100 GP. which are their 100% to distribute among their friends and other people. You allocate how much do you give out monthly and it gets distributed among people who are valuable to you. This way, if some open source folks are bought by some company, that company can compensate the loss by donating weekly funds over the person's distribution graph.

[techtonik]

Please tell us something surprising or amusing that one of you has discovered.

Lawyers can't give official advice in public even if client demands it.

[chadwhitacre]

Let's compute a team growth rate from the chart on stats:

[chadwhitacre]

We want a monthly growth rate.

  211 - 144 = 67
/ 226 - 187 = 39
  --------------
               1.718 teams / wk
x              4.333 wks / mo
  --------------
               7.444 new teams / mo
/            211 teams
  --------------
               0.035

So call it 3.5% team growth right now.

[chadwhitacre]

https://twitter.com/whit537/status/781969297132228608

[chadwhitacre]

Hey @kaguillera, we have an appointment Monday at 11 with @dowski (video call) to go over our application together.

@dowski Are we gonna live-stream that? 😳 (Could go either way, up to you ...)

[chadwhitacre]

https://twitter.com/whit537/status/782049483207041025 💃

[chadwhitacre]

I love how it's leaping. "Ooh! Ooh! Pick me! Pick me!" 😆

[chadwhitacre]

https://github.com/AlDanial/cloc

$ npm install cloc
[...]
$ cloc --exclude-dir=env,node_modules,vendor .
     716 text files.
     643 unique files.                                          
     323 files ignored.

github.com/AlDanial/cloc v 1.70  T=4.21 s (94.0 files/s, 19233.7 lines/s)
--------------------------------------------------------------------------------
Language                      files          blank        comment           code
--------------------------------------------------------------------------------
HTML                            133           4206              5          25730
Python                          131           3617           3339          10894
YAML                             13              0              0          10549
JavaScript                       37           2385           2413          10314
SASS                             52            182             12           3175
SQL                               9            226            128            930
Markdown                          4            267              0            802
CSS                               6            190             39            754
Bourne Shell                      7             96             47            300
DOS Batch                         1             29              1            212
make                              2             38             11            120
Bourne Again Shell                1              3              1              7
--------------------------------------------------------------------------------
SUM:                            396          11239           5996          63787
--------------------------------------------------------------------------------
$

[chadwhitacre]

$ cloc --exclude-dir=tests/py/fixtures .
--exclude-dir 'tests/py/fixtures' :  cannot specify directory paths
Use '--fullpath --not-match-d=REGEX' instead
Died at ./node_modules/.bin/cloc line 640.
$

Blech.

$ cloc --exclude-dir=env,node_modules,vendor --exclude-lang=YAML .
     716 text files.
     643 unique files.                                          
     336 files ignored.

github.com/AlDanial/cloc v 1.70  T=4.63 s (82.8 files/s, 15226.3 lines/s)
--------------------------------------------------------------------------------
Language                      files          blank        comment           code
--------------------------------------------------------------------------------
HTML                            133           4206              5          25730
Python                          131           3617           3339          10894
JavaScript                       37           2385           2413          10314
SASS                             52            182             12           3175
SQL                               9            226            128            930
Markdown                          4            267              0            802
CSS                               6            190             39            754
Bourne Shell                      7             96             47            300
DOS Batch                         1             29              1            212
make                              2             38             11            120
Bourne Again Shell                1              3              1              7
--------------------------------------------------------------------------------
SUM:                            383          11239           5996          53238
--------------------------------------------------------------------------------
$

[chadwhitacre]

Okay, call it 50,000+ LOC.

[chadwhitacre]

Are we gonna live-stream that? 😳 (Could go either way, up to you ...)

Per private email, we're not gonna live-stream.

[chadwhitacre]

What if standard practice in the tech industry were to give 1% of profits to open source?

https://en.wikipedia.org/wiki/List_of_the_largest_information_technology_companies

The top 13 largest sum to 1.3 trillion in revenue.

Let's estimate 2x for the long tail, so 2.6 trillion in revenue for the industry as a whole.

Apple's profit margin is 40% and Amazon's is less than 3%. Let's estimate a 10% average. Actually, let's go with 7%.

So that's $182 B in annual profit.

If 1% went to open source, that'd be $1.82 billion in revenue for open source, or about the size of Red Hat.

We want to build the next Red Hat, but decentralized.

[chadwhitacre]

Okay!

!m @dowski 🌻 💃

[chadwhitacre]

We spent 20 minutes on a video call together. @dowski grilled us for most of it, a little interview role-playing. The interviews themselves are only 10 minutes long, clearly not enough time to get everything across. I guess mostly it's a chance to demonstrate that we're competent, that we understand what problem we're trying to solve and how we intend to solve it, that we're the experts on this problem.

[chadwhitacre]

Which we are.

[chadwhitacre]

"I'm here to redo these."
"Of course you can, you're at FedEx Office, player!"

💃

[chadwhitacre]

Just talked to @kaguillera by phone. Doesn't make sense for him to (complete the) trek down here, I'm gonna be out soon.

[kaguillera]

Well I got my exercise for the day. Finally reached back home. I missed the bus going and coming 😦

[chadwhitacre]

Just mailed this to print:

yc-brochure.pdf

[kaguillera]

Posting the pics even though they are of the old brochure





And yes I am smiling.

[chadwhitacre]

Okay! Printed and folded. I was able to reuse the pennies from the previous brochures along with their stickies, so your handiwork is still going to YC, @kaguillera. 👍

Gonna check out and head home ...

[kaguillera]

👍

[chadwhitacre]

Handout code is in gratipay/gratipay.com@2919d10.

[chadwhitacre]

So to be honest I think @nobodxbodon is right. We have no business presenting a handout with #fail as the central, most prominent thing on it. This echoes Walker's advice in point (5) at #836 (comment). I think we maybe need to let go of the past. Let go of the Gittip bump. Start fresh. Beginner's mind. Think of ourselves as pre-launch.

[nobodxbodon]

Sorry to bring this up at this moment. Pitifully I missed the discussion of switching to non-profit, which I vaguely recall but can't find the exact issue (could you point me to it?). I heard from a speech about non-profit business that, non-profit usually is harder to run than for-profit business, which I believe you already know well. IMO, a for-profit business can stand out more with the openness and transparency, compared with a non-profit can (openness in non-profit is basically obligation).

Another aspect is, there are way more for-profit businesses than non-profit, and if Gratipay can find a successful business model, it will be more influential, especially if it's potentially reproducible by other for-profit businesses.

[chadwhitacre]

Good points. My current thinking is that my main goal today is to meet people. I am backing off from #867 (comment), in other words. Now thinking, "I was hoping I could hang out and watch and maybe help with some mock interviews."

[mattbk]

if Gratipay can find a successful business model, it will be more influential, especially if it's potentially reproducible by other for-profit businesses.

👍 I think we want to be influential and show that this stuff can work (along with letting people run experiments*, of course).

I'd argue additionally that, unless there is a pressing reason, there is no need to move toward nonprofit right now.

*I don't know who brought this up (maybe @rohitpaulk? My search-fu is weak this morning see ETA at bottom), but one potential use for allowing multiple split-up-the-money methods (beside take-what-you-want) and multiple ways of giving (bundles, corporate, others we've brainstormed) is to see what works financially and what doesn't. With enough entities using Gratipay, it could be a good resource for some economic papers.

ETA: It was this issue and also this comment within this very issue!

[kaguillera]

FYI got a text from @whit537 a while ago, that he landed and is heading to YC office.

[chadwhitacre]

Golly, what a day! Let's see here ...

[chadwhitacre]

could you point me to it?

#72 (comments copied)

[chadwhitacre]

Alright, so! I arrived in Mountain View at about 12:30 pm local time. I went to Olympus for lunch. Then I remembered how well it went over at All Things Open (#757) when I bought a dozen good croissants and brought them back to the conference to share (the Percona CEO in particular was quite surprised—"Wait ... you're bringing us croissants at our table?" 💃 ). So I went back inside after lunch and bought a half-tray of baklava (and an espresso ;). While waiting, I made small talk with a DoorDash driver to remember how awesome humans are, and then I sallied forth!

It was about a mile walk outside of downtown to get to the Y Combinator offices. They have two buildings, across the street from each other on a cul de sac. Teams of founders were meandering about in various states of nervousness and relief. Occasionally a pod of YC partners would emerge. Registration was in one building—the one with the iconic sign out front (I learned from one founder that it's made from what is I guess called "eighty twenty," an archetypal prototyping material)—and interviews were in both buildings. So people were constantly walking back and forth, and the weather was fantastic and the street was quiet (since it's a dead end), so people were milling about outside as well.

I greeted a few founders on the way up to the main building entrance, and entered with nervous confidence. The registration desk was staffed by one Scott Bell. Our conversation went something like this:

"Hi, is this the registration desk?"
"Yes, it is."
"My name's Chad, what's your name?" [extends hand]
"Scott." [shakes hand] "What company are you with?" [looks at iPad]
"Soooo Scott, here's the thing. I wasn't invited to be here. I was sort of hoping I could just hang out."
[adopting firm, bouncer-ly tone] "Absolutely not. You have to leave. We have a lot of founders that were invited to be here and it wouldn't be fair to them."
"Okay." [pause] "May I leave this here, though?"
"Um, what is it?"
"Baklava."
[half-smiling, I think] "Sure, okay."
[walking away, calling back] "I don't even want to interview, I just want to hang out."
"Nope."
"Okay! Okay ... I'll just be outside ..."

And there you have it! I decamped to the rain gutter across the street, plopped my bags down by the fire hydrant, and proceeded to have a fantastic time meeting other founders for the next five hours. It was awesome because I had absolutely no stress, since I wasn't interviewing. I became very comfortable with the proceedings. I heard probably fifteen companies stories (so I met probably 40 or 50 people). In turn, I was able to tell Gratipay's story to a number of them. "We help companies fund open source. Do you use open source? Do you pay for it? Who does?" Etc. I took a lot of pictures of teams posing by the sign. I cheered, especially for the color-coordinated crews ("Go red team!").

Re: Gratipay. There were a couple guys whose names I don't remember who really put the heat on, which was great! One of the two in particular zoomed right in on the weakest part of our story: companies have only weak incentives to pay for open source software. "Who pays for it?" "Red Hat and Intel." The guy (I wish I remembered his name! Azi?) was familiar with the concept of a public good and said straight away, "The only two ways to pay for a public good are to enclose it so it's not a public good anymore, or to tax people for it." I didn't come back strong at the time, but on the train I realized a good way to frame the third option that we're pitching: peer pressure.

Peer pressure is precisely why YC would be so valuable for us. They're such a tight-knit community. If YC buys into the idea of tech companies giving 3% of their profits to open source, then they have a great deal of power to establish it as an unwritten rule, a new social norm. I imagine spending all day every day of interviews next cycle parked out front of YC. I bet we could come away from there with a dozen new companies signed up to give on Gratipay—if it's easy enough to do so. I didn't even try today ("Now sign in with your company's Twitter account, now verify your email, now put in your credit card, now find someone to give to ..."). If YC were on board, it'd be magnitudes easier.

The one partner I mustered the chutzpah to introduce myself to was Sam Altman. I had seen him wander across to Building 2, and then as he was cruising back to Building 1 (Scott's Domain), I intercepted him:

"Sam!"
"Hi!"
"May I shake your hand? Chad."
[shaking hands]
"You rejected me, and I've been having a wonderful time hanging out in the gutter over there. I'll see you again next year."
[friendly, striding away towards the Scott-guarded door]
"Great! What's your company?"
"Gratipay. Funding open source."
[filing it somewhere?] "I'll keep an eye out."
"Thank you!"
[turns away]
[I chase him down]
"By the way, here's my business card." [pressing heart coin into his hand]
[automatically] "Thank you."

At dusk, with a gorgeous sunset away over downtown Mountain View, Founder Kingsley asked me about the most valuable thing I learned today. "I learned that I love it here. I came here not knowing whether I would dig this scene or not, and I really dig it. There's so many awesome people working on such great stuff!" Here are the ones I ended up scribbling down or otherwise remembering:

1. http://www.cbas.global/
2. https://angel.co/nordigan
3. https://maker.me/
4. http://www.deckard.ai/
5. http://www.clubinternet.co/
6. robot convenience store in Sweden
7. nanny service, same cost as day care but only two families' kids
8. http://vize.io/
9. using GPUs to shorten early drug discovery from two years to one month
10. using GPUs to do something else drug-related (this was Azi's)
11. low-cost internet in India
12. micro-loans in Nigeria

Okay! A stimulating day! ☕️

And I think with that, we can go ahead and close out this ticket. We are now a blip on the YC radar, and based on what I experienced today, I would really love to be more!

When will the application form for the next cycle be online?

Usually at least a few months before the deadline, which is usually in October for winter batches and March for summer batches.

http://www.ycombinator.com/faq/

[chadwhitacre]

I feel like I'm missing something. Like, maybe I should have tried harder to sell (but I don't like selling at someone else's event without their permission—I was on thin ice as it was). Or maybe I should've stayed to meet the other partners (but the last thing they want at the end of a long day is to be accosted by some rando, right?). Oh, well. Hopefully whatever it was I may have left on the table is still there next time. :-/

[chadwhitacre]

Here are the ones I ended up scribbling down or otherwise remembering:

Also:

1. kids learning robots from Australia (the first founders I talked to!)
2. crew from Canada (Spiffy?) helping liquor brands sell more liquor

[ehmatthes]

Thanks for the writeup, that's a more positive day than I would have anticipated. They really can't grant you an interview after not inviting you. With the volume of applications they receive, that would be a terrible precedent to set: "If you get rejected just show up in person, and maybe if you're forward enough we'll grant you an interview on the spot!" If they had interviewed you and accepted you everyone would know it, and they'd be swamped on the next interview cycle.

That said, the idea of hanging out with a bunch of people who did get invited for an interview is a really interesting idea. It sounds like you got a lot out of that, so good move overall! I'll be curious to read about the rest of the trip.

[chadwhitacre]

There were definitely a couple bro-y moments. This German crew was really keen to show me the soft-core that one of them shot in his spare time ("That's his girlfriend! That's his girlfriend!"). Another had a backronym featuring the word "cock." I wish I did a better job of pushing back. :-/

[chadwhitacre]

... though I think UK slang may differ sufficiently? The first situation was definitely slimey.