[00:05.480 --> 00:06.100]  Thank you.
[00:06.220 --> 00:07.980]  And please, it's a freeish country.
[00:08.160 --> 00:09.720]  Turn the microphones on, please.
[00:10.800 --> 00:11.860]  It's a freeish country.
[00:12.120 --> 00:13.120]  Hey, is Strick here?
[00:13.380 --> 00:15.820]  Who else here has cooked with us ever at a food hacking dinner?
[00:15.980 --> 00:16.740]  Raise your hand.
[00:17.080 --> 00:17.240]  Anybody?
[00:18.460 --> 00:19.760]  All right, Strick, get up here.
[00:19.900 --> 00:20.340]  Whoever else.
[00:20.420 --> 00:20.740]  Who's that?
[00:21.480 --> 00:22.460]  Hey, Tim, get up here.
[00:22.500 --> 00:23.520]  Will you help us pass out dishes?
[00:23.780 --> 00:26.000]  So we're going to do... this is how the talk's going to go.
[00:26.160 --> 00:29.180]  I'm going to talk a little bit about food hacking history a little bit.
[00:29.280 --> 00:31.740]  And we're going to drop our O-Day, which is this food.
[00:31.920 --> 00:33.960]  So we're going to feed as many people as we possibly can.
[00:33.960 --> 00:35.120]  I think we have enough food for everybody.
[00:35.360 --> 00:36.760]  We made some really weird food for you.
[00:36.900 --> 00:42.360]  And then we're going to demonstrate our software platform, which is called the Food Genome, which is a culinary informatics platform.
[00:42.620 --> 00:43.760]  So first of all, how's everybody doing?
[00:43.840 --> 00:44.440]  You having a good hope?
[00:46.060 --> 00:46.460]  Beautiful.
[00:46.900 --> 00:47.200]  Beautiful.
[00:47.800 --> 00:50.940]  So how many people here have seen Jason Scott's BBS documentary?
[00:51.860 --> 00:52.260]  Right?
[00:52.480 --> 00:53.080]  And get Lamp.
[00:53.180 --> 00:56.220]  And how many here know that he actually went out for the live action character of Alf?
[00:57.020 --> 00:57.420]  Anybody?
[00:57.800 --> 01:06.920]  So the interesting part about Jason Scott's BBS documentary, this is a little aside about food hacking history, is it's told, like in a studs turkle, it's told by someone who's part of the scene, you know?
[01:07.040 --> 01:08.960]  It's not the one true perspective.
[01:08.960 --> 01:10.200]  It's many people's voices.
[01:10.280 --> 01:11.160]  And I like that a lot.
[01:11.240 --> 01:12.880]  I think 2600 has a similar quality.
[01:13.080 --> 01:22.560]  And the thing that kind of freaks me out a little bit is, you know, Hollywood telling the story of hackers and Hollywood saying that, you know, we all look like Angelina Jolie or Keanu Reeves or whatever.
[01:22.740 --> 01:27.140]  They kind of had that influence in the 90s and that kind of made hackers start to look like that.
[01:27.140 --> 01:33.300]  And I think that's an interesting thing because we have this collective story, but we also all have our own individual stories.
[01:33.920 --> 01:35.740]  You know, this is me when I was a kid at my...
[01:35.740 --> 01:37.020]  You guys can't even see that, can you?
[01:37.620 --> 01:37.940]  Beautiful.
[01:38.080 --> 01:39.480]  Let me see if this... Does that not work?
[01:39.580 --> 01:39.780]  All right.
[01:40.480 --> 01:42.500]  So it's actually sloppy on the screen.
[01:42.780 --> 01:44.200]  Anyway, we all have our own stories.
[01:44.520 --> 01:50.000]  We basically, you know, we can get co-opted by the media, we can get co-opted by Hollywood, but we all have our own stories.
[01:50.400 --> 01:53.940]  And so I like to think of it as a white hat versus a black hat.
[01:54.020 --> 01:55.740]  How many people here identify as black hat?
[01:55.740 --> 01:56.820]  Can I just see you raise a hand?
[01:57.320 --> 01:58.000]  Yeah, a little applause.
[01:58.160 --> 01:59.440]  Two black hats in the audience.
[01:59.940 --> 02:04.460]  Well, I submit to you that there is a thing such as white hat food hackers, in my opinion.
[02:04.800 --> 02:06.860]  They make books for people like O'Reilly.
[02:07.120 --> 02:08.700]  It's about cooking with Asperger's.
[02:08.840 --> 02:11.380]  And it's all about science and food, things like that.
[02:11.480 --> 02:12.560]  You know, it's all Alton Brown.
[02:12.800 --> 02:15.060]  And there's a little Adrian Lamo shout out at the top.
[02:15.400 --> 02:17.760]  Suck the shit out of my ass, you twitching narc f*ck.
[02:18.700 --> 02:19.860]  So that's pretty good.
[02:20.180 --> 02:27.180]  So anyway, people, you know, as we get co-opted and, you know, food hacking, things like that, started to get caught by the mainstream, they're trying to tell our stories.
[02:27.300 --> 02:28.380]  And so I'd like to tell our story.
[02:28.800 --> 02:30.300]  So I just want a little quick slide.
[02:30.460 --> 02:32.500]  This is the true history of black hat food hacking.
[02:32.720 --> 02:32.920]  All right?
[02:33.060 --> 02:34.340]  And I don't know if you can see Watermelon Girl.
[02:34.500 --> 02:35.320]  She's pretty happy, huh?
[02:35.980 --> 02:36.500]  All right, cool.
[02:36.600 --> 02:39.900]  So how is food hacking like regular hacking, right?
[02:40.000 --> 02:42.620]  Well, personally, I came up in the hacker scene and I started getting into food.
[02:42.720 --> 02:46.220]  And then I started making the food scene look as much like the hacking scene as possible.
[02:46.220 --> 02:46.560]  You know?
[02:46.700 --> 02:48.160]  So we have our BBSs, right?
[02:48.400 --> 02:49.620]  We have the Kitchen Hack Lab.
[02:49.620 --> 02:51.840]  It's a Bucky BBS for recipes and menu development.
[02:52.040 --> 02:53.100]  We have a food hacking wiki.
[02:53.100 --> 02:55.020]  It's a chronicling of all of our food hacking tours.
[02:55.200 --> 02:58.220]  And we have some secret software development BBSs and stuff like that.
[02:58.300 --> 03:00.480]  So I just want to show you just a couple little slides from it.
[03:00.680 --> 03:01.140]  Here's a bunch of...
[03:01.140 --> 03:08.080]  You know, we do our recipe kind of prototyping and software through this BBS software that we've actually kind of created.
[03:08.400 --> 03:09.140]  And it really...
[03:09.140 --> 03:12.160]  We give out accounts to other people that are doing interesting food things.
[03:12.240 --> 03:13.220]  And it's kind of that community.
[03:13.220 --> 03:14.220]  You know, it is a web BBS.
[03:14.320 --> 03:15.200]  No one's really dialing up.
[03:15.320 --> 03:21.180]  But we have this kind of community of people that are doing BBSs and are interested in sharing their databases.
[03:21.540 --> 03:21.940]  You know, people...
[03:21.940 --> 03:23.100]  We trade the wares, right?
[03:23.200 --> 03:27.700]  We have, you know, USDA databases, screen scrape databases, interesting software, things like that.
[03:27.800 --> 03:29.160]  And so that's been one way.
[03:29.520 --> 03:32.140]  We also have hacker kitchens, right?
[03:32.260 --> 03:35.800]  And so as food hacking, for the past five years, we've been going on tour.
[03:36.080 --> 03:37.280]  People have been building us kitchens.
[03:37.600 --> 03:47.180]  The one on the upper right is a kitchen that was created for us in Zagreb, Croatia, by Container, this art collective that flew us out there and built a kitchen for us in the middle of their art space.
[03:47.220 --> 03:50.000]  We kind of did the same thing at Seabase, actually, during the CCC Congress.
[03:50.220 --> 03:52.620]  There's a picture of our mother kitchen there, the Unicorn Precinct.
[03:52.800 --> 03:59.980]  And as we've gone out on tours to kind of cook with other people in hacker kitchens, we kind of figured out what makes a hacker kitchen.
[04:00.120 --> 04:05.820]  And it's cookbooks, it's equipment, it's, you know, soldering irons to be able to fix stuff, make interesting different equipment.
[04:06.040 --> 04:12.700]  And it's also this collaborative experiment, experimenting thing, where there are people coming into this kitchen all the time.
[04:12.800 --> 04:16.260]  The bottom left one is Lab 24-7 in Bed-Stuy.
[04:17.080 --> 04:18.500]  Take some food, by the way, passing out.
[04:18.560 --> 04:19.340]  There is meat in it.
[04:19.780 --> 04:21.020]  Does anyone here like meat?
[04:21.180 --> 04:21.520]  Raise your hand.
[04:21.580 --> 04:21.940]  You like meat?
[04:22.060 --> 04:22.580]  All right, meat.
[04:23.400 --> 04:24.660]  So be careful of that.
[04:25.080 --> 04:26.540]  So we have a lot of food hacker kitchens.
[04:26.660 --> 04:30.900]  And the trick about food hacker kitchens is that people have to go there all the time to eat.
[04:30.900 --> 04:34.780]  and have this constant creating process.
[04:35.180 --> 04:36.000]  So what else do we got?
[04:36.120 --> 04:37.540]  We have food hacking is illegal, right?
[04:37.900 --> 04:40.040]  So black hat hacking sometimes is illegal.
[04:40.240 --> 04:41.540]  And maybe there's black hat food hacking.
[04:41.660 --> 04:43.760]  So we have here on the upper left...
[04:43.760 --> 04:45.260]  Man, are these slides bright enough?
[04:45.380 --> 04:46.000]  What do you guys think?
[04:46.280 --> 04:46.540]  No?
[04:46.980 --> 04:47.780]  Can you see anything?
[04:48.080 --> 04:50.780]  Can you turn the lights down by any chance that are aimed at these guys?
[04:52.860 --> 04:53.220]  Anyway...
[04:53.220 --> 04:56.220]  Yeah, because I just want you guys to see this...
[04:58.840 --> 04:59.540]  What do you think?
[04:59.540 --> 05:00.320]  We're working on...
[05:00.760 --> 05:01.720]  Is that any better?
[05:01.980 --> 05:02.840]  Can you see...
[05:03.200 --> 05:03.840]  There we go.
[05:04.920 --> 05:05.240]  That's...
[05:05.240 --> 05:05.540]  Beautiful.
[05:05.760 --> 05:05.980]  All right.
[05:06.280 --> 05:06.960]  So we have...
[05:06.960 --> 05:08.240]  We've done a lot of insect dishes.
[05:08.460 --> 05:09.540]  We've really been...
[05:09.540 --> 05:17.880]  The thing that's kind of inspired us about food hacking is you see this get co-opted into like, you know, TV science, food, and stuff like that, is how do we go in a direction that can't be done on TV, that can't be done legally?
[05:18.060 --> 05:21.440]  You know, we do dinner parties where people are all in the kitchen kind of cooking together.
[05:21.440 --> 05:25.120]  You really can't do that as a restaurant or as a food cart or anything like that.
[05:25.240 --> 05:26.860]  We've also done a lot of drug-related cooking.
[05:26.980 --> 05:27.860]  Does anyone here cook with drugs?
[05:28.000 --> 05:28.400]  Raise your hand.
[05:28.480 --> 05:28.640]  Anybody?
[05:29.600 --> 05:44.200]  So, this up on the upper left is different isohash trials where we basically soaked like mugwort and wormwood and tonkamine, different psychoactive things in different kinds of alcohol, and then kind of boiled them down into tar hashes and made a vaporizer course as part of our meals.
[05:44.820 --> 05:48.280]  We've also used parts of the body for a thing.
[05:48.340 --> 05:49.820]  This is the toe cheese on the bottom left.
[05:49.900 --> 05:50.980]  I don't know if you guys can see that.
[05:51.220 --> 05:51.820]  This is a...
[05:51.820 --> 05:53.660]  My toenail got ripped off.
[05:54.120 --> 05:54.840]  I dropped my...
[05:54.840 --> 05:56.740]  Anyone dropped a laptop on their toe before?
[05:56.960 --> 05:57.440]  It hurts.
[05:58.000 --> 06:00.760]  And so we Dremeled that down to make toe cheese out of it.
[06:01.120 --> 06:04.560]  So it kind of flavored the milk and you can make the cheese on the rind of the toe.
[06:04.820 --> 06:05.960]  It's a personalized thing.
[06:06.120 --> 06:08.840]  Anyway, these are things you can't have in restaurants, right?
[06:09.060 --> 06:12.980]  And so that really kind of drives me a little bit too, is feeding insects to people, feeding interesting things.
[06:13.620 --> 06:15.460]  So open source is the nuclear option.
[06:15.660 --> 06:17.660]  This may seem like this is a random slide to put here.
[06:18.160 --> 06:19.540]  Well, we've been working on all this software.
[06:19.980 --> 06:22.920]  This is a thing called the food genome, this big culinary brain I'm going to tell you about.
[06:23.220 --> 06:26.480]  And we've been kind of trying to build this big open source food taxonomy.
[06:26.660 --> 06:33.340]  But now all these .com, Web 3.0, 2.0 companies are trying to, you know, monetize food, monetize food on the Internet.
[06:33.720 --> 06:35.180]  Google just bought MetaWeb.
[06:35.280 --> 06:35.860]  Did anyone hear that?
[06:36.080 --> 06:36.440]  Freebase?
[06:36.780 --> 06:38.340]  Anyone know what Freebase is?
[06:38.340 --> 06:46.320]  You know, so there are other food taxonomies out there and Freebase was just acquired by Google, which means this huge open food taxonomy is, you know, going to be owned towards a certain direction.
[06:46.540 --> 06:58.160]  So that's why I'm going to talk about Food Genome in a second and talk about how open source is the nuclear option about our software that we've been working on to make interesting, you know, food applications and culinary informatics platforms.
[06:58.300 --> 07:00.820]  We actually are just now open sourcing it so that everybody can do that.
[07:00.920 --> 07:03.720]  And it's not just our private little tools in the lab.
[07:03.860 --> 07:05.260]  And I just wanted to do a side...
[07:05.260 --> 07:06.820]  Who here voted for Ralph Nader ever?
[07:06.820 --> 07:07.460]  Raise your hand.
[07:07.700 --> 07:08.080]  Ralph Nader?
[07:08.220 --> 07:08.800]  Oh, that's some people.
[07:09.040 --> 07:13.080]  Ralph Nader is 70 years old, 74, and he knows nothing about computers.
[07:13.320 --> 07:15.600]  He uses notebooks and manila folders.
[07:15.840 --> 07:18.420]  And talking about open source, he said, are you for or against open source?
[07:18.520 --> 07:21.480]  I said, well, people seem to co-opt open source and use it to sell things.
[07:21.680 --> 07:28.440]  His idea, which I want to repeat to hear you now, is it should be regulated by the government in a way where if you... it's like organic, right?
[07:28.820 --> 07:35.400]  And if you go against open source, if you actually don't release a source code, you should be, you know, persecuted, you know, prosecuted basically for false advertising.
[07:35.400 --> 07:38.680]  Anyway, I think that's an interesting idea that someone who doesn't know how to use computers does it.
[07:38.800 --> 07:42.900]  So we are going to open source this open food taxonomy, which is the food genome.
[07:43.120 --> 07:45.900]  So now we're at the part of the talk where I'm going to talk about the food genome.
[07:46.740 --> 07:48.160]  Has anyone had any food so far?
[07:48.840 --> 07:49.460]  What do you think?
[07:49.800 --> 07:50.260]  What is it?
[07:50.260 --> 07:51.100]  It's weird.
[07:51.180 --> 07:51.280]  Okay.
[07:51.320 --> 07:53.660]  We'll talk about it a little bit, actually, before we jump into the food genome.
[07:54.240 --> 07:55.460]  So it's...
[07:55.460 --> 07:56.740]  Who here likes fruit roll-ups?
[07:56.960 --> 07:57.680]  Fruit roll-ups?
[07:57.860 --> 07:58.500]  Delicious, right?
[07:58.680 --> 08:00.200]  Very easy to make your own fruit leather.
[08:00.360 --> 08:04.200]  So it's a fruit leather cannoli that we made out of roasted pineapple and star anise.
[08:04.320 --> 08:08.180]  And then we also took some fennel root and mixed that up with cayenne and salt and pepper.
[08:08.380 --> 08:11.380]  And so the fruit leather is half pineapple and half spicy fennel.
[08:11.380 --> 08:13.500]  And then that's wrapped around this fennel sausage.
[08:13.800 --> 08:17.320]  And then that has star anise and other different spices, pastry cream in it.
[08:17.380 --> 08:20.240]  So it's similar to a cannoli, except with a fruit pastry around the outside.
[08:20.460 --> 08:23.640]  And one of the projects we did with the food genome was the cookbook Iditarod.
[08:23.760 --> 08:25.580]  We're trying to read all these cookbooks and cook...
[08:25.580 --> 08:27.220]  I have like 400 cookbooks in my kitchen.
[08:27.500 --> 08:34.100]  And so we're trying to cook through as many cookbooks as possible and review them and, like, distill that information and feed it into this food genome brain.
[08:34.520 --> 08:36.700]  And so we're reading a lot of Indonesian cookbooks.
[08:36.780 --> 08:40.400]  And Indonesians have these things called a sambal, where they kind of mix fruit and chili and spiciness.
[08:40.400 --> 08:41.740]  And so that's kind of where it's inspired.
[08:42.000 --> 08:44.360]  So these food hacking volunteers over here...
[08:44.360 --> 08:48.040]  Just give them a round of applause for making food in front of hundreds of people.
[08:50.020 --> 08:52.500]  So they're feeding you and everyone's helping out, hopefully.
[08:52.780 --> 08:54.680]  So let's talk about the food genome a little bit.
[08:54.760 --> 08:59.500]  So the food genome is a giant hungry brain that scrapes the Internet and learns about food.
[08:59.720 --> 09:01.740]  And it basically quantizes food.
[09:01.900 --> 09:02.340]  It has this idea.
[09:02.500 --> 09:05.240]  It's based on this surrealist food game we built called Delicious Corpse.
[09:05.420 --> 09:07.880]  And Delicious Corpse has kind of generated...
[09:07.880 --> 09:09.040]  Thank you for keeping that up.
[09:09.040 --> 09:10.120]  Yeah, no, thank you.
[09:10.380 --> 09:10.980]  So it's...
[09:10.980 --> 09:12.360]  Anyone here like Surrealism?
[09:12.480 --> 09:12.760]  Raise your hands.
[09:12.960 --> 09:13.280]  Surrealism?
[09:13.440 --> 09:13.540]  Yeah?
[09:13.680 --> 09:13.900]  Positive?
[09:14.200 --> 09:15.100]  So Surrealists had this game.
[09:15.200 --> 09:16.600]  They did a lot of weird food games actually.
[09:17.900 --> 09:20.400]  Leonora Carrington used to chop off people's hair in the middle of the night.
[09:20.560 --> 09:23.440]  And in the morning make them breakfast with an omelet with their own hair in it.
[09:24.400 --> 09:25.100]  That's pretty cute.
[09:25.300 --> 09:25.560]  I don't know.
[09:25.640 --> 09:26.560]  For like 1930s?
[09:26.660 --> 09:27.760]  That's pretty avant-garde.
[09:27.940 --> 09:29.980]  I'm sure people were doing that back in the day for other reasons.
[09:30.140 --> 09:31.620]  But like voodoo and stuff.
[09:31.620 --> 09:34.100]  So we took these surrealist food games.
[09:34.320 --> 09:38.640]  There's this game called Exquisite Corpse where you chop up words and you make sentences out of it basically.
[09:38.780 --> 09:39.800]  And everyone contributes words.
[09:40.000 --> 09:41.300]  So we kind of were doing that with food.
[09:41.460 --> 09:42.440]  So you'd make up...
[09:42.440 --> 09:47.140]  You'd come up with things like chicken wing and apple candy, delicata squash and Jolly Rant...
[09:47.640 --> 09:48.420]  What is it?
[09:48.480 --> 09:50.860]  Blueberry baguette and hydrogenated chicken breast milk.
[09:50.860 --> 09:53.280]  And it did some interesting things because it actually...
[09:53.280 --> 09:53.720]  Yeah.
[09:53.840 --> 09:54.520]  Chicken breast milk.
[09:54.720 --> 09:55.300]  Does that sound good?
[09:55.740 --> 09:56.600]  Raw chicken breast?
[09:56.860 --> 09:57.480]  Blended with milk?
[09:57.720 --> 09:58.040]  Positive?
[09:58.400 --> 09:58.500]  Yeah.
[09:59.520 --> 09:59.740]  Yeah.
[10:00.440 --> 10:01.340]  Chickens don't make milk.
[10:01.460 --> 10:02.120]  But pigeons do.
[10:02.300 --> 10:02.900]  Anyone know that?
[10:03.020 --> 10:03.380]  Pigeons?
[10:03.640 --> 10:04.380]  They do make milk.
[10:04.480 --> 10:05.080]  They have this milk.
[10:05.200 --> 10:05.980]  It's called cropped milk.
[10:06.000 --> 10:07.060]  And it's at the back of their neck.
[10:07.160 --> 10:08.620]  And they feed their young with it.
[10:08.720 --> 10:09.260]  It's kind of gross.
[10:10.640 --> 10:11.040]  So...
[10:11.480 --> 10:12.500]  I haven't tried it.
[10:13.000 --> 10:13.400]  So...
[10:13.400 --> 10:14.080]  But I...
[10:14.080 --> 10:14.160]  You know.
[10:14.860 --> 10:15.260]  Someday.
[10:15.380 --> 10:16.600]  You make a pigeon milkshake.
[10:16.820 --> 10:18.500]  So we use this to kind of generate.
[10:18.500 --> 10:23.340]  And the interesting thing about a surrealist food game is it breaks down your conceptions about what's possible.
[10:23.380 --> 10:25.160]  And it's a reflection of the collective stomach.
[10:25.360 --> 10:26.800]  Everyone writes down a piece of food.
[10:27.000 --> 10:27.560]  Try this sometime.
[10:27.680 --> 10:28.720]  Write down a piece of food on paper.
[10:28.900 --> 10:32.180]  And then everyone puts it together and you take four of them at random and then you make that thing.
[10:32.340 --> 10:35.920]  And it's super fun and it breaks down your own boundaries and kind of, you know, encourages this idea.
[10:36.160 --> 10:39.900]  So the other inspiration we have from the food genome is Mayor Gavin Newsom from San Francisco.
[10:40.180 --> 10:40.360]  Anyone?
[10:40.740 --> 10:41.100]  Alcoholic?
[10:41.480 --> 10:41.840]  Alcoholic?
[10:42.040 --> 10:42.440]  Serving drinks?
[10:42.560 --> 10:42.760]  Great.
[10:43.200 --> 10:44.960]  He had a big alcohol scandal.
[10:44.960 --> 10:51.860]  And also Ferran Adria from El Bulli had this interesting weird taxonomy of how he would organize food.
[10:51.860 --> 10:54.200]  And he had these really strange stylized icons.
[10:54.520 --> 11:02.160]  And then in the upper right corner we had this kind of cooking for bachelors book that was up in... from the 1950s where every step didn't really have words.
[11:02.280 --> 11:03.320]  It would have one or two words.
[11:03.460 --> 11:04.440]  And then they would show...
[11:04.440 --> 11:09.260]  You can see here three cups of vermouth, one cup of cognac, three cups of Chablis.
[11:09.580 --> 11:10.220]  So you would actually...
[11:10.220 --> 11:11.320]  Or shots of those actually.
[11:11.320 --> 11:13.060]  And so these are very visual ideas.
[11:13.260 --> 11:20.940]  And so as we started with the food genome, this giant... you know, figure out how to make a food program that would generate menus like this, but actually have them be delicious.
[11:21.240 --> 11:22.980]  And at least have some sort of a feedback engine.
[11:23.040 --> 11:25.620]  So you could go in and you could say thumbs up, thumbs down.
[11:25.820 --> 11:27.520]  And then the side effect of that...
[11:27.520 --> 11:28.800]  I'm just going to jump to this one.
[11:28.940 --> 11:30.760]  Is that it would actually build your flavor profile.
[11:30.860 --> 11:34.180]  So the food genome is this engine that you talk to and tell it what you're eating.
[11:34.300 --> 11:36.040]  You can either have it as a food diary or tell it what you like.
[11:36.120 --> 11:39.600]  You could say yum or yuck about things and it learns what you like to eat.
[11:39.600 --> 11:41.120]  It builds a profile about it.
[11:41.260 --> 11:42.540]  It can learn what your kids like to eat.
[11:42.620 --> 11:46.500]  You could say what I like to eat in, you know, last month or last year.
[11:46.860 --> 11:47.240]  Things like that.
[11:47.340 --> 11:57.680]  And because we actually have an open API where you can query an engine and actually do really interesting kind of geolocative taste gestalt things where you see what is everyone eating in Chicago.
[11:58.100 --> 12:04.840]  I have a lot of old restaurant cook menu books from the past 30 or 40 years from like San Francisco and New Orleans and Chicago.
[12:05.040 --> 12:14.980]  So I think it's really interesting personally to map the taste gestalts of those eras and see how the cuisine in Chicago has changed or see how the cuisine in New Orleans has changed with the oil spill.
[12:15.220 --> 12:18.540]  So I'm just going to talk a little bit about the food genome platform right now.
[12:18.540 --> 12:20.400]  So this is kind of an example...
[12:20.400 --> 12:22.040]  Or this is kind of a schematic of the diagram.
[12:22.200 --> 12:26.740]  And this is like an XML scraped menu right here where we take menus from restaurants.
[12:27.020 --> 12:28.960]  They put them up in PDFs and doc formats.
[12:29.140 --> 12:31.820]  And we scrape it and we tokenize it and we feed it into the genome.
[12:32.080 --> 12:36.340]  So the genome itself is this big database that knows about all these kind of objects and things like that.
[12:36.480 --> 12:44.340]  It has a scraping API that you feed physics data, nutrition data, price data into, and also food modelers that you can build new kinds of food.
[12:44.340 --> 12:45.980]  Like you can model shit, right?
[12:46.100 --> 12:51.740]  You can say the color is brown, and it has a stinky flavor to it, and it's kind of liquid, and it's kind of solid.
[12:51.960 --> 12:59.300]  And so the physics engine on the base of it knows about things like liquid and solid, and you can't chop something if it's liquid unless you freeze it, right?
[12:59.400 --> 13:00.080]  That kind of a thing.
[13:00.220 --> 13:06.840]  And so then the API to it is available for applications, for recipe tweakers, and search engines.
[13:06.960 --> 13:09.460]  So you can basically search through the database and ask it questions.
[13:09.600 --> 13:12.160]  You can ask the genome, what red food did they eat in China?
[13:12.160 --> 13:17.340]  Or what would Julia Child make for me if I was in Thailand and I have a nut allergy and I'm vegan?
[13:17.480 --> 13:19.760]  Just these really complicated kinds of things, right?
[13:19.840 --> 13:23.580]  And so there's all these food 3.0 kind of customized food sites coming down.
[13:23.780 --> 13:34.720]  And I think this is the secret sauce that a lot of those are missing, I guess, because the way we're able to map this using semantic web technology, RDF technology, is a narrow enough domain.
[13:34.720 --> 13:37.100]  There's only about 60 objects you got to know about, right?
[13:37.260 --> 13:43.040]  Like a type of a food and a geo region and a granularity and actions and stuff like that.
[13:43.140 --> 13:44.460]  So I'm going to dive a little bit deeper into that.
[13:44.580 --> 13:50.600]  But you can see here, when we scrape this menu, this is a fresh mint sprigs, right?
[13:50.700 --> 13:51.360]  That's in a cocktail.
[13:51.640 --> 13:54.360]  And it's three fresh mint sprigs and it got parsed to the quantity of three.
[13:54.780 --> 13:57.860]  Sprigs is the unit, mint is the product, and fresh is the quality.
[13:57.860 --> 13:59.620]  And all those things mean things.
[13:59.780 --> 14:04.900]  And because it's semantic web, each of those tokens that it's parsed out has a URL, which you can link into other things.
[14:05.020 --> 14:06.740]  And you can say, well, show me all fresh food.
[14:06.880 --> 14:08.660]  Or show me everything that has three of anything.
[14:08.880 --> 14:11.300]  Or show me anything that's measured as a sprig, right?
[14:11.820 --> 14:13.840]  So I think that's really the...
[14:13.840 --> 14:15.180]  And of course, there's regions too.
[14:15.340 --> 14:18.180]  There's like Cuban cocktails and dishes and stuff like that.
[14:18.280 --> 14:21.540]  So we're going to dive a little bit into some of the use cases for the food genome.
[14:21.640 --> 14:24.240]  The first thing we kind of built was what should I eat for dinner, right?
[14:24.360 --> 14:26.760]  And this is kind of the beginning of the visualization.
[14:26.760 --> 14:29.460]  Because the interesting thing is that we have the colors of these foods.
[14:29.600 --> 14:31.060]  I think that might even be the next...
[14:31.060 --> 14:31.220]  No.
[14:31.320 --> 14:32.040]  It's a couple slides down.
[14:32.680 --> 14:34.260]  We actually have these colors of these foods.
[14:34.380 --> 14:36.960]  So this mint pesto pasta for some reason is super pink.
[14:37.060 --> 14:37.580]  Why is that?
[14:37.760 --> 14:38.920]  Oh, because it has tomatoes in it.
[14:39.560 --> 14:41.760]  This one is pink because it has shrimp and tomatoes.
[14:41.880 --> 14:47.780]  And so we're able to kind of look at food, at recipes at a glance and say, this doesn't look very healthy because it's very brown.
[14:47.860 --> 14:49.260]  But this is very red and green.
[14:49.380 --> 14:50.460]  That actually does look very healthy.
[14:50.580 --> 14:52.760]  And it's especially helpful for when you have your own...
[14:53.420 --> 14:54.040]  Like if you've...
[14:54.040 --> 14:55.960]  Like this is my diet over a month, right?
[14:55.960 --> 14:59.120]  And so I eat a lot of vegetables and maybe, you know, a lot of meat.
[14:59.220 --> 14:59.900]  Like the pink colors.
[15:00.100 --> 15:04.020]  And so you can kind of look at that at a glance, like a mood ring and say, are you eating healthily or not?
[15:04.660 --> 15:05.260]  What time is it?
[15:05.320 --> 15:05.660]  4.04.
[15:05.820 --> 15:06.640]  How's this going so far?
[15:06.720 --> 15:07.540]  Is this all right, you guys?
[15:07.680 --> 15:08.200]  Is it a little fast?
[15:08.320 --> 15:08.620]  All right, good.
[15:09.020 --> 15:10.320]  Because we've got a lot more.
[15:10.780 --> 15:18.540]  The next part of the talk, after we kind of talk about the basics of how the food genome data scraping works, is going to be the visualization stuff that we've been working on for a while.
[15:18.660 --> 15:20.740]  And that's the stuff that I think is super impressive.
[15:21.040 --> 15:23.100]  So this is a what should I eat for dinner where you just...
[15:23.100 --> 15:23.740]  You ask your...
[15:23.740 --> 15:25.180]  You know, everyone wants this as a killer app.
[15:25.240 --> 15:27.500]  You ask the Internet, what the f*ck do I want, right?
[15:27.500 --> 15:28.260]  I want tomatoes.
[15:28.540 --> 15:31.060]  And most recipe sites will be like, here's some tomato recipes.
[15:31.300 --> 15:34.200]  But if you have the history, you can be like, oh, well, you had Chinese food yesterday.
[15:34.340 --> 15:36.100]  You had Italian food two days ago.
[15:36.260 --> 15:38.280]  So we want to take you in a totally different direction.
[15:38.280 --> 15:43.680]  And of course, another important thing about knowing about the yuck is trying to get you to like foods that you hate, right?
[15:43.860 --> 15:45.820]  So we know that Dylan hates mushrooms, right?
[15:46.280 --> 15:48.600]  But we also know that he loves Choco Tacos.
[15:48.860 --> 15:51.420]  So have you tried Choco Tacos with mushrooms in them?
[15:52.780 --> 15:53.580]  Sounds kind of gross.
[15:53.880 --> 15:58.000]  Let's say he likes Hawaiian food, but he hates green pepper.
[15:58.220 --> 16:01.480]  Well, has he tried Hawaiian skewers with pork and pineapple and green pepper on it?
[16:01.540 --> 16:07.860]  It's a way to get into, you know, to try and have this little step that helps you like the food that you f*cking hate, right?
[16:07.860 --> 16:10.420]  And I think that's especially good for kids and stuff like that.
[16:10.480 --> 16:11.160]  Who here has kids?
[16:11.420 --> 16:11.660]  Anybody?
[16:11.940 --> 16:12.140]  Kids?
[16:12.340 --> 16:12.720]  All right, good.
[16:12.960 --> 16:13.780]  Everyone has kids these days.
[16:13.880 --> 16:14.080]  It's great.
[16:14.960 --> 16:16.540]  So then we have allergies as well.
[16:16.660 --> 16:20.180]  You're not only learning what you like or what you don't like, but also what you're allergic to.
[16:20.260 --> 16:23.880]  You can say, I don't care for green peppers so much, but peanuts will kill me.
[16:23.980 --> 16:24.740]  And because we have this...
[16:24.740 --> 16:25.080]  Yeah, what's up?
[16:29.200 --> 16:30.760]  Oh, you don't have to pass that one plate for person.
[16:30.860 --> 16:32.100]  Just keep holding the plate again.
[16:32.780 --> 16:33.940]  You're not passing that one plate for person.
[16:35.340 --> 16:37.280]  Anyway, let's talk to this.
[16:37.280 --> 16:40.300]  Yeah, leave the empty plate at the end is what Veggie has to say.
[16:41.520 --> 16:43.600]  Give it up for the death vegetable, ladies and gentlemen.
[16:43.940 --> 16:44.260]  All right.
[16:44.380 --> 16:45.200]  Passing food out.
[16:45.360 --> 16:45.600]  Beautiful.
[16:47.560 --> 16:48.280]  Death vegetable...
[16:49.340 --> 16:52.660]  Death vegetable also knows how to hold a cleaver.
[16:54.720 --> 16:55.240]  All right.
[16:55.680 --> 16:56.260]  That's scary.
[16:56.420 --> 16:56.520]  Okay.
[16:56.680 --> 16:58.960]  So we talk about allergies a little bit.
[16:59.000 --> 17:04.280]  And because we have this international database, right, that knows about the taste-gestals around the world, we have a big language lookup table and stuff like that.
[17:04.420 --> 17:14.920]  This is an app that our friend mocked up for the iPhone, which is Peanuts Will Kill Me, which is whatever country you're in, right, you want the language to say Peanuts Will Kill Me in whatever language, you know, the country is.
[17:14.980 --> 17:15.680]  You don't want to have to...
[17:15.680 --> 17:18.600]  Does anyone have to go to another country and explain how vegan you are, right?
[17:18.720 --> 17:20.220]  And you go to Germany and they're like...
[17:22.280 --> 17:22.600]  So...
[17:23.040 --> 17:25.140]  Or Mexico is another example.
[17:25.560 --> 17:25.780]  All right.
[17:25.900 --> 17:35.480]  So as we started, you know, scraping all this data, we basically scraped hundreds of thousands of recipes off the Internet and started modeling them and looking for similar things that were among them.
[17:35.540 --> 17:37.560]  And so then we started talking about how to organize the foods.
[17:37.680 --> 17:40.660]  There's this whole periodic table of condiments ideas where how do you...
[17:40.660 --> 17:43.440]  What are all the different ways that you can group different kinds of foods?
[17:43.820 --> 17:45.900]  So here's how we actually did scrape this recipe.
[17:45.980 --> 17:47.100]  I'll just go through the thing really quick.
[17:47.280 --> 17:50.700]  So these fields right here are the recipes, tags, ingredients, and directions.
[17:50.700 --> 17:51.440]  That's basically...
[17:51.440 --> 17:52.440]  Every recipe has that.
[17:52.860 --> 17:53.260]  And you can...
[17:53.260 --> 17:56.560]  You write a personalized scraper for each recipe website that you're grabbing.
[17:56.600 --> 18:00.000]  100,000 recipes are 30,000, 40,000 per website usually for the big ones.
[18:00.160 --> 18:03.320]  And so you grab the recipe, you know about the ingredients and the directions and the tags.
[18:03.500 --> 18:08.080]  And then we're actually able to scrape them using these components, buckets, and statements into making...
[18:08.080 --> 18:10.400]  This is our first recipe tweak or prototype we made.
[18:10.500 --> 18:11.460]  It was this French onion soup.
[18:11.700 --> 18:13.740]  And you see that it detected that it was...
[18:13.740 --> 18:16.180]  Well, it knew that it was France and it was onion, and it was a soup.
[18:16.280 --> 18:18.040]  But onion was the main ingredient, French was the region.
[18:18.040 --> 18:28.300]  But if you click on France, which doesn't work anymore, of course, on our software, and then change it to Morocco, it would change it to a Moroccan onion soup, and that would figure out which of these ingredients is French and replace them with Moroccan,
[18:28.300 --> 18:30.440]  so it would add, like, cumin and orange zest, right?
[18:30.620 --> 18:36.260]  Or you change it from French to Italian onion soup, and it would add, you know, hella herbs and, you know, whatever, you know, mark or something like that.
[18:36.580 --> 18:41.260]  And onion, you could click on onion, change it to butternut squash, and make it a butternut squash soup with the same ingredients.
[18:41.300 --> 18:48.020]  Or change it from soup to a souffle, and it would change, kind of using these same ingredients, and be able to use these visualized processes would all change.
[18:48.040 --> 18:49.500]  And you'll try to change to make that different thing.
[18:49.740 --> 18:52.580]  And so the way that we're able to do that... in this prototype level.
[18:52.780 --> 18:56.540]  was by scraping these ingredients and directions... into buckets and having statements.
[18:56.740 --> 19:01.680]  And the buckets and statements... basically work like... you chop garlic and you put it in a pan.
[19:01.840 --> 19:03.340]  The pan becomes the default bucket.
[19:03.660 --> 19:04.640]  And then you add wine.
[19:04.840 --> 19:08.700]  And then the statement... you're parsing the recipe, just the pros of the recipe...
[19:08.700 --> 19:11.520]  and you know that you're adding wine to the default bucket that was last-defined.
[19:11.660 --> 19:12.800]  So it has a null in that field.
[19:12.800 --> 19:16.060]  And then you would say, heat, milk, and then combine the two.
[19:16.220 --> 19:22.660]  And so we kind of know which of those were buckets, what equipment was being used, which each step, which products were being used in each step.
[19:22.760 --> 19:25.520]  And so this is kind of just an example of how that pipeline works.
[19:25.660 --> 19:29.260]  And then we're able to kind of super group the recipes and components.
[19:29.400 --> 19:30.860]  Oh, components are super important.
[19:31.000 --> 19:31.900]  Like wet versus dry.
[19:31.960 --> 19:32.500]  Who here bakes?
[19:32.560 --> 19:33.280]  You ever make a cake?
[19:33.620 --> 19:33.760]  Right?
[19:33.880 --> 19:34.140]  All right.
[19:34.300 --> 19:34.620]  Muffins?
[19:34.760 --> 19:35.380]  Anyone muffins?
[19:36.080 --> 19:36.640]  Muffins, anyone?
[19:38.220 --> 19:39.920]  So you have the wet and the dry components.
[19:39.920 --> 19:43.800]  And so our algorithm is able to detect, well, this is the vegetables.
[19:44.240 --> 19:45.320]  And these are the herbs.
[19:45.620 --> 19:46.640]  This is the wet stuff.
[19:46.920 --> 19:48.080]  These are all used in this step.
[19:48.200 --> 19:50.320]  This is the frosting for a cake, basically.
[19:50.640 --> 19:54.540]  So the also thing we're able to parse is geographic content.
[19:54.740 --> 19:58.400]  And so we basically built a geotaxonomy of food, right?
[19:58.500 --> 19:59.780]  So it's not all countries in Asia.
[19:59.900 --> 20:02.400]  It's just all the countries in Asia that have distinct cuisines.
[20:02.500 --> 20:03.780]  And I'm sure we're totally offending people.
[20:03.920 --> 20:09.460]  And I think that's another idea I want to talk about, actually, is according to Indian food, this is a good pairing.
[20:09.460 --> 20:11.580]  According to Indonesian food, this is a good pairing.
[20:11.740 --> 20:15.260]  But it's all according to the food genome or according to this cookbook.
[20:15.400 --> 20:18.200]  Those are the ideas in semantic web is the idea of contexts.
[20:18.420 --> 20:21.220]  A semantic web statement is a subject predicate object, right?
[20:21.420 --> 20:24.560]  So you'd say tomato is a key ingredient in Italian food.
[20:24.980 --> 20:28.300]  And the context is according to Mario Batali, right?
[20:28.540 --> 20:30.620]  Or maybe it's a canonical statement.
[20:30.880 --> 20:40.720]  And so all the statements in the food genome are either canonical, if they're real things, like celery seed is a part of the plant of celery root, and celery seed is a part of the plant of celery.
[20:40.880 --> 20:42.620]  Those are like kind of canonical fact statements.
[20:42.840 --> 20:49.500]  But then there's also context statements where you say, according to Persians that live in Los Angeles, right, rice is a Persian food.
[20:49.660 --> 20:54.940]  But according to Iranians that live in Iran, you know, that's a Tehrani food, because that's where the rice is.
[20:55.040 --> 20:56.220]  Bread is everywhere else in the country.
[20:56.320 --> 20:58.540]  And so you'll be able to see from these different kinds of perspectives.
[20:59.120 --> 21:01.760]  And then the whole point of this is to, oh, yeah, here's a tokenizer.
[21:01.960 --> 21:05.620]  So here we have a recipe here for Jamie Oliver's minestrone soup, right?
[21:05.720 --> 21:10.160]  And so here it's parsed down this display in a large stock pot over medium low, and it's parsed out all the buckets.
[21:10.400 --> 21:12.040]  It knows that there's a large stock pot.
[21:12.140 --> 21:13.260]  It knows all the ingredients in this step.
[21:13.320 --> 21:14.060]  It knows the actions.
[21:14.400 --> 21:17.880]  It's figured out that you're sauteing, you're chopping, those kinds of things.
[21:18.000 --> 21:30.440]  And it also, it's able... One of the things I don't think I mentioned was the slider technique, which is, I don't think, in this thing where you're able to kind of have a slider on a recipe and tweak a recipe and say, make it vegetarian or make it vegan or speed it up,
[21:30.660 --> 21:30.800]  right?
[21:30.860 --> 21:32.300]  I don't want to have to make roast garlic.
[21:32.480 --> 21:33.880]  I'll just put the chopped garlic in, right?
[21:34.100 --> 21:36.880]  And so just a way to, like, kind of tweak a recipe and make it cost less.
[21:37.540 --> 21:41.480]  Just those are different parameters that you're able to tweak once you kind of quantize this recipe.
[21:41.940 --> 21:44.160]  So here's kind of how our tokenizer works a little bit.
[21:44.240 --> 21:48.940]  So we have things like, you know, two tablespoons of minced garlic, and it says that's a quantity, a unit of preparation.
[21:48.940 --> 21:56.380]  It detects articles like ore, non-alcoholic as a quality, unrefined as a quality, preparation is sliced.
[21:56.580 --> 22:01.140]  And so it basically parses through each of these lines, try and find a unit or quantity first.
[22:01.340 --> 22:03.080]  And if there is no unit or quantity, it's always each.
[22:03.240 --> 22:06.160]  Because you have a recipe that calls for orange, right?
[22:06.340 --> 22:07.060]  So what does that mean?
[22:07.180 --> 22:15.160]  So we have this physics engine on the back end that I'll show you on the next slide that knows what, if a recipe just says orange, it kind of knows what the proportion of the recipe that is.
[22:15.500 --> 22:23.840]  And then it identifies the other types of tokens, interprets grammar stop words, and it has to resolve type conflicts like cloves of garlic, like two cloves of garlic.
[22:24.060 --> 22:27.440]  Is that cloves the food or is that cloves the unit, right?
[22:28.140 --> 22:28.700]  How's that going?
[22:30.140 --> 22:30.460]  Beautiful.
[22:30.880 --> 22:31.180]  All right.
[22:31.300 --> 22:31.700]  Nice job.
[22:31.860 --> 22:32.620]  What do you think of the food?
[22:32.780 --> 22:33.140]  Is it weird?
[22:33.340 --> 22:33.900]  What do you think?
[22:33.940 --> 22:34.240]  Is it good?
[22:34.380 --> 22:34.620]  All right.
[22:34.680 --> 22:34.900]  Beautiful.
[22:35.120 --> 22:35.400]  Thank you.
[22:37.640 --> 22:38.800]  So we have a...
[22:38.800 --> 22:42.180]  So the physics engine on the back, we basically scraped this thing called the book of yields.
[22:42.320 --> 22:51.040]  It's kind of a common book that they use in commercial kitchens where it says a head of cabbage yields this many cups of chopped cabbage, right?
[22:51.040 --> 22:52.060]  And so we have this...
[22:52.060 --> 22:56.260]  You're able to convert between the units of mass as purchased and volume, right?
[22:56.420 --> 23:02.280]  So the engine that we have, this live API, is able to say one cup of apple weighs 13 ounces, right?
[23:02.340 --> 23:04.080]  Or one onion weighs that many ounces.
[23:04.140 --> 23:06.240]  One pot of cardamom is 0.37 grams.
[23:06.500 --> 23:08.280]  And that's an important...
[23:08.280 --> 23:14.540]  Every recipe website or conversion or even USDA website has this algorithm and lookup table with all of its special cases.
[23:14.720 --> 23:18.220]  And so we're kind of making this public so that people are able to convert these things if they want.
[23:18.220 --> 23:23.320]  As purchased is a unit that's used in restaurants a lot, like a case of bananas, right?
[23:23.500 --> 23:24.560]  Or a clove of garlic.
[23:24.760 --> 23:28.180]  It's something that's neither mass nor volume, but has a mass and has a volume.
[23:29.160 --> 23:30.100]  So that's kind of interesting.
[23:30.340 --> 23:31.280]  Oh yeah, this is also...
[23:31.280 --> 23:32.460]  Because this is horrible graphics.
[23:32.580 --> 23:33.420]  I take full responsibility.
[23:33.640 --> 23:36.240]  All the beautiful graphics, by the way, are done by Jules, the co-chef.
[23:36.300 --> 23:36.800]  Give him a hand.
[23:37.280 --> 23:38.040]  Jules right there.
[23:38.340 --> 23:40.580]  We'll give him more of a hand once we show the beautiful graphics.
[23:41.040 --> 23:42.420]  Here's the shitty graphics that I made.
[23:42.500 --> 23:43.100]  This is me eating.
[23:43.220 --> 23:48.800]  This is me charting the lifeline of how dead a piece of fruit is, or a piece of food is that I'm eating, right?
[23:49.000 --> 23:50.660]  So the top one is salad dressing.
[23:50.920 --> 23:53.700]  So salad dressing, when you create it, is this point right here.
[23:53.980 --> 23:57.040]  The black pepper is alive for a while and then dead while it's dried.
[23:57.140 --> 23:57.860]  Mustard, same thing.
[23:57.960 --> 23:59.400]  Onion powder is alive until it's dried.
[23:59.540 --> 24:00.900]  Rapeseed oil is harvested and dried.
[24:01.020 --> 24:02.520]  Egg solids are alive and then dead.
[24:02.800 --> 24:05.380]  And they're all dead when you mix them together and make salad dressing.
[24:05.560 --> 24:07.220]  And then it sits on the shelf forever and you eat it.
[24:07.560 --> 24:08.940]  Compare that to sourdough, right?
[24:09.020 --> 24:10.820]  Where you have a starter, which is a live part.
[24:10.880 --> 24:12.100]  This green line means it's alive.
[24:12.280 --> 24:13.980]  It's an alive part of the food that you mix in.
[24:13.980 --> 24:16.200]  And the wheat was alive for a while and the sugar was alive for a while.
[24:16.280 --> 24:20.840]  But the thing that you're eating is dead because it sits on a thing for a while and the cooking process actually consumes it.
[24:21.080 --> 24:22.120]  Compare that to miso soup.
[24:22.160 --> 24:23.000]  Who likes miso soup?
[24:23.140 --> 24:23.320]  Everybody?
[24:23.540 --> 24:24.480]  Yeah, miso!
[24:25.060 --> 24:25.520]  All right.
[24:26.080 --> 24:28.960]  Miso is alive when you eat it because you're not supposed to boil it, right?
[24:29.040 --> 24:33.320]  So miso is this living culture that grows on soybeans and different kinds of starches.
[24:33.500 --> 24:34.780]  The green onions are alive for a while.
[24:34.880 --> 24:38.940]  Bonito flakes and seaweed are kind of good for you because they've been dead for a long time, but they're dried.
[24:39.020 --> 24:42.800]  A lot of Japanese cuisine is a lot of dried food because of the seasons there and harsh winters.
[24:42.800 --> 24:47.840]  And so when you make the miso soup, you're mixing these dead things and this alive thing, but it's actually alive and it puts it into you.
[24:47.980 --> 24:49.300]  So I don't know if you ever eat...
[24:49.300 --> 24:50.520]  Who here likes turkey?
[24:50.700 --> 24:50.980]  Turkey?
[24:51.140 --> 24:51.780]  Delicious turkey.
[24:52.020 --> 24:53.800]  You ever eat turkey and go to sleep, right?
[24:53.920 --> 24:54.840]  And you're like halfway through the day?
[24:55.000 --> 24:59.220]  Well, I have this friend who only eats live food during the day, like salads and stuff like that.
[24:59.300 --> 25:02.680]  And those are enzymes that you're putting into your body so that you have the energy for the rest of the day.
[25:02.800 --> 25:10.900]  If you're eating dead food in the beginning part of the day, your body has to exert energy to generate enzymes to break down all that stuff.
[25:10.900 --> 25:16.980]  And so I think it's interesting to kind of quantize how alive the food is that you're going to put in your body.
[25:17.480 --> 25:19.840]  Oh yeah, this is kind of a description of what I just said.
[25:20.000 --> 25:20.240]  Awesome.
[25:22.800 --> 25:23.940]  Who here's vegetarian?
[25:24.440 --> 25:26.360]  All right, let's put some vegetarian ones out there.
[25:26.520 --> 25:26.780]  Beautiful.
[25:29.540 --> 25:31.880]  Yeah, some vegetarian come up and pass it out to the vegetarians.
[25:31.980 --> 25:33.360]  I'm sure you know the secret handshake, right?
[25:34.200 --> 25:34.800]  See your vegetarian?
[25:34.960 --> 25:37.240]  It's just, yeah, it's a right angle with a carrot.
[25:37.800 --> 25:39.540]  Okay, so let's also talk about quantizing.
[25:39.540 --> 25:42.400]  Who remembers sentence diagramming in high school or junior high?
[25:42.460 --> 25:43.060]  It's junior high, right?
[25:43.500 --> 25:43.640]  Right?
[25:43.760 --> 25:44.400]  Sentence diagramming.
[25:44.540 --> 25:46.600]  So we do the same thing with food where you take...
[25:46.600 --> 25:50.100]  The scraper reads this recipe for vegetable stock and it's able to...
[25:50.100 --> 25:51.740]  I talked earlier about components, right?
[25:51.960 --> 25:57.380]  Well, the components of vegetable stock is you roast these savory vegetables and you add this into a stock pot.
[25:57.460 --> 26:03.260]  And so we kind of built this recipe diagram generator where you would have all the savory vegetables right here.
[26:03.500 --> 26:04.760]  Then the action was roasting it.
[26:04.840 --> 26:07.100]  And then this is the preposition, really, the 45 minutes time.
[26:07.100 --> 26:11.420]  and then these are all the herbs that are getting added into the stock po after the vegetables get roasted.
[26:11.680 --> 26:18.520]  This is the white wine that is getting added to this equipment and then deglazed and that is added to the stock later because it is the flavors, the fond.
[26:19.120 --> 26:24.380]  And then you have this stock pot here with water being added to it and you are simmering it for an hour and you are straining it through cheesecloth.
[26:24.460 --> 26:27.020]  What this allows you to do is it allows you to replace different parts of it.
[26:27.420 --> 26:31.980]  You say, well I actually want to put some fennel in my product or put some beets or whatever...
[26:31.980 --> 26:34.060]  you're like, okay, that is part of the раст vegetable part.
[26:34.100 --> 26:34.760]  That goes in here.
[26:34.920 --> 26:36.240]  Or you say, okay, I have some basil.
[26:36.240 --> 26:38.800]  You're like, okay, well, that goes into the herb component list right here.
[26:38.900 --> 26:46.340]  Or you say, well, I don't have a cheesecloth, and it kind of downgrades using our taxonomy to, you know, pantyhose or a strainer or something like that.
[26:46.460 --> 26:49.860]  This is kind of a longer one called rhubarb syrup, where you're dicing these chunks.
[26:50.120 --> 26:54.140]  And one of the prepositions that's interesting here is thick syrup consistency, right?
[26:54.140 --> 26:59.360]  You're boiling this down for this amount of time or a thick syrup consistency, and removing it from heat and adding lemon.
[26:59.500 --> 27:01.140]  And what lemon's doing here is adding acid.
[27:01.260 --> 27:04.500]  So, of course, if you don't have lemon, you add a little bit of apple cider vinegar and things like that.
[27:04.500 --> 27:10.880]  So, kind of visualizing the food this way helps you think about what replacements to do, and also kind of what all the buckets are for this stuff.
[27:10.980 --> 27:12.580]  So, now we'll talk a little bit about the colors of food.
[27:12.700 --> 27:16.460]  So, this is early part of the genome graphics that we kind of started measuring.
[27:16.640 --> 27:18.720]  What are all the colors of all the food that we have, right?
[27:18.900 --> 27:24.520]  So, here you have like ahi tuna and apple and banana and beets and beans and stuff like that.
[27:25.120 --> 27:26.500]  And then here's kind of an idea.
[27:26.580 --> 27:27.900]  This is Chris's food that he's been eating.
[27:27.980 --> 27:31.040]  It's all gray because he's only filled up his personal food genome a little bit.
[27:31.180 --> 27:32.760]  But this is, you know, cremini mushrooms.
[27:32.760 --> 27:37.060]  And I really think that's important to look at the color of the food that you've been eating before you eat it, not after you eat it.
[27:37.140 --> 27:39.160]  Because mine's a little red and brown, if you know what I mean.
[27:39.600 --> 27:49.060]  And to basically be able to look at all these people's food genomes and say, this person's not very healthy because they have some tomatoes and things, but they're eating a lot of white food and a lot of brown food and a lot of yellow food, like oils and stuff like that.
[27:49.120 --> 27:53.820]  So, I think color is an important part, especially with kids, like how to think about food.
[27:53.820 --> 27:56.340]  Now we come to the trippy food mandalas.
[27:57.120 --> 27:58.440]  What do you think of that shit, huh?
[27:58.760 --> 28:01.200]  Another hand for Jules and the trippy food mandalas.
[28:01.500 --> 28:02.040]  All right.
[28:02.360 --> 28:04.200]  So, the trippy food mandalas.
[28:04.360 --> 28:09.820]  This is the history of food and the middle column is kind of, this is some Dr. Seuss truffle the tree shit right here, right?
[28:10.020 --> 28:14.000]  It's how diverse your cuisine it is, the cuisine that you eat.
[28:14.120 --> 28:16.340]  And so we're able to map different corpuses.
[28:16.340 --> 28:21.920]  Now that we've used the food genome to scrape these and we know about the context and we know this is a corpus from allrecipes.com website.
[28:21.920 --> 28:23.640]  This is a corpus from this cookbook.
[28:23.800 --> 28:26.020]  This is a corpus from this person's Twitter feed.
[28:26.200 --> 28:29.540]  We're able to kind of visualize those and be able to say things like, well, recipes.
[28:29.740 --> 28:37.460]  So this is a recipe source for all the antipasto recipes on allrecipes.com and a lot of them have red bell pepper and I think there's a further filter on this.
[28:37.660 --> 28:39.760]  And this is all the sub antipastos, right?
[28:40.080 --> 28:42.120]  So this one's down here all, you know, very green.
[28:42.320 --> 28:44.380]  These ones are pretty brown because they have a lot of dough in it.
[28:44.480 --> 28:45.480]  It's like antipasto with dough.
[28:45.640 --> 28:47.880]  And this one's a very cheese heavy because it's got orange, right?
[28:48.160 --> 28:49.200]  These ones, they look pretty...
[28:49.200 --> 28:52.660]  So you can see, just by looking at all these, how diverse each of these recipes are.
[28:52.720 --> 28:59.660]  These are visualizations of recipes by their color, how healthy they are, how diverse they are, and about what the proportions are of all the food.
[28:59.920 --> 29:01.000]  So that's kind of cool, huh?
[29:01.140 --> 29:02.300]  There's like little percentages or whatever.
[29:02.460 --> 29:04.100]  This is actually for regions of cuisine.
[29:04.100 --> 29:11.060]  So given a corpus of recipes, these are Chinese recipes, these are Japanese recipes, these are Indian recipes, these are Korean recipes.
[29:11.060 --> 29:16.060]  So you have like a lot of onion and milk and rice and shit in the Indian food, versus Singapore and Thailand.
[29:16.060 --> 29:22.440]  So you're able to kind of look at these cuisine types by region and say, oh, this is different than this cuisine, and maybe this is a little more diverse.
[29:22.640 --> 29:31.880]  Like Singapore... I mean, also, it's kind of reflective of how shitty our recipe source is, that the Singapore recipes aren't more diverse than Japanese food, because Singapore food is actually pretty diverse.
[29:32.060 --> 29:33.400]  So that's interesting.
[29:33.480 --> 29:38.420]  We're able to use these food mandalas to look at recipes, regions, and compare your diet over time, right?
[29:38.600 --> 29:47.600]  So this is, you know, day one, day two, day three, day four, day five, day six, and then it kind of, you know, gets these extra little layers here of this column rising up.
[29:47.800 --> 29:56.440]  And because it's by quantity, because it's by percentage of mass, it's interesting because a lot of the stuff that builds up at the tip is like herbs and spices and things like that.
[29:56.620 --> 29:59.080]  All right, we still have a beautiful... plenty of time.
[29:59.280 --> 30:00.740]  It's almost 420, everybody.
[30:01.100 --> 30:01.660]  Just so you know.
[30:02.140 --> 30:02.380]  All right.
[30:02.720 --> 30:04.580]  This is a measure of my food over time.
[30:04.760 --> 30:08.300]  So then we paid my friend Chloe, who's an amazing anime artist.
[30:08.300 --> 30:10.700]  I should even just like look at her website, right?
[30:10.920 --> 30:13.140]  Or our website, foodgenome.com.
[30:13.380 --> 30:15.200]  She just drew every food, right?
[30:15.360 --> 30:17.300]  She drew about like 500 of the main foods.
[30:17.980 --> 30:18.920]  We're on the web, right?
[30:19.160 --> 30:20.880]  Maybe a lot of people are looking at our website right now.
[30:21.000 --> 30:26.980]  So she drew all the food groups and they're all cute and smiley and anime and stuff like that.
[30:27.120 --> 30:28.080]  I know it's super adorable.
[30:28.280 --> 30:39.200]  And so we started using this platform of the Food Genome to make, you know, it's all well and good to make interesting kind of chef tools up to this point where we kind of say, here's all the food that we have in the kitchen.
[30:39.200 --> 30:39.940]  What should we make?
[30:40.040 --> 30:43.240]  And kind of understanding the extent of the food data out there.
[30:43.400 --> 30:45.020]  But how do we make that appealing to people?
[30:45.200 --> 30:47.520]  And so we started printing them out and building little food games, right?
[30:47.640 --> 30:53.360]  Like building these little cards and kind of like the delicious corpse stuff earlier, where you're kind of getting an idea of what to cook.
[30:53.640 --> 31:00.300]  You're actually, you know, building these things together and charting, you know, if these are, you know, we had refrigerator magnet backings on some of these, like on the fridge.
[31:00.420 --> 31:06.280]  And so just imagine with your kids, you have every food, you know, there's sugar, and they have these cute little anime faces and stuff on them.
[31:06.700 --> 31:09.480]  You know, just kind of putting them on the fridge and saying, this is the food that I want.
[31:09.600 --> 31:11.160]  And basically being able to play with your food.
[31:11.620 --> 31:16.040]  And to truly, because food is a, for someone to cook, it takes a big commitment.
[31:16.820 --> 31:22.160]  And there's games out there, like these restaurant, virtual restaurant games in Farmville or Cafe Land or whatever.
[31:22.360 --> 31:26.880]  Those are kind of interesting, but they don't actually, you know, translate into, maybe they do translate into real life.
[31:27.000 --> 31:28.000]  I haven't really played them very much.
[31:28.000 --> 31:37.360]  But I think the challenge here with the food genome is to use that as a platform to build these food games on it that actually do affect how people see foods and give them different ideas for things.
[31:37.600 --> 31:43.360]  Like if you ever play Scrabble, you have a bag of chips and you just reach in and grab some letters and build dirty words and stuff like that, or refrigerator magnets.
[31:43.640 --> 31:47.480]  You can do the same thing with food where you're just grabbing two things and you say, oh, lettuce and balsamic vinegar.
[31:47.600 --> 31:48.200]  That sounds good.
[31:48.240 --> 31:48.780]  What else goes?
[31:48.920 --> 31:50.400]  Okay, not this, not this, not this.
[31:50.520 --> 31:55.860]  And it's a lot easier than going to a market for some people because a lot of people don't have markets to go around to.
[31:56.200 --> 31:57.280]  So who remembers Shanghai?
[31:57.280 --> 31:57.440]  Shanghai.
[31:57.580 --> 31:58.720]  Everyone loves Shanghai, right?
[31:58.860 --> 31:59.100]  Beautiful.
[31:59.300 --> 32:10.280]  So we started looking at, you know, how to build food games that kind of, you know, instill in someone this idea that you can play with your food and that you should break down your barriers of what you expect.
[32:10.420 --> 32:15.400]  And so we take these, you know, tiles from Food Shanghai and start looking at pattern matching, you know, just among it.
[32:15.480 --> 32:22.400]  And you say, well, these are, you know, when someone's playing this Food Shanghai game, they're looking at a lot of, it looks like a lot of visual noise, but there are these patterns that you're detecting.
[32:22.520 --> 32:27.220]  And the patterns are things like count and color and shape of things, right?
[32:27.220 --> 32:35.440]  There's a lot of long, stocky things and round things and different amounts of things versus three or many and different kinds of colors of these foods.
[32:35.600 --> 32:40.940]  And because this game is from the Super Nintendo, they don't really have a very good color palette.
[32:41.060 --> 32:48.000]  And so we definitely have a richer color palette, but we got a lot of ideas for just when people would use food for art for games from back in the day.
[32:48.200 --> 32:55.520]  And so we started building, with Fugino, things like food Tetris, things like food Minesweeper, things like food Risky Bus or Chaos.
[32:55.520 --> 32:57.860]  Who here remembers Risky Bus or Chaos Bot on IRC?
[32:57.980 --> 32:58.280]  Anybody?
[32:58.540 --> 32:58.860]  Risky Bus?
[32:59.020 --> 32:59.160]  Nobody?
[32:59.480 --> 33:00.820]  It was basically, you don't know Jack.
[33:01.440 --> 33:04.180]  And we built this... well, Jules built this chat bot, right?
[33:04.320 --> 33:07.840]  It would have a new game and it would say, hey, the game, the current game is tomato and basil.
[33:08.160 --> 33:09.020]  You're all logged in.
[33:09.180 --> 33:11.140]  What is your good idea with tomato and basil?
[33:11.360 --> 33:15.540]  And so you message the bot and you say, well, I would make a tomato basil milkshake.
[33:15.700 --> 33:18.060]  And someone else would say, okay, I would make a tomato basil casserole.
[33:18.240 --> 33:20.860]  And they would post the results and people would vote on these things.
[33:20.980 --> 33:22.700]  And so it's kind of a mechanical turkey, if you will.
[33:22.700 --> 33:28.600]  This head-to-head multiplayer food Twitch game that helps you kind of...
[33:28.600 --> 33:33.600]  Because that's really the key about having an algorithm like the Food Genome is it needs human input.
[33:33.760 --> 33:36.240]  It needs people to go in and say, this is gross, this is good.
[33:36.320 --> 33:37.240]  This is gross, this is good.
[33:37.320 --> 33:40.580]  Or this is gross to Europeans and this is delicious to Americans, right?
[33:40.760 --> 33:45.240]  And so we were trying to build these Twitch games that are these ideas of how to...
[33:45.680 --> 33:53.700]  of how to take this firehose of food data that's coming in on the existing Food Genome website and the Twitter bot where you're messaging the Twitter bot what you're eating and building a food diary.
[33:54.100 --> 33:56.240]  Or you're just going in and teaching it, I like this, I don't like this.
[33:56.460 --> 34:01.000]  And trying to make sense of that in a really simple kind of like Tetris-y Twitch game kind of idea.
[34:01.860 --> 34:02.940]  So here's, yeah, tweet your meat.
[34:03.080 --> 34:03.560]  That's pretty good.
[34:03.700 --> 34:08.380]  So I send Food Genome, this Twitter bot, you know, what I've been eating that I ate.
[34:08.540 --> 34:12.020]  Indonesian pineapple, sausage, fennel, and garlic that's in this stuff today.
[34:12.020 --> 34:19.840]  And Food Genome kind of writes back and helps me, gives me a bitly address or nomly address, a tiny URL to that URL, right?
[34:20.540 --> 34:22.600]  And so it's kind of just building this genome, this concept.
[34:22.840 --> 34:26.700]  And so these are little game blocks that can be used in other kind of games.
[34:26.880 --> 34:28.880]  Someone else can look at that and say, that's disgusting.
[34:29.020 --> 34:31.680]  Maybe there's a Canadian chat room that just says, that does not sound Canadian.
[34:31.840 --> 34:35.000]  It's just the 100% official, you know, Canadian stamp of approval or whatever.
[34:35.280 --> 34:43.420]  We also built this iPhone app for doing like a recipe ideas on the go where we loaded a corpus of about 60,000 recipes on an iPhone app and said, Hey, I've got blueberries.
[34:43.580 --> 34:44.180]  What goes with blueberries?
[34:44.380 --> 34:44.880]  Oh, honeydew.
[34:45.040 --> 34:47.060]  Okay, here's ideas for honeydew and blueberry, right?
[34:47.320 --> 34:49.120]  And it also has regions in it as well.
[34:49.200 --> 34:51.920]  So you could say, I want, what would I do with blueberries that's French?
[34:52.140 --> 34:53.480]  Just really simple kinds of ideas.
[34:53.640 --> 35:02.960]  So these are just kind of, after we finished this kind of the back end of this Food Genome platform, we started building these apps and games on top of it and trying to figure out what we wanted to play around with food.
[35:03.620 --> 35:04.320]  What else we got?
[35:04.440 --> 35:06.000]  Okay, here comes the cool visualization shit.
[35:06.180 --> 35:07.860]  So this is the corpus of Mango.
[35:07.860 --> 35:12.500]  So this is like 199 mango pairings that are all kind of compared across these axes.
[35:12.820 --> 35:13.840]  What do the axes mean?
[35:17.430 --> 35:19.020]  Right, he says, don't worry, I report it.
[35:19.450 --> 35:22.540]  He says the higher up these pairings are, the better they taste with mango.
[35:22.700 --> 35:27.720]  So lime juice and salt are up here, but like broccoli and apple juice and tuna are kind of down here.
[35:27.820 --> 35:30.280]  So these are all things that go with mango, maybe, but maybe not.
[35:30.460 --> 35:36.160]  And then this is two axes where you have the Y axis is mango and then the X axis is fruit salad.
[35:36.160 --> 35:44.880]  And it says, if mango is in fruit salad, it goes really well with apple and lemon and marshmallow, but does not go very well with paprika and ham and ground nutmeg.
[35:45.100 --> 35:51.600]  But those are still things that go well with mango because they're up here, whereas they don't go well with fruit salad because they're on this side of the axis.
[35:51.600 --> 35:55.560]  So these are just like the preliminary kind of, you know, plotting stuff we worked on.
[35:55.660 --> 35:57.300]  And then we started moving into this crazy shit.
[35:57.380 --> 35:58.000]  You ready for this?
[35:58.100 --> 35:59.260]  The food star charts.
[35:59.580 --> 36:01.260]  Maybe I'll even, I think there's a monkey view.
[36:01.460 --> 36:02.340]  Here's a monkey view.
[36:03.360 --> 36:05.960]  So Joel started making these food star charts, right?
[36:06.060 --> 36:06.800]  This is an interesting one.
[36:06.960 --> 36:08.360]  This is time versus time.
[36:08.580 --> 36:11.140]  This is everything that goes with time.
[36:11.340 --> 36:13.680]  Oh, I don't know if I can view that whole thing on the screen.
[36:13.840 --> 36:15.660]  So these are things that don't go with time at all.
[36:15.660 --> 36:17.300]  And then these are things that do go with time.
[36:17.480 --> 36:21.500]  And they're in... it's like a corpus of food recipes is built... is fed into this.
[36:21.640 --> 36:31.360]  And it says, according to this corpus of 40,000 recipes from whatever website, these are the foods that are mostly with... you know, aggregate all the mass that they were in the recipes with time, right?
[36:31.520 --> 36:32.840]  And so these ones... what is the one?
[36:32.950 --> 36:36.800]  Beef broth, mushrooms, chardonnay, okra.
[36:36.980 --> 36:38.620]  All those things go really heavily with time.
[36:38.900 --> 36:41.980]  And so I'm just going to jump back to this and say, this is salt.
[36:42.080 --> 36:44.640]  This is food that goes with salt versus food that goes with sugar, right?
[36:44.640 --> 36:52.340]  And so you get these really beautiful kind of trippy images where you're saying, you know, sugar has a lot more of this big fruit cluster down here and then this savory cluster up here.
[36:52.460 --> 36:55.700]  Whereas all the stuff that goes with salt is, you know, doesn't have those clusters.
[36:55.930 --> 37:04.540]  And I really recommend that you check out kind of our website and monkey view for some of these... for some of these images because they're... they're f*cking gorgeous.
[37:04.880 --> 37:06.520]  Where... just different ingredient pairings.
[37:06.860 --> 37:10.100]  And you can just zoom way in and kind of explore them kind of like star charts.
[37:10.300 --> 37:17.760]  And we have an open API for visualization where you can grab these and you're really just requesting what goes with this, weight them, and what are the colors of these food.
[37:18.000 --> 37:21.260]  And the thing that we've really learned from doing this is that all foods go together.
[37:21.540 --> 37:23.430]  Like seriously, all foods f*cking go together.
[37:23.500 --> 37:26.000]  You just have to find something else to link it in.
[37:26.180 --> 37:27.000]  You know, you could say...
[37:27.000 --> 37:29.860]  You could say chocolate doesn't go with squid, right?
[37:30.120 --> 37:32.840]  But it does if you mix milk and fennel with it, right, Veggie?
[37:33.160 --> 37:33.430]  Yeah?
[37:33.660 --> 37:33.930]  No?
[37:34.380 --> 37:35.060]  You weren't into that?
[37:35.240 --> 37:39.930]  He brought me a squid once and woke me up in the morning, put a three-foot squid next to my head and said, make this.
[37:39.930 --> 37:41.300]  And I was like, all right, I'm going to make it chocolate.
[37:41.660 --> 37:43.040]  So we made this chocolate squid.
[37:43.240 --> 37:44.880]  And I thought it was pretty good, but he thought it was gross.
[37:45.700 --> 37:51.540]  So this is kind of our visualization again of these nice little anime sprites and trying to figure out, you know, which foods kind of go together.
[37:52.060 --> 37:53.900]  So we have, you know...
[37:53.900 --> 38:00.360]  But foods go together according to me, according to my kid, according to the country of France, according to the country of France in the 1800s.
[38:00.480 --> 38:05.360]  We have all these different contexts and recipe sources that allow you to kind of figure out what your angle is.
[38:05.360 --> 38:13.500]  And you can say, chocolate doesn't go with squid, except for this one really weird British recipe from the 1600s or something like that, which is before they had chocolate, which is trippy.
[38:14.380 --> 38:14.740]  All right.
[38:15.040 --> 38:17.620]  So then we started also using this API to build...
[38:17.620 --> 38:20.280]  Well, Jules and Elise made these...
[38:20.280 --> 38:22.560]  These are food place mats.
[38:22.680 --> 38:26.060]  And this is United States cuisine in the background, right?
[38:26.180 --> 38:29.840]  Which is all burgers and fries and shit with burgers and fries on it, right?
[38:30.000 --> 38:31.720]  This is, I believe, French food, right?
[38:31.840 --> 38:33.480]  With an omelet and stuff.
[38:33.480 --> 38:35.800]  Oh, no, this is Spanish food, actually, with, like, red pepper.
[38:36.080 --> 38:37.040]  And so this is also a placement.
[38:37.140 --> 38:39.760]  These are available on Etsy.com, I think, or maybe just the tote bags are.
[38:41.680 --> 38:42.780]  Where's the Choco Tacos?
[38:42.820 --> 38:44.360]  We haven't actually added that to the genome yet.
[38:44.360 --> 38:44.600]  Sorry.
[38:44.720 --> 38:45.720]  You can be the first...
[38:45.720 --> 38:47.180]  You can be the mayor of Choco Tacos.
[38:47.450 --> 38:58.520]  And so we're basically able to kind of bring the genome into this offline world where we are creating art out of these things and making huge posters that do the star charts and are actually, you know, building, you know, food.
[38:58.640 --> 38:59.800]  You're eating off your own diet.
[38:59.960 --> 39:03.780]  You're looking at this and you're saying, well, this is American cuisine and I'm eating American cuisine.
[39:04.040 --> 39:05.100]  It just messes with your head.
[39:05.360 --> 39:09.140]  And you look at it and you're like, man, there's a lot of mayonnaise in the corner right over here.
[39:09.400 --> 39:12.200]  There's a lot of, you know, milk and brown stuff and soy and whatnot.
[39:12.980 --> 39:13.760]  This is beautiful.
[39:13.900 --> 39:14.720]  A little unicorn break.
[39:14.880 --> 39:15.340]  What do you think?
[39:15.860 --> 39:18.320]  So we live in a hacker house called Unicorn Precinct 13.
[39:18.380 --> 39:27.380]  We've been doing these dinner parties for the past, what, 2004, 2005, where five to 35 people come over and who here has cooked with us before?
[39:27.540 --> 39:28.200]  Tim, where are you at?
[39:28.300 --> 39:28.820]  Dylan, everyone?
[39:28.820 --> 39:29.820]  Yeah, all right, beautiful.
[39:30.300 --> 39:34.600]  So people all come over, bring in groceries or chip in for groceries, and we all cook together.
[39:34.680 --> 39:39.300]  And I have all the crazy science equipment because I worked at the Fat Duck and trained in molecular gastronomy there.
[39:39.380 --> 39:42.120]  So liquid nitrogen and the vacuum pumps and all that shit.
[39:42.340 --> 39:45.800]  And I'm kind of over all that stuff, but it's just fun for people to come over and kind of play with that.
[39:46.040 --> 39:47.950]  And so at the Unicorn Precinct, we have these dinner parties.
[39:48.020 --> 39:49.660]  People come over and we all kind of cook together.
[39:49.800 --> 39:59.760]  And it's different than working in a restaurant or it's different than going to an underground cafe because we have this anarchist sensibility of there's the head chef who isn't going to fire you and you're not trying to execute their vision.
[39:59.930 --> 40:01.360]  It's like an anarchist pirate ship.
[40:01.500 --> 40:02.950]  It's like, follow me to certain plunder.
[40:03.120 --> 40:03.980]  What do you want to make?
[40:04.220 --> 40:04.860]  Oh, you're the...
[40:04.860 --> 40:05.500]  Okay, congratulations.
[40:05.500 --> 40:06.480]  You're the pirate captain now.
[40:06.600 --> 40:07.300]  You should take over.
[40:07.640 --> 40:10.500]  Like, Tim should make some pork loin stuff because he brought the pork loin.
[40:10.560 --> 40:11.320]  What do you want to do with it?
[40:11.450 --> 40:15.840]  And it's this idea that cooking, similar to programming, is a very collaborative art.
[40:15.950 --> 40:18.140]  And you're making something in the kitchen with a bunch of those people.
[40:18.140 --> 40:20.100]  And so we took this on tour a couple times.
[40:20.620 --> 40:22.680]  We went around the United States, around Europe a little bit.
[40:22.760 --> 40:26.520]  We've been in Japan where we just kind of hook up with people and curate.
[40:26.660 --> 40:28.580]  We'd, you know, drive for eight or nine hours, pull into Chicago.
[40:28.760 --> 40:31.340]  This is a tour we did when we did HOPE, I guess, here in 2006.
[40:31.720 --> 40:35.000]  We were in the middle of a tour where we went from San Francisco to Maine to Florida and back.
[40:35.300 --> 40:37.520]  10,000 miles in a biodiesel station wagon.
[40:37.880 --> 40:39.900]  We had 16 dinner parties in 32 days.
[40:40.040 --> 40:48.140]  We'd, like, drive for a while, pull into a town, go to the farmer's market, go to our friend's house, cook dinner with 25 people, make a six or seven course meal, and then drive to the next town the next day.
[40:48.640 --> 40:49.180]  And it was super fun.
[40:49.200 --> 40:50.880]  We got to cook with tons of our friends and family.
[40:50.950 --> 40:58.360]  And I think that's an interesting thing about Black Hat Food Hacking is that you are, you're creating a restaurant where the people get to touch the food.
[40:58.680 --> 41:01.180]  And as, you're not supposed to be able to do that in a restaurant.
[41:01.320 --> 41:06.760]  If you go to an underground supper club, there's people creating and executing this, but it's just their vision and you're the customer going in.
[41:06.840 --> 41:08.220]  You're not part of that experience.
[41:09.040 --> 41:13.020]  So this is kind of a big mosaic of all the food that Chloe drew for us.
[41:13.240 --> 41:14.340]  You should check her stuff out too.
[41:14.400 --> 41:20.880]  I especially like this weird kale thing up here that's in the unicorn's hair and then the little creepy lychee hanging up there.
[41:21.980 --> 41:26.300]  And of course, we also made a bunch of zines for a lot of different steps.
[41:26.430 --> 41:34.800]  One of the things we did with software development is we did this like report back idea where you have to take what you've done in the past two weeks and make a zine out of it and just kind of staple it together and then pass it out to your friends and be like,
[41:34.880 --> 41:36.360]  hey, look, we did visualization this week.
[41:37.800 --> 41:38.080]  Hungry.
[41:38.160 --> 41:39.620]  I don't know what the goat feed is for.
[41:40.400 --> 41:45.360]  It's kind of a weird slide, but I'm hungry just from talking to you guys.
[41:45.360 --> 41:46.200]  Let's get talking about food.
[41:46.360 --> 41:47.140]  This is the HOPE Zero Day.
[41:47.320 --> 41:48.260]  So Jules made this stuff.
[41:48.430 --> 41:49.260]  I helped out as well.
[41:49.280 --> 41:50.200]  I made all this fruit leather.
[41:50.340 --> 41:51.900]  I was blending all the...
[41:51.900 --> 41:58.100]  I roasted a bunch of fennel, but then that was too dry, so I had to put some parchment over it so it would steam and we'd keep some of the moisture inside of it.
[41:58.200 --> 42:03.400]  And I tossed it with olive oil, salt, and cayenne and then roasted that and then took it out and blended that with some pineapples.
[42:03.500 --> 42:07.280]  I had poached the pineapples, just kind of cooked them down by their self with some star anise.
[42:07.450 --> 42:17.840]  And then I blended that together in the Cuisinart and kind of spread it out eighth inch thick on cookie sheets and did them in the dehydrator and then rolled them up after about 12 hours of dehydrating and brought them here to HOPE.
[42:18.040 --> 42:21.540]  And then I brought some sausages from San Francisco as well that were cured so that you guys wouldn't get food poisoning.
[42:21.660 --> 42:21.980]  You're welcome.
[42:23.620 --> 42:24.340]  And then...
[42:24.340 --> 42:25.120]  Yeah, have some, man.
[42:25.200 --> 42:25.600]  Do you want some?
[42:26.320 --> 42:27.000]  Yeah, beautiful.
[42:27.240 --> 42:28.500]  That's my egg guy right there, Tim.
[42:29.080 --> 42:30.500]  He brings eggs over sometimes.
[42:30.760 --> 42:34.200]  And then Jules made these different pastry creams with different kinds of spices in them and stuff like that.
[42:34.260 --> 42:35.740]  So it's like a star anise pastry cream.
[42:35.880 --> 42:38.320]  I think the other pastry cream was like bay leaf and cardamom.
[42:38.500 --> 42:39.380]  And this is just...
[42:39.380 --> 42:39.640]  I don't know.
[42:39.740 --> 42:45.140]  It's just fun to cook for a bunch of people and to have it on stage and to have something extra instead of kind of going up here and trying to sell my book or something like that.
[42:45.200 --> 42:45.620]  I don't have a book.
[42:45.760 --> 42:48.620]  But, you know, trying to sell a book and kind of describe something to you.
[42:49.380 --> 42:54.880]  I feel like you kind of get what we're talking about a little bit about the direction that we're going in as far as like cuisine goes.
[42:55.900 --> 42:56.500]  So that's...
[42:56.840 --> 42:59.660]  Now I'm just going to say I've described you the Fuge genome.
[43:00.220 --> 43:02.320]  We are open sourcing it, making the source code available.
[43:02.520 --> 43:06.500]  It is a platform for other people to make food apps and food games.
[43:06.640 --> 43:08.880]  So if you have any ideas, please get in touch with us.
[43:09.080 --> 43:10.140]  Please Twitter with us.
[43:10.140 --> 43:11.500]  Talk to us about food.
[43:11.900 --> 43:12.780]  Tell us what you've been making.
[43:13.260 --> 43:15.950]  Make that a part of the dialogue of the hacker community.
[43:16.080 --> 43:20.340]  Just tell your own story because, you know, people always say, I'm not a cook, but how long have you been eating?
[43:20.450 --> 43:21.340]  You've been eating for decades.
[43:21.580 --> 43:23.360]  You know, everyone is an expert on what it is they like.
[43:23.450 --> 43:24.040]  Don't be afraid.
[43:24.320 --> 43:25.520]  And that's the end of the talk.
[43:25.620 --> 43:26.160]  Thanks very much.
[43:34.330 --> 43:39.630]  Just another hand for all of our volunteers from the food hacking team for making that awesome food and distributing it.
[43:41.870 --> 43:44.690]  And we'll take some questions if that's chill.
[43:44.690 --> 43:45.550]  Is there any questions?
[43:46.030 --> 43:46.370]  No?
[43:46.530 --> 43:46.930]  All right, thanks.
[43:47.910 --> 43:48.470]  What do you got?
[43:48.970 --> 43:51.510]  Just say who you are and what your favorite food is.
[43:57.040 --> 43:57.880]  Is there a switch?
[44:01.550 --> 44:02.430]  Try the back mic.
[44:06.010 --> 44:11.870]  And the talk, I really could even go into more stuff because we have these BBSs of all this kind of weird food.
[44:12.150 --> 44:12.590]  Is anyone here...
[44:12.590 --> 44:13.570]  Actually, I'm just going to go into one more thing.
[44:13.630 --> 44:14.930]  Does anyone know about Rotten.com?
[44:15.150 --> 44:15.330]  Anybody?
[44:15.570 --> 44:16.330]  Rotten.com, right?
[44:16.490 --> 44:17.070]  Yeah, Rotten.com.
[44:17.070 --> 44:21.770]  So we did this dinner for Rotten.com some sort of anniversary and we tried to make the most disgusting food possible.
[44:22.570 --> 44:24.830]  So this is, I think, blood...
[44:24.830 --> 44:28.450]  This is black custard where we took pig's blood and mixed it with...
[44:28.450 --> 44:29.610]  Let me see if it'll zoom in.
[44:29.810 --> 44:34.310]  We took pig's blood and mixed it up with cream and then kind of made a little custard.
[44:34.530 --> 44:35.970]  It's like half and half pig's blood and cream.
[44:36.550 --> 44:37.950]  And let's see if this loads.
[44:38.190 --> 44:38.570]  Here we go.
[44:39.750 --> 44:40.950]  Okay, this is black custard.
[44:41.110 --> 44:46.290]  So it's pig's blood with cream in it and then we chilled it in the fridge and then put a little sugar on it and torched it.
[44:46.470 --> 44:49.770]  And it was like cold blood milk and then the scab was on top.
[44:49.870 --> 44:51.970]  And so you taste it and it tastes exactly like a scab.
[44:52.230 --> 44:53.050]  It's pretty awesome.
[44:53.150 --> 44:56.730]  We also did durian fruit ice cream which, you know, tastes like nastiness.
[44:57.270 --> 44:57.690]  What else?
[44:57.770 --> 44:59.230]  We also did black...
[44:59.930 --> 45:01.630]  Kind of made like a black olive oil.
[45:01.770 --> 45:02.330]  This is kind of good.
[45:02.470 --> 45:05.470]  Black cod poached sous vide in its own...
[45:05.470 --> 45:06.510]  In black oil.
[45:07.310 --> 45:08.770]  So it's like a fish in black oil.
[45:08.890 --> 45:10.430]  So how do you make black oil, right?
[45:10.770 --> 45:11.790]  Well, you can't...
[45:11.790 --> 45:13.430]  It's just black olive oil, right?
[45:13.570 --> 45:14.670]  But how do you get a black olive oil?
[45:14.730 --> 45:15.310]  Well, you just...
[45:15.730 --> 45:21.870]  You get a bunch of olives and you blanch them four times to get rid of the brine flavor out of them and then you blend them...
[45:21.870 --> 45:23.890]  You dehydrate them so it gets all the water out of them.
[45:24.010 --> 45:24.770]  Let's see if I can load it again.
[45:25.030 --> 45:26.110]  Oh, I might be off the Internet.
[45:26.890 --> 45:27.310]  And then you...
[45:27.310 --> 45:27.910]  Here we go.
[45:28.030 --> 45:28.590]  It's on now.
[45:28.970 --> 45:30.290]  You blanch them four times.
[45:30.450 --> 45:31.630]  You take the olives out.
[45:31.770 --> 45:38.130]  You dehydrate them to get all the water out of it and then you just have these fatty, you know, fat solids of olives.
[45:38.150 --> 45:41.270]  And you blitz them with just olive oil and it turns it black, right?
[45:41.390 --> 45:43.930]  So it's this evil black, but it tastes just like olive oil and it's kind of nice.
[45:44.050 --> 45:44.730]  So it's like an oil spill.
[45:45.130 --> 45:45.910]  All right, what's the question?
[45:46.010 --> 45:46.630]  Who's got a question?
[45:46.790 --> 45:48.330]  All right, so can you make that...
[45:48.930 --> 45:56.290]  Does that soccer let you take the food genomes of, let's say, the whole family and figure out recipes that everyone's going to like?
[45:56.730 --> 45:58.010]  Yeah, I should have said that, huh?
[46:00.170 --> 46:00.530]  Yes.
[46:01.250 --> 46:02.350]  Okay, that's very cool.
[46:02.570 --> 46:04.550]  Yeah, so what's your favorite food?
[46:05.450 --> 46:08.950]  Well, I can't have it that often, but a really nice roast beef sandwich.
[46:09.250 --> 46:15.630]  Yeah, we didn't really touch much about the diet stuff, but we also know what food's vegetarian, what's vegan, what's good for celiacs, what's good for diabetics.
[46:15.690 --> 46:20.190]  We have an API for that where you can feed it like lasagna and it suggests replacements for things like that.
[46:20.250 --> 46:32.870]  And I think that's a key thing, especially with families, where you say, well, I've got these four kids and maybe there's a stupid cell phone game that teaches the kids to put in the foods that they hate or that they already had or that they don't like or that they want to try or that are interesting.
[46:32.870 --> 46:37.530]  Hopefully they're all cute and anime looking and then kind of generating recipes based on that.
[46:37.930 --> 46:38.590]  So, yeah, what do you got?
[46:39.230 --> 46:42.510]  This is very similar to the previous guy's question, but it's actually the opposite.
[46:43.090 --> 46:52.190]  If I have a whole bunch of different food and I don't know what to make from it, is it possible using your API or an application that myself or someone else could write and say, how did these go good together?
[46:52.330 --> 46:53.570]  Which one of these could I put together?
[46:53.730 --> 46:54.470]  And what could I even make?
[46:54.930 --> 46:56.110]  Yeah, yeah, 100%.
[46:56.110 --> 46:56.910]  So it's like, what's...
[46:57.290 --> 47:02.510]  So a lot of the recipe websites are doing that now in terms of, it's just a search.
[47:02.670 --> 47:04.050]  You say, I have salt and bread.
[47:04.190 --> 47:06.610]  And it says, okay, make a salt sandwich or whatever it tells you to make.
[47:08.290 --> 47:09.550]  I haven't found one of those yet.
[47:09.650 --> 47:10.630]  Do you know of one that does that?
[47:10.770 --> 47:11.330]  They all suck.
[47:12.090 --> 47:12.370]  Exactly.
[47:12.710 --> 47:13.890]  I think, like, I don't know.
[47:13.950 --> 47:14.550]  There's a bunch of them.
[47:14.630 --> 47:16.290]  Just search for, you know, let's see this.
[47:16.730 --> 47:17.090]  What is it?
[47:17.370 --> 47:18.430]  I don't remember all the names.
[47:18.610 --> 47:23.130]  But I think the interesting thing about ours is that it learns what you like and what you have and haven't had.
[47:23.130 --> 47:36.190]  And in addition to that, we have some weird prototype software that actually creates, use it for creating weird dishes and just kind of suggesting strange dishes and strange combinations of food versus figuring out the ones that people, we know that we already like.
[47:36.270 --> 47:40.050]  And so we can look at you and say, well, you're vegan and you hate pasta.
[47:40.150 --> 47:41.650]  So these are some foods that you would like.
[47:41.750 --> 47:42.930]  It's kind of looking at the whole...
[47:42.930 --> 47:47.630]  It's very similar to Amazon or Pandora where it's saying people that, you know, you type in all the food that you've been eating.
[47:47.870 --> 47:48.970]  You know, you've been eating a lot of Chinese food.
[47:49.050 --> 47:49.750]  You've been eating a lot of sandwiches.
[47:49.770 --> 47:54.250]  And it says, well, here's food that you would probably like because other people who also eat these things would like these things.
[47:54.530 --> 47:58.970]  So yes, you can query the stuff that's in your fridge and it can tell you these are recipes you can make out of those.
[47:59.930 --> 48:00.810]  Okay, that's fine.
[48:01.270 --> 48:01.550]  Thanks.
[48:03.830 --> 48:04.850]  I got a question.
[48:04.930 --> 48:05.510]  What do you got?
[48:06.610 --> 48:10.110]  I don't have any myself, but a friend of mine is jewish.
[48:10.290 --> 48:12.010]  Can you put on restrictions for kosher?
[48:12.250 --> 48:13.890]  Yes, we have kosher, halal.
[48:14.090 --> 48:15.150]  What other religions do we have?
[48:15.230 --> 48:17.010]  It's part of the world taxonomy.
[48:17.010 --> 48:25.730]  If you go to food genome slash region, there's like all the continents and that we might be super slow because of everyone hitting it and we suck at scaling.
[48:26.750 --> 48:28.670]  Someone should rewrite our website for us.
[48:28.710 --> 48:29.150]  That'd be awesome.
[48:29.610 --> 48:37.910]  Yeah, there's a religion taxonomy that allows us to... that has like Buddhist food, kosher, halal, all that stuff.
[48:38.290 --> 48:38.970]  One more though.
[48:39.330 --> 48:40.230]  Is there any...
[48:40.230 --> 48:45.250]  And suggest replacements for those things to say to make this lasagna kosher, take the cheese out of it.
[48:45.490 --> 48:45.990]  Sounds gross.
[48:45.990 --> 48:50.090]  Is there any ability to expand this into something like mixed drinks or whatnot?
[48:50.330 --> 48:55.390]  Like to get different ones from around the world, people posting what they like, what just weird things that people try together were.
[48:55.630 --> 48:58.590]  Yeah, so the question is, is there any possibility to expand it into mixed drinks?
[48:59.130 --> 49:01.810]  Yeah, we have a lot of mixologist friends who feed us their data.
[49:02.010 --> 49:05.270]  I think the other interesting thing is feeding a chef's au boi into it.
[49:05.370 --> 49:15.770]  Feeding all of Julia Child's cookbooks versus just one Julia Child or all the chefs that are in a city and say, according to Thomas Keller from the French Laundry plus McDonald's, what would he make for me right now?
[49:15.910 --> 49:17.330]  By the way, I'm vegan.
[49:17.790 --> 49:20.790]  And by the way, you know, beef tallow gives me the shits.
[49:20.990 --> 49:22.470]  So don't mix it in with something.
[49:23.350 --> 49:25.110]  So yes, that is a...
[49:25.750 --> 49:26.510]  Yes, yes.
[49:26.630 --> 49:27.030]  Next person.
[49:27.210 --> 49:27.550]  Yeah.
[49:27.690 --> 49:32.630]  Is there a way to filter out according to, you know, I want to set my diet according to the tools I have.
[49:32.810 --> 49:38.110]  For example, I like to play around with my food, but most of the times I only have like a frying pan.
[49:38.110 --> 49:38.730]  Right, right, right.
[49:38.870 --> 49:39.830]  Yeah, that was one of the early...
[49:39.830 --> 49:40.730]  So we have equipment lists.
[49:40.850 --> 49:42.130]  You can basically build a kitchen object.
[49:42.290 --> 49:44.390]  And your kitchen object is all the stuff that's in your kitchen.
[49:44.510 --> 49:45.710]  The food that's in your kitchen, the equipment.
[49:45.930 --> 49:47.070]  So you can say, I'm camping.
[49:47.190 --> 49:49.850]  I just have a Dutch oven and a river, right?
[49:50.250 --> 49:52.850]  And it's like, all right, well, you got to chill the pasta down.
[49:52.910 --> 49:54.670]  So put it in the bag and put it in the river.
[49:54.810 --> 49:55.490]  Do you have pantyhose?
[49:55.750 --> 49:57.370]  You know, kind of a thing.
[49:57.510 --> 50:00.190]  So yes, there's an equipment abstraction that knows.
[50:00.350 --> 50:01.490]  There's a whole taxonomy that's on the way.
[50:01.550 --> 50:06.150]  I think it's live that understands the equipment that's in your kitchen.
[50:06.150 --> 50:08.590]  And also, indeed, the chefs that are in your kitchen.
[50:08.830 --> 50:15.790]  So a kitchen like ours, the Unicorn Precinct, you have chefs come by and you can say, given all the chefs that have been in our kitchen the past year, what will we make?
[50:15.930 --> 50:17.850]  And based on our equipment in our kitchen.
[50:17.930 --> 50:18.870]  What equipment do you have in your kitchen?
[50:18.950 --> 50:19.390]  Frying pan?
[50:20.170 --> 50:23.410]  A frying pan, maybe something to stir stuff in, and a cutting board.
[50:23.490 --> 50:24.190]  That's basically it.
[50:24.310 --> 50:24.710]  Right, right, right.
[50:24.810 --> 50:25.910]  We should get a knife, for one.
[50:26.290 --> 50:26.490]  Right.
[50:28.310 --> 50:30.910]  Yeah, no, that's... yeah, so you feed... it's the same thing.
[50:31.010 --> 50:33.350]  It's a different... it's just... I talked earlier about there's 60 different objects.
[50:33.630 --> 50:40.030]  Basically, one of them is equipment, and it has all the sub-equipment too, so you can say spoon, or slotted spoon, soup spoon, all that other shit.
[50:40.290 --> 50:40.850]  All right, thanks.
[50:41.030 --> 50:41.110]  Yeah.
[50:43.830 --> 50:47.550]  Yeah, as I was going back to my seat, I realized I had a couple other questions.
[50:48.890 --> 50:53.550]  So, how do you encompass the memes that are kind of outside of what you...
[50:54.350 --> 50:56.110]  what you're describing in your taxonomy?
[50:56.250 --> 51:06.510]  Like, for example, modern font software deals with things like serifs and kerning, and what happens when certain letters go together.
[51:06.850 --> 51:17.550]  So, for example, let's say you want to make that black olive oil, or you want to make cookies that are either crunchy or soft, but have similar ingredients, but you just treat them differently.
[51:17.830 --> 51:18.150]  Right.
[51:18.270 --> 51:24.510]  Or if you want... or like my grandmother taught me how to make crepes that have a handle on them by just pouring a little differently.
[51:24.950 --> 51:25.090]  Sure.
[51:25.410 --> 51:26.510]  So, there's a little bump on it.
[51:26.670 --> 51:27.290]  You know, stuff like that.
[51:27.650 --> 51:28.550]  So, rephrase your question.
[51:28.850 --> 51:29.230]  Okay.
[51:29.450 --> 51:30.270]  So, how do you...
[51:30.270 --> 51:36.290]  So, the kind of stuff that's outside of a recipe, that's outside of just ingredients...
[51:37.210 --> 51:37.390]  Okay.
[51:37.890 --> 51:38.910]  Yeah, so techniques...
[51:38.910 --> 51:41.070]  That changes... that can drastically change...
[51:41.070 --> 51:41.710]  Yeah.
[51:41.790 --> 51:43.310]  It's a hard thing.
[51:43.510 --> 51:43.970]  And I think...
[51:43.970 --> 51:50.150]  I mean, ideally we set off as being this... you know, our ultimate goal is to be the programming language for the replicators on Star Trek, right?
[51:50.150 --> 51:54.150]  You go, tea, Earl Grey hot, and it f*cking makes you tea, and it's Earl Grey and it's hot.
[51:54.530 --> 51:57.010]  But, you know, how... do you... do you over-steep it?
[51:57.070 --> 51:57.590]  Do you steep it?
[51:57.670 --> 52:01.630]  So, we hopefully break it down to that level where you are as far as the techniques go.
[52:01.830 --> 52:03.230]  But yeah, as long as you can find...
[52:03.230 --> 52:14.970]  It's also interesting because... because of the main language that we started in was English, it's not a very good culinary language versus like, say, French or Chinese, which are these huge ancient mother cuisines that have specific words for what you just said.
[52:15.010 --> 52:22.390]  Adding a handle onto a crepe or skimming scum off the top versus skimming fat off the top these are, you know, different words and different concepts.
[52:22.650 --> 52:24.590]  And so, I think we didn't really touch on that.
[52:24.690 --> 52:34.430]  There is this internationalization aspect where you would be able to, if we brought in a bunch of Brazilian food and a bunch of Chinese food, be able to ask it, you know, I have... I live in China.
[52:34.430 --> 52:35.430]  I speak Spanish.
[52:35.790 --> 52:38.810]  I want to, you know, what is... I want to go to the store in China.
[52:39.050 --> 52:42.370]  I need the words of the ingredients and the recipe in Spanish.
[52:42.510 --> 52:45.250]  This idea of kind of changing the internationalized things.
[52:45.250 --> 52:47.990]  But I think that's a trick about having different concepts.
[52:48.210 --> 52:48.810]  But, yeah, I don't know.
[52:49.370 --> 52:52.910]  It's such a narrow problem domain that I think that's a solvable problem.
[52:53.110 --> 52:54.410]  Like, how many terms are there?
[52:54.470 --> 52:56.470]  How many special, you know, a shake...
[52:56.470 --> 52:57.390]  We have a pinch, right?
[52:57.430 --> 53:00.050]  A pinch is like a sixteenth of a teaspoon or something, right?
[53:00.190 --> 53:02.470]  So, we can take these colloquial terms.
[53:02.470 --> 53:06.530]  But because they have contexts and they say, well, this is from the 50s.
[53:07.050 --> 53:08.590]  This teaspoon is from the U.S.
[53:08.710 --> 53:10.590]  50s versus a current teaspoon or from England.
[53:10.750 --> 53:15.790]  And so, you have these different ideas that the different units of measure are different in different regions as well.
[53:16.110 --> 53:16.370]  So...
[53:16.370 --> 53:19.910]  Well, another example was, you probably know this better than me.
[53:20.030 --> 53:25.270]  I don't remember the author who wrote this, but there was a 15-page how to fry an egg.
[53:25.450 --> 53:26.070]  Right, right.
[53:26.070 --> 53:32.390]  Which involved how to bring up the chicken and what to feed it during all those years and so forth.
[53:32.390 --> 53:33.210]  I mean, I'm not...
[53:33.210 --> 53:35.870]  We're not trying to be the editorial solution of getting rid of all recipes.
[53:36.130 --> 53:40.010]  And, you know, we're just trying to, you know, think of how many recipes there are on the Internet.
[53:40.250 --> 53:40.990]  And there's about a million.
[53:41.190 --> 53:47.010]  And we just want to, you know, raise it by a factor of a hundred or something and make tons more and, like, mutate them all.
[53:47.130 --> 53:51.310]  And if you look at a recipe, there's always secrets that you're not actually doing in the recipe.
[53:51.490 --> 53:52.790]  But it should at least give you...
[53:52.790 --> 53:57.230]  Recipes are a small part of what the food genome is because it's about the pairings of that recipe.
[53:57.350 --> 54:01.690]  It's taking that recipe out and saying, well, this recipe has meat and herbs, right?
[54:01.690 --> 54:05.590]  And abstracting it out so that you know that that meal you had yesterday had meat and herbs.
[54:05.730 --> 54:07.810]  And it's time for you to have something that's just vegetables and herbs.
[54:07.970 --> 54:08.730]  Because you already had meat.
[54:08.770 --> 54:12.370]  And you set this threshold that says you only want to have meat two or three times a week.
[54:12.590 --> 54:15.510]  So, I don't think we'll ever get rid of editorial recipes.
[54:15.930 --> 54:18.510]  But, you know, we're trying to quantify as much as possible.
[54:18.870 --> 54:20.250]  And you can add more objects to it.
[54:20.330 --> 54:24.630]  The whole point, I'm just saying, is to leave room for expansion for that kind of stuff.
[54:24.630 --> 54:24.770]  Yes.
[54:24.890 --> 54:25.130]  Yeah.
[54:25.230 --> 54:26.190]  Hopefully, it's just woke up.
[54:26.310 --> 54:26.810]  It's a brain.
[54:26.990 --> 54:28.550]  It'll hopefully learn more if people teach it.
[54:28.550 --> 54:28.830]  It's really cool.
[54:29.190 --> 54:29.370]  Thanks.
[54:31.710 --> 54:32.070]  Hi.
[54:32.970 --> 54:34.250]  This is going to sound really random.
[54:34.350 --> 54:38.390]  But I recently signed up for this thing called 23andMe where you spit in a cup and they give you a genome.
[54:38.770 --> 54:44.110]  And they give you raw data back to you eight weeks later.
[54:45.430 --> 54:52.270]  So, perhaps, would you consider taking raw human genome data and putting it into the food genome system?
[54:52.590 --> 54:52.690]  Yeah.
[54:52.710 --> 54:53.550]  I think that's super interesting.
[54:53.730 --> 54:55.670]  I think if I had a genome sequencer, this is what I do.
[54:55.770 --> 54:57.310]  I'd make it the perfect date night food.
[54:57.550 --> 55:00.550]  What I'd do is, you would set up, you'd make a reservation to a fancy restaurant.
[55:00.750 --> 55:01.350]  You and your date.
[55:01.510 --> 55:02.150]  It's your first date.
[55:02.310 --> 55:04.250]  And you send your genome profiles in, right?
[55:04.430 --> 55:08.030]  And then you make chewing gums that have the genetics encoded, right?
[55:08.130 --> 55:09.770]  And then you get to swap it, right?
[55:09.910 --> 55:13.670]  You get to, like, chew on each other's genetics at the beginning of the meal.
[55:13.670 --> 55:16.250]  And see, oh, it tastes kind of stinky.
[55:16.390 --> 55:18.210]  I don't think I want to have this relationship.
[55:18.630 --> 55:18.930]  So, yeah.
[55:18.970 --> 55:20.250]  I think there's a huge opportunity.
[55:20.310 --> 55:26.330]  And I welcome anyone suggesting any weird combinations of colors of things or human genetics.
[55:26.570 --> 55:27.770]  I'm big into eating people.
[55:28.010 --> 55:28.590]  So, yeah.
[55:29.410 --> 55:32.730]  There's also, oh, on that note, Hufu.
[55:32.950 --> 55:34.510]  Are you familiar with Hufu?
[55:34.750 --> 55:39.670]  It's a vegan human flesh substitute.
[55:39.910 --> 55:40.530]  Right, right, right, right.
[55:40.630 --> 55:40.810]  Wow.
[55:40.810 --> 55:43.770]  Yeah, we have a test where we did vegan blood for vegan goths.
[55:44.430 --> 55:44.710]  Right?
[55:44.830 --> 55:45.890]  Because they want to drink blood.
[55:46.070 --> 55:47.490]  But, you know, it would be real blood.
[55:47.630 --> 55:50.230]  And so it was like, I think it was ended up being, it was like beet juice, of course.
[55:50.450 --> 55:53.850]  It was like dehydrated parsley for the chlorophyll for the kind of tangy taste.
[55:54.830 --> 55:58.570]  And then a little bit of soy milk and then some nigiri, which is the stuff that firms up tofu.
[55:58.570 --> 56:01.310]  And so you can actually take the blood and use it as a coagulant.
[56:01.390 --> 56:03.830]  You can use it as a, it would firm up basically.
[56:04.090 --> 56:06.890]  Because blood is used as a thickener if you want to put it in a vegan sausage.
[56:07.070 --> 56:11.970]  So you can make a vegan blood sausage out of the vegan blood because it would thicken up because it had this nigiri in it.
[56:12.190 --> 56:12.490]  Which is...
[56:12.490 --> 56:13.110]  That's awesome.
[56:13.330 --> 56:13.650]  Thank you.
[56:13.650 --> 56:13.670]  Thank you.
[56:13.810 --> 56:13.950]  Yeah.
[56:16.810 --> 56:21.590]  Have you guys done anything with like brewing or cheese production inside of this?
[56:22.350 --> 56:23.190]  Well, toe cheese.
[56:24.550 --> 56:24.890]  Yeah.
[56:25.030 --> 56:25.190]  I don't know.
[56:25.250 --> 56:26.730]  We have a cheese taxonomy.
[56:27.070 --> 56:29.110]  I just, I know nothing about brewing.
[56:29.310 --> 56:33.810]  I'm one of those people who just leaves huge areas of knowledge.
[56:34.410 --> 56:39.050]  I like to be ignorant in huge areas of knowledge so that eventually when I discover it, like I just started drinking coffee about a year ago.
[56:39.170 --> 56:40.470]  I don't know if you guys can tell, I just had a coffee.
[56:40.650 --> 56:41.170]  It's pretty good.
[56:42.270 --> 56:44.910]  But now, it's been really fun for me because I'm discovering it for the past year.
[56:45.010 --> 56:49.770]  So I would be totally open to discovering more about brewing and trying to even understand how that taxonomy would be built in.
[56:49.870 --> 56:52.330]  But definitely in terms of cheese, especially in the...
[56:52.330 --> 56:59.850]  We've done a lot of development in the vegan cheese area too with all the thickeners and the rejuvelac and different kinds of nut milks that are fermented and stuff like that.
[57:00.030 --> 57:01.530]  You said you guys are in Bed-Stuy?
[57:02.230 --> 57:03.910]  Well, that was one place that we did a dinner at.
[57:04.010 --> 57:04.930]  We're actually based out of San Francisco.
[57:05.330 --> 57:05.810]  Ah, okay.
[57:06.570 --> 57:06.890]  Yep.
[57:08.270 --> 57:08.630]  All right, guys.
[57:08.690 --> 57:09.050]  What do you think?
[57:09.090 --> 57:09.850]  Is it fat or what?
[57:10.290 --> 57:10.610]  Yeah?
[57:11.170 --> 57:11.650]  All right.
[57:16.670 --> 57:17.170]  All right.
[57:17.890 --> 57:19.130]  Well, I'm just...
[57:19.130 --> 57:24.530]  I don't really have anything else to say, probably, except you should get in touch with us.
[57:25.190 --> 57:25.610]  I don't know.
[57:25.670 --> 57:29.110]  I had a bunch of other weird slides I wanted to show you guys, but they all went away.
[57:29.170 --> 57:30.190]  Well, I'm fine with that.
[57:30.530 --> 57:31.170]  Yeah, go ahead.
[57:31.230 --> 57:31.510]  What do you got?
[57:31.510 --> 57:34.450]  You made a food map for this is why you're fat.
[57:35.590 --> 57:36.070]  Right.
[57:36.390 --> 57:38.330]  Me in particular or just people in general?
[57:39.650 --> 57:40.130]  Okay.
[57:40.550 --> 57:40.950]  Yeah.
[57:41.330 --> 57:42.470]  This is why you're fat.
[57:42.810 --> 57:44.870]  Yeah, that's an interesting idea in terms of diet.
[57:45.070 --> 57:48.370]  I think that's super important, especially with obesity epidemic.
[57:48.370 --> 57:49.270]  People are eating corn.
[57:49.410 --> 57:49.910]  People are eating...
[57:49.910 --> 57:58.870]  I think that's a really kind of interesting visualization to do is how the food breaks out for the beef that you're eating and how that breaks out into corn versus grass versus soy and things like that.
[57:58.970 --> 58:00.410]  I think that's super interesting.
[58:00.410 --> 58:03.230]  I would like to be a platform that people would use for that sort of a thing.
[58:03.730 --> 58:04.430]  So, yeah.
[58:05.450 --> 58:05.730]  All right.
[58:05.810 --> 58:05.890]  Cool.
[58:06.010 --> 58:06.590]  Thanks a lot, everybody.
[58:06.910 --> 58:07.170]  All right.