Dr. Bruce Damer on Genesis Engines for Biota at Netflix HQ (Oct 28 2013)

Just another WordPress site

Dr. Bruce Damer on Genesis Engines for Biota at Netflix HQ (Oct 28 2013)

it’s working so should I stand up there or should I stand down here you don’t mind if I stand down here so I don’t accidentally pitch over the edge so this is actually the first public talk I’ve done in two years there’s a lot has been incubating and i’m going to introduce something old and then something new the old project is called the evil grid or evolution grid and the successors called the genesis engine and how does this fit a evolution in the cloud or conscious in the cloud this is a conception they’ve actually been working on since I was 14 years old in 1976 I was wandering around in the sagebrush behind my parents house in Kamloops BC Canada and I had this flash which was what would the coolest nerd project V to work on and I looked down at all the plants and bugs and things and well how on earth did all this get self-assembled to get going how did molecules get together and make this and that became my kind of life direction I started on a PhD in 1985 at university of southern california and i thought i had it all together i had a vax 11 750 and old-timers would only remember what that thing is a tektronix color display in an ARPANET account and i thought now i can build a hemodynamic simulator that might show emergence then I looked around there were no libraries no tools no field no colleagues no publications a few self-replicating automata books and papers by Russians and by a guy named Johnny von Neumann and I thought oh I’m too early I’m too early so 22 years later I restarted the project after hosting conferences on this theme and building a network and doing a decade of work with NASA I Sri started the project when there was sufficient so for all of you PhD students who are taking a long time there is hope and i finished the PhD finally from the original conception in 2011 that i’d started in 85 and had the idea in 1976 but it’s an ongoing story so there is hope for us us middle-aged PhD people so life intelligent or not how much if any is out there one of the mechanisms that has been used to ask that question not necessarily answer it is the famous Drake Equation everybody knows with the Drake Equation here it’s a set of probabilities of you know rate of star formation fractional the stars that have planets etc etc and you add them all up for a galaxy and a length of the last ones length of time such revelations a release detectable signal so it’s a it’s a by frank drake an idea of how many signals are we going to detect on our radio astronomy gear and you can vary all these probabilities all you want I chose to expand upon you know the function l lu PL there the fraction of the above that actually go on to develop life at some point so how many life generating places are there in the universe or in a given galaxy and so expanding upon the Drake Equation I came down to if you’re looking at chemical life if you’re looking at aqueous chemical life what are the factors that have to be in there so this is a miniaturization of the Drake Equation into this one function what how it has to be there in any system so that life emerges if something is missing life won’t emerge so what are the minimum what is the minimum that you need and I’ll come back to this so this is just a little little preview of this concept so back in two thousand eight I with the team our NASA team we had a little extra time so we produced this fun YouTube video which I’ll show you which is the concept Shin of what if you could use computers to do a gigantic simulation the chemical simulation in digitally to then sweetness with all the properties in origin of life inside the network now this is a great theme in science fiction artificial life people working on this you know great science fiction writer writers like greg egan and whatnot this is a common theme and then they’ll take the next step since you can’t simulated

the chemistry pretty faithfully could you then fabricate that life in some kind of 3d printer and so that was this idea so I’ll roll this fun movie for you the other it I’m gonna school you an illusion she here we see a good simulation you really process and it’s behind Markel’s of you parts of retarding the inner half and as something to da beleza processors and in a Dublin hand and I’m starting to see some Anakin 2d phenomenon so let us die in and CBC’s coming Anthony and it looks like that our loser programmer reminding something interesting we’re not a little bit of salt organization is Anthony here oh that’s it’s gotta come inside this to go on a few billion generations and we ask something more interesting why are you vesicles yeah look at my charger in this one and building cherishes later we’ve got a little bit angry on some more of anarchy and every time I said it either let’s push this long and put it into our scam scam taken apart diddly diddly all being coerced and build a little casual about data now I’m going to drive Minotaur and if you have are happy to have a fantasy is co2 ammonia and go to water issue oh and instead of a family to our digitally altered into the original chemistry maybe she’s Sam screams in reacting parts and channels are creatures finished being credited and at the end of the last you will come our fab empty ready for a swan dive into a deeper in the years are related like original tempo as you are recognizing the chemicals is a running algorithm stands when I glasses seems to be operative the start a new life in the world of the molecule this was the conception that we put out to the world about a pathway to a second Genesis of life you know starting in the digital domain and then the next one I went to talk to many people about this so one of the people I went to talk to his Freeman Dyson at the Institute for Advanced Study and I’ve been going to see him now and then to show him progress on the work it has comment the simulation should be truly messy ie nature’s not clean and neat as you are showing in the movie cells are more like dirty water surrounded by car surrounded by garbage bags and of course if you read his ideas about the double origins of life in his book you’ll you’ll understand what that means and many other comments so for example while I was at the Institute I went into the archives and I went into Robert Oppenheimer’s files about the electronic computer project at Princeton which produced this machine which is was widely copied and may have been the first true digital computer had registers and serial processing of commands and secondary and tertiary memory and this is john von neumann standing next to it and this machine’s i went through all these files and find out this oh this is a contingency design this was just to get something that worked for more than 30 minutes at a time and actually could do real things and we’re stuck without architecture it turns out that that architecture is really poor for doing nature really poor in fact it’s of the seven properties of computing systems that Peter Bentley at UCL University College London lays out and the seven properties of natural systems none of them are come our work and so it’s like computers are incredibly bad at doing natural systems so back to building life are you going to you so in some sense you can’t use computers to simulate chemistry to any degree to help you with origin of life stuff you just can’t wear four orders of magnitude off we may no real to do it so

building life the hard way in chemicals what if you could do chemical experiments all at once but marry them with computers this is the famous miller-urey simulation of the reducing atmosphere back in 1953 showing the amino acids that they produce neighbor news 21 people didn’t realize at the time they looked at they the vials today and he actually produced a huge number of amino acids more than they found in the 50s this is Steen Rasmussen in Denmark proposing radical new ways of trying in chemistry trying to create a chemical life cycle that would create a kind of a Genesis but this this architecture turned out to be too radical to even be doable in the first stages in the lab and the project lost you know it finished its funding I don’t think with any significant progress but this is the equipment they’re using for their chemical origins of life this is the kind of gear you use not you’re not sitting at a screen you’re doing basically glassware work and running things are very expensive tests and missing things most of the time so this is the process by which one lab was attempting attempting to do it so how easy is this going to be how easy is it going to be it turns out it’s incredibly hard individual origin of life team’s focus on one narrow area like polymerization or finding metabolic cycles or working on vesicles and compartmentalization they tend to work in very very narrow areas with experiments done by hand like this guys doing and penny Boston commented she’s a well-known cave explorer looking for deep in the earth and she she commented that you just have to find a model that we’re in chemistry you can support ever-increasing levels of emergent complexity and that nature had this computer to do it with you know and had at least several hundred million years even if it may have gotten the through panspermia the organisms may have come from elsewhere but it had an awfully big system to to make an origin of life happen so here you then go down in the sort of craig venter mode and say what’s a minimal cell that actually works what parts do you actually have to have and this is a one bottle of it it’s a lot i’m moving parts it’s a heck of a lot of moving parts that have to work with ninety-nine point nine percent 99 percent accuracy for this thing not to fall apart and do the very difficult job of mitosis of dividing the extremely different technical job of dividing so how do you get from basic simple mixtures of atoms to something that’s like the Queen Mary of complexity the simplest cell is still a vast machine with a hundred billion carbon atoms how do you do it how do you map that computer onto this one you know there’s a computer from my collection a commoner pip many of us grew up on these they our computers aren’t much more powerful than that in relative terms so in 2008 through 2011 our team built a distributed artificial chemistry simulation called the evil grin and we first ran it in my barn on basically old servers until we burned every machine out and then UC San Diego donated a rack of about thirty cores for us to complete the work what we did was we said all right since you cannot you have to simulate nature to it’s true in a faithful way you can’t just do a conway’s game of life you have to simulate atomistic environments and as as faithfully as you can and we chose the simplest one which is interstellar chemistry cosmochemistry where you have a lot of free atoms move and you have Isis and dust and everything to you can’t some like those are too big but a domestic chemistry is a very simple soup very few reactions happen in an interstellar chemistry but enough happen over hundreds of million years to make the organic molecules that you are made out of so by the time solar systems accrete they’re crammed with all this stuff I mean you can you can look at the sky and you can see all this material gradually forming over time so our whole goal in the evil grid was to see in an atomistic soup how does nature wiggle along in this extremely noisy environment of just atoms going every which way how can you cut through the complexity and the noise to get accretion of molecular bonds faster and faster because otherwise none of these experiments are going to finish in a human lifetime so this was the result and it turned out to you know I allow me

to finish the PhD finally and this chart the second chart here showing these stair casing was the signature and experiment number six that ran for about 40 days non-stop will we found that nature uses or at least in this simulation uses some combination of stochastic hill climbing to get around in those hilly landscapes and to form bonds and to keep the bonds formed and stochastic hill climbing as you get up to a maximum which might say I’ve got three bonds form and then you wiggle off of it and you try a thousand or a million more pathways finding a ridge to the next complexity which has four bonds form hallelujah and then it Wiggles off and down and it’s that you see that happening in evolution you see that happening in social systems in economic systems it’s stochastic hill climbing and it worked in really the most noisy environment we could find to simulate in nature which is a atomistic suit so here’s experiment number one just a few products after months of running here’s experiment number six rocking number 141 molecules formed in their incredible is a full order of magnitude so the next step was and I called this kind of the cosmic legal the wiggle that nature takes through space to accrete complexity and hold on to it so but then I was sitting on a park bench in Montpelier and zan and I were at this conference and I was depressed because I thought computers are just not up to it there is not going to be able to simulate enough digital stuff to show us how life may have started and this whole thing is a bust I found an optimization method but in my lifetime I’m never going to see it and then popped into my head was this vision of millions of chemical experiments millions are trillions of them somehow controlled by computers doing the same thing as we were doing so that the molecules do the walking they simulate themselves and the computers are the eyes and the selectors and the scores which they’re good at they’re not good at simulation and they help run it very much like PCR very much like the Human Genome Project very much like what’s called combinatorial chemistry so was born this concept of a Genesis engine a hybrid between computers and chemistry that is tuned also toured the human mind the designer of experiment who has teleological angles which may or may not be met and I drew some really simple diagrams of well here’s your chemical experiments on little shelves and and these little shelves could be controlled by a laptop and then some shelf is heated and some is vibrated and another shelf is has more things added to it now these are just little balloons I mean this is a very cartoony conception and then you might combine the results of experiments but you could you could monitor the best-performing little balloon put stuff in and out of it and wait for it to become a richer mixture and then combine it then go on to more little shelves of balloons and here’s a more logical you know here we have a set of millions of experiments around catalysis the millions around vesicle formation and millions of them around RNA formation and they’re all going through these gates and they’re going to big soups like these were the best ways we found to make long-chain RNA and here are the best ways we found to make balloon or bubbles out of lipid and here’s the best ways we found to create a metabolic cycle and we’ll dump them into a big soup in and see what happens coming out of the out of the big soup so this was this jet idea of fully automating origin of life experiments across all the experiments chaining them together and then looking for the results and actually built one of these shelves in 2011 22 see how we could do it on the total cheap or 80 bucks it turns out that at the same time Dave Deemer who’s Mike my colleague at UC Santa Cruz which I’ve just joined as of last week built something far more sophisticated and this is his their first step at doing a Genesis engine and hear what they’ve done is they’ve and this is a big thing for argent alive people they haven’t done automation before they haven’t done architectures before it’s all been pretty much by hand there been doing 19th century chemistry basically but the revolution in from a symmetric sand from combinatorial chemistry is starting to come into the origin of life field which is extremely underfunded field I mean these people are doing this on as a sideline mostly there’s very little funding for this but they board well

plates well holes in this aluminum plate found a forty dollar stepper motor to drive this and then they could with one set of leads they could place water containing lipid phospholipids with another set they could blast them with co2 to evaporate the water and force the lipids to go down into these liquid crystal matrices between which are captured basically nucleotides which then form RNA and then they would rhian and you’d see these long chain RNA molecules and then they would do it again this is stimulating a water geysers that are all over the world land-based geysers in Kamchatka and Mount lassen but not where the geyser shoots up and sprays amphiphilic molecules lipids onto rocks they dry out and you get these wonderful layering and in between are these wonderful factories to make these informational molecules and it’s been days thrust of his entire career is to show how this works so this is his first automated way of simulating a hot spring now there’s no computer watching this there’s no computer that’s saying oh I like that experiment in the 15th well plate I’ll start a million moral as you can see there’s very few well plates but it’s a very it’s a furnace the start to building a Genesis engine so back to the formula of the formulation this is now evolved into and this was developed with working with Dave Deemer and Stuart Kaufman and a number of others we asked the question what must you have in your origin of life experiments such that you could actually get a second Genesis there’s enough stuff there from which would emerge a prebiotic form that was very compelling and we kind of worked out that there are these factors and just briefly so the factors think of them like black boxes in a computer architecture you start with a mixture of simple kind of molecules floating around you run them through time through through all of these things churning away and it yields and what what I call a milieu which is a much more sophisticated mixture which has biological behaviors it may not be alive but it’s up the chain so you could think of it as this kind of a flow diagram and what are all these factors and I won’t go them in in super detail but it’s some of the obvious ones you need an energy source you need something to cycle the system you have tides you have heating and cooling you have all that you need something to concentrate chemical so they’re not diffusing all the place and they can get together you need something called condensation which means water goes out so the bonds can form the important bonds and informational molecules if you don’t have condensation you’re never going to get those you need catalytic action which is like a baguettes B which we get c which begets a so you can produce lots of a’s in a catalytic cycle that’s called metabolism that’s called producing products you need some kind of coding you need to be able to form informational molecules that’s coding you need encoding all over the place compartmentalization things need to be in cells I need to be in containers otherwise they just diffuse into the bulk and they’re gone you have to have them you need and this is the terms that I brought in for my PhD work you need combinatorial action you need to try lots of solutions over millions of solutions over and over and over again you need some kind of climbing mechanism this is artificial evolution this is human beings coming in saying our machine has just seen a experiment number 1 million such-and-such which we really like so we’re going to make a trillion more of them and run them and look for the it’s still nature doing the emergence because nature produced the results you like your then allowing nature to track a vector to a trillion more experiments that you’re going to look at and you could be completely wrong but it’s a way to do climbing and one of the book authors in the book that I’m putting together called Genesis engines work at IBM just over here on net millipede memory and probably none of you remember this project millipede memory was where they would take atomic force microscope tips and they would move along and they would read and write memories a good idea right it was the highly dense memory it failed but one day in the lab as they were trying to get this thing to work a bunch of

moisture got in vapor got into the the box and suddenly there were droplets hanging off the end of the FM tips and when they would put them down the drops it would stay and they had a whole field of beads perfect you know spherical little beads and one of our offices remembered this is being one of the problems and then when we signed him up for the book he said wait a minute we computed we could get a trillion of those globules per square inch and we could we could move them around we could combine them we could put stuff into them we could pick them up and go and take them places I said you just created that’s a system of revolutionized chemistry in the 21st century can you imagine and you can wash them away so that has become another possibility out of this project doing a trillion experiments at once per square inch inside one of these machines the last of course is evolution the Zen was talking about did evolution come before life it probably did their chemical there’s a whole Center for chemical evolution at Georgia Tech that just studies chemical systems that exhibit Darwinian natural selection so if evolution kicks in the sooner it kicks in into your chemical system the more you’re going to drive toward innovation and there’s whole teams that are trying to figure out how can evolution start before life it kind of can it can be in in the system so just a few more slides we then worked out in any experiment what do you have to have what is the lowest level you have to have energy and you have to have cycling and concentration of some sort if you want to form molecules of any productivity condensation and then these are kind of all optional because most people don’t have the budget or the time to do the full monty to have a system that has all these properties so most of them are doing a few of those and their pick a few here from here from here from here and there they’re publishing the paper the concept we have is to go back to the origin of life field and say this is the totality of the space we’re working in and that any paper can be described any any research project can be described based upon what does it implement from this stack so for example you can take these terms and do fun things with them so example compartmentalization promotes condensation and concentration which thereby promotes coating the forming of polymers which in the coding promotes this as well so you can write these funny terms that show how everything into relates back and forth you want to use these terms and this applying this to Dave demers work at at nasa astrobiology and at UC santa cruz we wrote this formula this completely describes his experimental framework and how it works this one little compact term so a combinatorial action because he’s doing the plates rotating around on compartments which are these crystal matrices subject to energy cycling because it’s getting wet and dry and wet and dry promotes concentration and condensation blah blah blah blah yielding RNA strands so this one little term is a way to concisely describe any of these class of origin of life and probably most chemical experiments but focusing on origin of life another way of showing it is like well this combinatorial cycling helps compartmentalization which thereby helps these three things so it’s another way of graphically showing it another way of showing it is you start your experiment with some building blocks and some compartments and then halfway through you’ve added energy in cycling and you get increasing complexity and then your end product is double-stranded RNA another way of showing it as a cycle so really a versatile language last experimental show you this is a very famous one from Bartel and Shaw stack in 1993 and this is chemical evolution this is running ribozymes trillions of them through a column through beads and through cooling and heating and whatnot and you find that the ribozymes evolve and you have productivity in the ribozyme 7 million times more than if you just did a straight experiment and so we can’t with the compact description of that experiment and so in the future you’d have whole flows this is a blueprint for doing a Genesis engine so phase one of the work now which is going to last probably 50 years is to get this blueprint language down get it used in

the origin of life field and then there’s a project at McMaster for a million dollars they just received to build a Genesis engine or a piece of it if I go there and we can convince them to use this nomenclature everybody be stalking the same way writing blueprints just as von neumann’s blueprint for his computer was used by everybody it will have a common blueprint and that maybe in 10 years 20 years 30 years 50 years there will be some master diagram somebody puts together that they then build the hardware and they put together a true primordial soup comes out of the thing and I wrote this this nice ode to the Genesis engine Oh Jan this is the final slide Oh Genesis engine you great Rube Goldberg machine of the 21st century resplendent with all your pumps piping chemical soups around your computer eyes scanning for signs of complete competing lines of polymeric infused vesicles and your purring grids of silicon modelling yields and then selecting experiments to be robotically receded and inside of you one day perhaps decades hence alarm will sound in one lone experiment within your millions of distributed chemo grids a sample will be rushed for analysis and scientists will emerge breathlessly declaring that a second Genesis has occurred or rather is in the course of occurring if time and budgets permit you running you for another thousand years you will leave us all wondering what it all means but it will mark a major moment for our species as powerful as when our earth was first photographed from space for thanks to you we will know that we are most certainly not alone in the universe and in some sense we will have made first tagged and I’d like to thank the organizations and people that help support all this work over all these years and one last thought closing thought and thank you very much thank you Bruce any questions for Bruce read the Seas referring to t la tele ology and I wondered what you meant yeah the hears that the tricky thing teleology is sort of you kind of know the results you want so you design something that gets you toward the result you want the problem in emergent business especially in chemical emergencies it’s not about being trapped because if you said I know how I started it it’s lots of little balls floating around and this is what trap some scientists in the 1960s Sydney Fox being among them because they saw these balls moving around and fulfilling balls and they would divide they were like or so darn close so they focused all their experiment but nobody understood the machinery of the cell at that time so they didn’t realize they had Mount Everest to climb first the balls were like the minute the you know hardly even the first step so their entirety biological framework was get all these balls to do their thing but none no understanding of the mechanism inside so if you design a Genesis engine that sole focus is to get lots and lots of vesicles to form you’re gonna have a great vesicle forming thing but not a living system when it come when you come out of it and so what kaufman stearic hoffman has a fantastic new theory called enablement in the biosphere he calls it no entailing laws that as soon as life cranks up physics you can kind of like all right planets are here now we can use even Newtonian physics to figure out they’re going to be here you know in on Tuesday of 20 23 and a certain month they’re going to be there you can’t do that with biology as soon as a biological form in an ecosystem that subject to our any natural selection is around it takes all these crazy directions that are unpredictable in any way they’re never predictable there’s enablement to what he calls adjacent Possible’s a good example one example that Stewart uses all the time is the lungfish the lungfish is this was our fish back in the devonian or some period where would come up to the surface of the ocean or rivers or whatnot and it would suck in air it had a lung to breathe with so these lungfish were doing pretty well they weren’t very dominant but suddenly one day there was a mutation and the lungfish was unable to expel all the water back out because that’s how they it just had a problem with a flap or something so you thought it’s a doomed lungfish know what happened is the innovation that happened as a result of that exploded across the earth within a

couple of million years because the long fishin’s would try to expel all the water couldn’t in like oh float back down and then realize you know as if Ilan fish can realize that it’s now neutrally buoyant it’s got just enough air and just enough water that it’s hanging in the water like I don’t have to flap anymore to keep from floating down I’m affecting this chill I can chill and so less energy was used so this lungfish was incredibly successful and within 23 million years the swim bladder was born neutral buoyancy had never existed in the known universe it was a new physical property because you can’t have neutral buoyancy inside a star and a dead planet not only exists within a living system so there’s a whole new property of thought of a biological physics that just exploded unpredictable unplanned and this is how evolution goes and so that’s stewards so in in the design of Genesis engines we have to actually make them flexible enough and malleable enough that these molecular systems can go crazy directions within the system and take us in places we just can’t design with our you know puny monkey brains that’s the answer the teleological conundrum any other questions mark so you have you had a slide that showed a diagram of multiple experiments that you were then combining the results from can you put that do you remember where that would certainly can oh I’ll go back to oh it’s further down oh did I just go to the start yeah this one was down in here we are coming into their ear that was so I have two questions about that um the first is if you were to do that how much of the game would you be giving away by giving it that structure by setting up that structure from the beginning mm-hmm in the experiment that’s question number one in question number two is how how much can you use the examples the example of life that we have in front of us to guide the simulation so let me let me ask a version of that question so you’ve abandoned using digital simulation mm-hmm so the foundation of it but imagine that you were before you had shifted to this sort of hybrid system combined system how far are we from being able to simulate not evolve a cell but even simulate a cell know how many how many generations oh it’s so far off well in gym in generation human generations so get an example Anton which is the super computer built by David Shaw in New York he was a he’s a financial bad guy gone good he helped create what was it they not the junk bonds the hedge funds in the 80s but he was a science guy from Columbia so he sold his business for billions of dollars kept his building in Manhattan and packed it with scientists and they built the world’s greatest protein simulator it has like 10,000 or 20,000 d6 and it’s a huge cube and and they announced as a couple years ago was like a folding of a protein having 5,000 atoms doing a 2 nanosecond full sequence you know with the bulk phase with water bouncing around on this is in simulation and in tune a sex they didn’t only three weeks of flat out calculation three weeks and there’s 30,000 kinds of enzyme proteins in your one cell on human 30,000 types and so they’re so happy about these little point successes because you know one protein folding me done in a reasonable time and characterized the protein without denaturing it and that’s the best we’ve got and so a living a the bulk phase of just the water or minerals a bunch of amphiphiles enough enough of a sort of marine environment or a coastal environment or a hot spring environment to to simulate digitally to have any kind of regional complexities is a billion times bigger than ours is

so much bigger that’s why I threw my hands up on the park bench in montpelier like it and why why do it when we have perfectly good chemicals that are willing to do it in real time for us how much does this sort of structure focus an experiment look this well ironically on thursday i met with with dave and we started to think about this now that we are coming up with this language and he said here are the four experiments that i would chain together from these for Shaw Stags group this one from Gerald Joyce this one guided from road a in Germany was doing using salts he says they they tie together you can have the outputs of this one go on the inputs of this one and so it started to make sense to us that this this is even plausible but you’d have to put them in one machine and that machine would be so doggone flexible could run any experiment and so it would have you know it would be able to reconfigure itself to use as its arrays to do in numerous types of experiments and one of the concepts is if you could get it so Dave’s experience with this because he he made a discovery in the 80s that led to nanopore sequencing so in the 80s they were in the lab and they had a membrane and there was pores in the membrane and suddenly a bunch of DNA that was on one side of the membrane was on the other side it just got there it’s like instead of saying well just wipe the slide there’s no no this is amazing and it turns out the ends like them the DNA had found itself up against a poor and had corkscrewed its way through and what they ended up designing was a clicker a molecular clicker mechanism that would change potentials as different base pairs went vinyl click click and it would count individual base pairs and last februari net Oxford nanopore in Florida at a conference held up a box saying this box will sequence the human genome in two hours using nanopore sequencing and that’s all coming to market now so the idea is to build that kind of stuff into this know how many generations are we talking about the yucky that if you get any kind of polymer you can have a box in there that’s really small that will do the immediate counting of what you just got so rather than doing mass spec and TLC plates and all this you can do direct counting of the molecular contents really tough stuff I mean that the detection alone is hard but the goal would be say by the year 2050 you could have a box you could put in your basement pour your coffee into just in good old back to the future as coffee as everything you need takes it all apart and then it would do a little piece of the origin of life project which would then publish to the internet and then other Genesis engines would pick up and carry on bigger units so you could have distribute this all over the planet and plug them into different wall sockets and do massive distributing kind of Seti type this is City in Reverse it’s trying to to show how life might have started one of the long-term things if we get really good at this in 500 years we’ll be able to have a Genesis engine this goes back to some of your work at NASA and say okay we’re going to now sim we’re going to put in not a simulation but a piece of Marcia and ice cap you know we’re going to make it we don’t have to go get it and bring it back and sample return but we’re going to we’re going to now take a million pieces of Martian ice cap and we’re going to run this thing to see what life could have emerged in such environments so it’s like a telescope or a microscope on where in the universe good life actually come and then we got with a gradient of 0 life could be quite common because we found it emerges quickly in these environmental slower in these environments and not at all in those environments that it’s a crazy idea and then flipping it around say well given that we just showed genesis in a piece of frozen carbon dioxide or water for mars and an organism is now happily tolling its way through that piece of simulacra we can actually ship it to Mars and it’s totally at home it’s done the terraforming thing we just evolved organism after organism to do to make habitable places and any civilization that’s worth its salt probably is harness the power of evolution to do just that you know they’ve got the best nerds working on any more questions for Bruce well thank you for us of them