WEBVTT 00:27.330 --> 00:33.090 Greetings fellow hackers, biohackers, and friends of Four Thieves Vinegar. 00:33.090 --> 00:39.130 This is Michael Laufer, recording for you from a secure undisclosed location. 00:39.130 --> 00:46.670 And don't worry, we'll actually have a live Q&A after this session, so I will be available and excited to chat with all of you. 00:47.170 --> 00:54.750 I am chief spokesperson for the Four Thieves Vinegar Collective, and I do want to emphasize the fact that we are a collective. 00:55.050 --> 01:02.730 Oftentimes it gets forgotten because the many people who do the work for this are in the shadows. 01:02.730 --> 01:10.470 They have to, for understandable reasons, keep themselves safe and can't be known to the public. 01:10.470 --> 01:19.670 But there are so many people who work on the Four Thieves projects, and there are so many teams who do so much amazing work. 01:19.870 --> 01:31.950 And I just want to emphasize that it really does take a village, and when you're going for a project like this it takes a lot of people with a lot of expertise. 01:32.030 --> 01:38.010 And I just want to recognize how amazing all of these people that I'm lucky enough to get to work with are. 01:38.750 --> 01:47.470 And maybe some of you out there will be inspired to come and help us, because you can always use more people with great talents. 01:48.090 --> 01:54.710 So I want to go through and talk about some of the things that we have coming up. 01:54.830 --> 01:57.310 I can't talk about all of them, of course. 01:57.470 --> 02:07.010 We have about 11 different projects on the boards, but I'd like to talk about three or four of them, and specifically how they connect and relate to our main mission. 02:07.010 --> 02:27.230 Our main mission being that we look to develop ways that people can get access to medicines and medical technologies without having to interface with the medical infrastructure, do-it- yourself methodologies, which of course rings with the hacker spirit. 02:28.070 --> 02:33.030 First thing I want to chat about is drug delivery methodologies. 02:33.030 --> 02:42.510 Now we often sort of forget about this, that somebody says they're taking a drug, or that some medication was administered to them. 02:42.510 --> 02:45.790 But the way that drug is administered we sometimes forget. 02:46.450 --> 02:53.310 Usually we think about taking pills, but there are so many other ways that a drug can make its way into your body. 02:53.690 --> 02:55.990 Oftentimes you'll get an injection, of course. 02:55.990 --> 02:57.830 There are different types of injections. 02:58.470 --> 03:01.450 We've seen, of course, that there are also different types of pills. 03:01.450 --> 03:07.810 There are tablets, and there are capsules, and there are time-release versions, and instantaneous-release versions. 03:08.470 --> 03:12.050 Additionally, you see things that are more advanced nowadays. 03:12.610 --> 03:14.990 You see things like transdermal patches, of course. 03:14.990 --> 03:21.470 Everybody's familiar with nicotine patches, but you see that also with nitroglycerin, other medications that make their way through. 03:21.470 --> 03:25.190 Even fentanyl is delivered in a patch for people who have chronic pain. 03:25.330 --> 03:26.910 Inhalers, of course. 03:27.230 --> 03:44.470 And so looking at some of these images here, I wanted to just take a moment to remember this really fun workshop that we did last year at DEF CON here in Biohack Village, where we were talking about hacking different drug delivery methodologies. 03:44.470 --> 03:47.850 And that you see that that's actually, of course, not an inhaler. 03:47.850 --> 03:48.650 That's a vape pen. 03:48.650 --> 03:52.410 But you can hijack a vape pen to work as an inhaler. 03:52.410 --> 03:58.250 And there are certain things that make their way into your body more efficiently through the lungs than they do through other means. 03:58.370 --> 04:04.890 Transdermal patches, when you buy them off the shelf, of course, are made with these specialized gelatins. 04:04.890 --> 04:13.670 But you can make them yourself using gauze pads and DMSO and some saran wrap and some medical tape. 04:13.750 --> 04:15.950 These things are all hackable. 04:16.710 --> 04:19.270 But these aren't the ones I wanted to talk about. 04:19.270 --> 04:22.090 There's a different one that is very exciting. 04:22.450 --> 04:29.910 So take a moment and see if this looks familiar to you at all in any regard. 04:29.910 --> 04:37.190 What you're looking at here is a cylindrical piece of flexible plastic. 04:38.530 --> 04:43.830 And what it is, it is a piece of plastic, which is biocompatible. 04:43.990 --> 04:50.670 And it's doped with the active pharmaceutical ingredient, which needs to be put into a person's body. 04:50.670 --> 04:54.290 Now, I'm sure some of you have heard of this before. 04:54.390 --> 04:58.690 In the 90s, the first generation of this was called Norplant. 04:58.690 --> 05:01.590 And it was a birth control drug. 05:01.810 --> 05:10.130 It's placed in the upper arm and you can see the more modern version that's called Nexoplanon. 05:11.630 --> 05:20.830 And it's being inserted with a hollow needle here in the upper arm of someone you can see in the upper right hand corner. 05:21.390 --> 05:35.690 Now, for those of you who are not familiar with this, or just seeing it for the first time, and those of you who've seen it before, take a moment to just look and see if this reminds you of anything. 05:35.930 --> 05:39.070 And let that percolate in your brain while we move on. 05:39.070 --> 05:42.010 Now, the question is, of course, how is this made? 05:42.170 --> 05:44.490 And how do you use it? 05:45.010 --> 05:48.350 Now, the uses are incredibly widespread. 05:48.350 --> 05:50.490 Now, it's been already used for birth control. 05:50.490 --> 05:54.190 And it's currently being tested. 05:54.190 --> 05:58.770 I believe it's in phase 2A trials for pre-exposure prophylaxis, which is really huge. 05:58.770 --> 06:06.630 You think about something that could be put under your skin, that could protect you from HIV infection for a decade. 06:06.630 --> 06:09.110 That's extraordinarily powerful. 06:09.110 --> 06:19.370 And this really applies to anything that is something that anybody has to take regularly, and where adherence is a big problem. 06:19.370 --> 06:28.630 Sovaldi is arguably the most incredible thing that's ever been put out by the pharmaceutical industry. 06:28.630 --> 06:32.910 It literally cures hepatitis C in 12 weeks. 06:32.910 --> 06:46.090 The problem is, is that if you do not take that pill every day, in the same way, at the same time, and you fall off adherence, it doesn't take and then you have to start over. 06:46.090 --> 06:48.890 And that's an $82,000 treatment. 06:48.890 --> 06:53.430 It shouldn't be and needn't be for reasons that you all know. 06:53.430 --> 07:06.870 But of course, it's tragic when you get it together to get treatment to somebody and it doesn't happen because somebody just forgot a dose one day. 07:07.650 --> 07:24.450 If this could instead be put into an implant, and somebody could get it put under their skin and 12 weeks later gets taken out and their hepatitis C free, that would solve that particular problem, then you'd get 100% adherence instead of only 40%. 07:24.450 --> 07:53.110 Again, looking at all manner of things, if you think for a moment about whether you or anybody who know, has any medication that they take every day, why take the time and the attention to take it every day when you could merely have something put under your skin and it would just sit there and you wouldn't have to worry about that autoimmune issues. 07:54.590 --> 07:58.770 Hormone replacement therapy is one that I'm very excited about. 07:58.770 --> 08:05.750 I know so many trans people who are so worried about having their pipeline interrupted. 08:05.850 --> 08:25.210 And if we could instead place hormones in an implant, which we know is possible because it was done with birth control and say, hey, you don't have to worry about taking your shots or interface with the medical infrastructure for five, maybe 10 years at a time. 08:25.210 --> 08:28.390 I think that would solve a lot of really amazing problems. 08:29.990 --> 08:35.490 The naloxone pro drug that we were looking at a couple of years ago would be really great to be able to put into this. 08:35.490 --> 08:44.070 This would essentially be an overdose safeguard antidote that you could put under somebody's skin and they wouldn't overdose. 08:44.070 --> 08:45.550 I mean, amazing. 08:45.550 --> 08:48.850 Even things as basic as vitamins. 08:49.470 --> 08:52.890 Most people are vitamin B deficient. 08:52.890 --> 09:02.110 And you could just have a vitamin B complex implant that would continue to dose your body with vitamin B around the clock for years. 09:02.190 --> 09:09.350 And again, anything where people have to take something every day. 09:09.350 --> 09:17.190 If you know anybody over the age of 60 in the United States, they have several medications typically that they take every day. 09:17.690 --> 09:27.690 And again, as years go on and days blur together, it can be harder and harder to remember to take something at the same time every day. 09:27.690 --> 09:31.430 And that takes care of that particular problem. 09:31.430 --> 09:33.950 So it's a very exciting technology in general. 09:33.950 --> 09:36.290 Of course, the question is, you know, how do we hijack it? 09:36.290 --> 09:37.570 How is it made? 09:38.170 --> 09:43.610 So the plastic has to actually be extruded. 09:43.610 --> 09:59.690 And so you have this big industrial machine that's full of the active pharmaceutical ingredient as well as the plastic and then it's pushed down through this screw and then heated along the way. 09:59.690 --> 10:01.170 And it's pressed out a dye. 10:01.170 --> 10:12.890 Here's a diagram that sort of shows that and you can see the spot there where everything's put into the feeder and that's the pharmaceutical ingredient along with the plastic. 10:13.870 --> 10:19.530 The plastic is usually PLA or PLA and another plastic. 10:19.530 --> 10:23.670 Again, some of you will start to see where this is heading. 10:23.690 --> 10:31.870 And it gets pressed out a dye at the end and then you get these little pieces of filament which are then broken off. 10:32.650 --> 10:35.230 And the question is, well, how do we do that? 10:35.230 --> 10:45.770 Well, this is the same as the 3D printing filament that is made by hobbyists all over the world. 10:45.850 --> 10:54.470 People who are enthusiasts of 3D printing, of course, take their old failed prints, shred them up, and re-extrude them. 10:54.470 --> 11:03.890 And if you can look at the image on the right, you can, of course, get ones that are manufactured in such a way that they're kind of a commercial. 11:03.950 --> 11:11.070 There are the kits like the filler extruder in the upper right and there are ones that are home-built like on the lower right. 11:11.410 --> 11:14.130 And this is basically all it takes. 11:14.470 --> 11:23.870 Concerns, of course, of trying to make sure the densities are homogenous and you don't get the active pharmaceutical ingredients separating away from the plastic is simple to fix at a small scale. 11:23.870 --> 11:29.150 You merely extrude the plastic, shred it again, and then re-extrude. 11:29.150 --> 11:44.170 And eventually you can have something where you merely load it into the same sort of thing that you use to plant an RFID implant into your hand and be able to put these in however you need them. 11:44.910 --> 11:55.490 Now, of course, the natural question that comes up is, well, then what do you do about getting the active pharmaceutical ingredient? 11:55.490 --> 11:58.870 And, of course, there are ways to source those in some cases. 11:59.090 --> 12:06.210 But one of the big things that we've been focusing on for a very long time is how do you manufacture your own drugs? 12:06.390 --> 12:12.390 And so I'm very excited to present to all of you the new Apothecary Microlab. 12:12.390 --> 12:20.570 And this is a picture of me holding the 3D printed case that has the brains of the machine in it. 12:20.570 --> 12:27.830 And you can see at the top there, there's the embossed logo and there's the touchscreen interface. 12:27.830 --> 12:39.330 And inside you have a Raspberry Pi and a shield that sits on top of that with relays. 12:39.330 --> 12:46.010 And it is connected then also to an Arduino which has a CNC shield on top of it. 12:47.010 --> 12:53.350 And I'll show you a little diagram for each of those sub-assemblies. 12:54.050 --> 12:59.790 Here's the Raspberry Pi and the relay board and then the touchscreen interface. 12:59.850 --> 13:06.030 And then this is a separate part, the CNC shield that sits on top of the Arduino. 13:06.110 --> 13:12.850 Arduino sends over G-code then sends information. 13:12.850 --> 13:30.710 So I put this together because the part of this graphic was actually put together by some yellow journalists who were trying to shame us when I was quoted as saying, shouldn't be any harder than putting together IKEA furniture to build the micro lab. 13:30.710 --> 13:33.170 And at this point, it really isn't. 13:33.170 --> 13:42.550 And the hardware team has done a just spectacular job of coming together with finding parts that are literally off the shelf. 13:42.550 --> 13:45.170 No manufacturing required and no soldering required. 13:45.170 --> 13:48.130 These literally snap together, which is so cool. 13:48.130 --> 13:50.630 It really means that anybody could build it. 13:50.630 --> 13:53.950 And here's a little diagram of how everything's connected. 13:54.070 --> 13:56.830 Those are stepper motors that drive the syringe pumps. 13:57.510 --> 14:01.230 And also that drive the stirring motor. 14:01.230 --> 14:07.150 And there's a sensor with the heat thermocouple and all bits and pieces. 14:07.250 --> 14:21.910 What you're looking at here is a video in real time of the head of the hardware team actually assembling the entire micro lab. 14:21.910 --> 14:33.650 And in addition to the fact that this is happening real time, what you see on the table here as he's working is everything that is required. 14:33.890 --> 14:43.850 And including the tools, I guess it's just slightly out of frame, but his three tools are a hex key, a small screwdriver, and a pair of scissors. 14:43.850 --> 14:49.010 And using nothing else, it takes him about 25 minutes to assemble this. 14:49.010 --> 14:54.330 So it probably won't finish quite by the time I'm done. 14:54.330 --> 15:09.570 But hopefully if the editing goes well, what will occur here is that as I go to the next slide, this video will go into a corner and you'll be able to see it continue in real time as I continue to talk about some of the other stuff that we've been working on. 15:10.590 --> 15:15.970 So you see these 3D printed parts that he's bolting onto the stepper motors. 15:15.970 --> 15:17.690 Those are the syringe pumps. 15:18.050 --> 15:25.110 The electronics that he snapped together just a moment ago was the control module that I talked about. 15:25.190 --> 15:34.170 And the circular piece that's 3D printed on the lower right of the screen that you see is the core for the actual reaction chamber. 15:34.170 --> 15:51.290 And then you see some miscellaneous parts along the left side of the screen tubing, the fish pumps that circulate the water, along with the bars and the threaded rod that allows the syringe pumps to work. 15:51.450 --> 16:02.670 So hopefully this will now move into the corner and you'll be able to continue to watch it while I talk about the other developments. 16:03.490 --> 16:10.150 Now, of course, the big question always is the chemistry. 16:10.150 --> 16:30.610 And this is where our chemistry team has done such great work in the past of trying to find ways to synthesize, to actually manufacture the active pharmaceutical ingredients from things that were easily available, cheap, having high margins for error, 16:30.610 --> 16:39.410 having no specialized materials needed or specialized techniques used. 16:39.530 --> 16:43.070 This was fairly tricky. 16:43.170 --> 17:08.310 And some of you, I'm sure, remember that in the early days, we had a deal with a company called Comatica that had developed some very sophisticated machine learning algorithms that would crawl through the databases of chemistry and give good guesses as to what might or might not work in terms of trying to do retrosynthesis on chemistry. 17:08.310 --> 17:18.590 Now, they, of course, were bought by Merck and that process is now proprietary, used internally, they don't farm it out, they don't share it. 17:19.130 --> 17:33.110 And one thing that we had this idea to do was to say, gosh, you know, we have a data science team, and they looked at some things and they said, you know, we could probably do the same thing, it wouldn't be that hard. 17:33.110 --> 17:38.770 And if you look on the right, this is the the NIST database that we had used with only 1000 reactions. 17:38.770 --> 17:40.790 And of course, it had the same structure. 17:40.810 --> 17:48.330 But Reaxis, the one that Comatica used, which is premium and paid for, of course, has 6 million reactions. 17:48.330 --> 17:55.290 Well, some of you remember the narrative and one thing and another, we got our hands on that data. 17:55.550 --> 18:01.010 And we're able to actually run an analog of that software. 18:01.890 --> 18:06.270 Now, the problem is, is you need a supercomputer for this. 18:06.550 --> 18:14.190 A very congenial fellow was working with me, and he tried to put together a Raspberry Pi cluster. 18:14.190 --> 18:20.530 And then when he looked at the software, the data science team had passed along, he said, Oh, this isn't going to work. 18:20.530 --> 18:21.930 And I asked why. 18:21.930 --> 18:35.410 And he said, Well, this is not something that lends itself to parallel computing, you need a traditional supercomputer, you need something with a tremendous amount of RAM, and you need a lot of cores running. 18:35.570 --> 18:41.990 And so when you need a supercomputer, the regular thing that most people will suggest is using Google Cloud. 18:41.990 --> 18:44.750 And I have some opinions on that. 18:44.750 --> 18:47.950 And then other people will suggest using Amazon Web Services. 18:47.950 --> 18:53.230 And I have feelings about that too, which I'm sure most of you share. 18:53.230 --> 18:59.530 And so the question is, if you need a supercomputer, and you don't want to play ball and pay for it, what do you do? 18:59.530 --> 19:03.790 Well, we figured out a solution to this. 19:03.790 --> 19:06.350 And we got very lucky. 19:06.870 --> 19:10.250 This is the Glitter supercomputer. 19:10.270 --> 19:13.690 And it's an HP ProLiant DL580. 19:13.730 --> 19:17.230 It's a 4U rack mount unit. 19:17.490 --> 19:20.910 And it's extremely powerful. 19:20.910 --> 19:32.030 If you look at the specs, this is a 40 core unit with half a terabyte of RAM. 19:32.030 --> 19:37.730 It's optimized in a lot of ways, it draws a tremendous amount of power. 19:37.730 --> 19:46.490 But we were very lucky that we managed to find one used for $300. 19:47.690 --> 19:52.570 Usually these go for upwards of $6,000. 19:52.590 --> 19:54.810 I mean, I think that's what the baseline unit costs. 19:54.810 --> 19:59.730 And I think it's more like, you know, double that for something with the specs that we have. 19:59.730 --> 20:06.190 But we just got very lucky that we found a used one that some company just wasn't using and managed to get our hands on it. 20:06.190 --> 20:21.390 In addition to running the software, we're hoping that this can actually become an open source supercomputer that everybody in the hacker community can use. 20:21.390 --> 20:28.770 So if somebody has something that they want to run, and they need a supercomputer for it, that they can just use some time on the supercomputer. 20:29.010 --> 20:32.770 So I'm gonna try and pull it up here. 20:33.790 --> 20:36.690 This is sort of what it'll look like. 20:37.150 --> 20:48.130 These are sort of mockups, but just so you can get sort of a sense of what things should look like when the machine is booting and running. 20:48.430 --> 20:50.670 Just wanted to give you guys a sense. 20:50.670 --> 20:55.530 And of course, you know, you want things to look pretty while they're starting. 20:55.530 --> 20:57.570 So if you'll bear with me here. 20:57.570 --> 21:04.890 Alright, so this is sort of what the should look like when you start things up. 21:04.970 --> 21:16.250 And when it says the smile string smiles is a it's a protocol for being able to use essentially ASCII to input the structure of molecule. 21:16.250 --> 21:18.910 And you'll see this looks a little goofy. 21:19.170 --> 21:27.610 But at the same time, you can see it's sort of suggested of of the structure of molecule that, of course, is a double bonded O. 21:27.630 --> 21:33.870 And here you have, again, another group that's specified and where it is. 21:33.870 --> 21:36.310 And this is all standardized in databases. 21:36.310 --> 21:38.750 So it's it's good, and it's easy to use. 21:38.950 --> 21:45.390 And if this is all work, then this should be go there. 21:45.610 --> 21:51.050 And you'll see the computer come up here now with recommendations. 21:52.370 --> 22:01.050 And along the right column, you'll see how likely it is that this is actually going to be plausible. 22:01.410 --> 22:05.470 And point 999, of course, is is really good. 22:05.470 --> 22:09.090 When you see the one that means they are actually found it in the literature. 22:09.250 --> 22:11.190 994 may seem like it's pretty good. 22:11.190 --> 22:12.970 But again, this is fairly iffy. 22:12.970 --> 22:24.950 The thing that's important to remember when looking at this sort of thing is that the really great strength of machine learning and AI is not to replace human intuition, but to augment it. 22:24.950 --> 22:46.130 So this is something that could change things so that instead of needing, say, oh, months to years with a team of really gifted chemists, you could have this with a good chemist, and just one of them and be able to do it in a few days. 22:46.130 --> 22:55.650 When you see these things that say cannot buy, that's part of one of the kinks we're working out is, you know, trying to show what's available commercially. 22:55.650 --> 23:00.530 And if something isn't available commercially, you'll notice that each of these are, of course, hyperlinks. 23:00.530 --> 23:11.070 And what you can do is you can actually click on them and then do the retrosynthesis on one of those and work your way back to see, do you have to make the thing that you need to make the next thing? 23:11.070 --> 23:16.970 And again, what is purity and yield look like and how you work forward from there? 23:17.470 --> 23:27.050 So to sort of sum up as we're going, you look for treating yourself. 23:27.190 --> 23:39.790 When you look for taking control of your own health, the thing to do, of course, is at the end is you need to package your treatment, right? 23:39.790 --> 23:50.250 You need that delivery mechanism, whether it's something that you load into a syringe or something that you pack into a tablet, or something to load into an inhaler, you need to package it. 23:50.250 --> 23:58.790 But of course, before that, you need to manufacture it, you need to generate that active pharmaceutical ingredient, somehow. 23:59.350 --> 24:09.950 And again, the packaging, we have these ideas for specifically, we're trying to develop these, these implants, which is really exciting. 24:09.950 --> 24:15.410 The manufacturer, we're hoping to make happen with the new version of the micro lab. 24:15.410 --> 24:25.910 And of course, developing a process for manufacture, we're hoping to do with the glitter computer and chemhactica, the new system for trying to do retrosynthesis. 24:25.910 --> 24:35.110 But there's one extra piece, which of course, before all of that, having all these things at your fingertips means almost nothing. 24:35.110 --> 24:42.190 If you don't know what it is you actually need to manufacture. 24:42.710 --> 24:47.510 This is sort of the antecedent problem that comes before everything. 24:47.610 --> 24:52.570 And the question comes, of course, how do you determine what it is you need? 24:52.570 --> 25:09.630 We all have health issues, but going from I understand what's wrong, or I understand what is not right, I recognize my symptoms, to understanding what is ailing you and what would ameliorate that is quite a jump. 25:09.790 --> 25:14.710 So I'm very excited to announce this that we're working on. 25:14.710 --> 25:32.610 This is the anarchist medical book, and we're trying to develop a compendium that will be a reference that people can use when they are not well, and yet do not wish to interface with the medical infrastructure. 25:33.290 --> 25:34.430 We're really excited. 25:34.430 --> 25:36.630 We have a bunch of things that are going into this book. 25:36.630 --> 25:54.870 And of course, to all of you out there, if you have things that you think should be included about this, if you know some cool trick, or some secret, some treatment that is not popular, that very few people know about, that other people could benefit from, 25:54.870 --> 25:59.330 please, please, please get in touch and let us know about it so that we can include it. 25:59.330 --> 26:05.090 We're trying to make this the best book we can, and try to include as much as we can gather. 26:05.090 --> 26:13.030 And so all the secret little folk wisdom of all the wonderful hacker anarchist communities out there, we'd really love to pack it in. 26:13.030 --> 26:15.010 So please send us anything you can. 26:15.010 --> 26:16.710 We'd love to see all of it. 26:17.570 --> 26:20.450 This is about where I end. 26:20.510 --> 26:23.230 And I always end with this slide. 26:23.790 --> 26:35.190 Because it's a, it's an important reminder that that helps not on the way that we really need to work to help each other, keep each other safe. 26:36.230 --> 26:41.010 And there's nobody but us. 26:41.250 --> 26:44.530 So let's take care of each other. 26:45.290 --> 26:46.770 Thank you all so much. 26:46.770 --> 26:48.770 It's been wonderful being with you. 26:48.770 --> 26:52.990 And after this video closes out, we'll have a little q&a. 26:53.250 --> 26:55.930 And I look forward to talking with all of you. 26:56.190 --> 26:57.830 Thanks so much 27:01.020 --> 27:01.520 . 27:01.920 --> 27:03.480 Rest in peace. 27:03.480 --> 27:06.140 Here's another warrior song. 27:10.000 --> 27:14.000 Here's a toast to the dead for children with cancer and AIDS. 27:14.000 --> 27:16.840 A cure exists and you probably could have been saved. 27:16.840 --> 27:19.400 Sad to see medicine divorce morality. 27:19.400 --> 27:22.200 Corporate homewreckers pipping up the salary.