[Music] welcome everyone to the third annual stern innovation conference this conference is hosted by nyu stern's fubon center which was made possible through an endowed gift from one of our alumni richard ming zing zai who's the chairman and ceo of fubon financial holding company the fubon center serves as the school's hub to support facilitate and enhance cross-disciplinary collaboration among its existing areas of excellence including fintech data analytics and artificial intelligence and technological innovation it's designed to bring together scholars managers and students to conduct and disseminate world-class research on fintech tech innovation data analytics artificial intelligence and data science for business in addition to the annual stern innovation conference we also host an annual fintech conference and several speaker series on fintech data analytics artificial intelligence and technological innovation so please check out our website and when you get a chance come join us for some of our other events we have a really exciting lineup today we have some of the very top scholars from around the world on remote innovation and collaboration and they're here to tell us about their latest research findings we also have top managers from companies like microsoft and paypal here to tell us about what they have learned about remote collaboration and to also tell us about the latest technological advances for achieving it so without further ado i'm going to start introducing our first speaker our first speaker today is professor hila lifshitz assaf she's a professor here at new york university in the department of information operations and management sciences she's also a faculty associate at harvard university in their lab for innovation science professor liv shisoff's research focuses on developing an in-depth understanding of the micro foundations of scientific and technological innovation and knowledge creation processes she explores how the ability to be innovate is being transformed as well as the challenges and opportunities the transformation means for r d organizations professionals and their work her three-year longitudinal field study of nasa's experimentation with open innovation with online platforms and communities received the best dissertation grigor mclelland award at the european group for organizational studies she also it also earns the best administrative quarterly asq paper based on a dissertation and the best published paper in communications and information systems at the academy management so a lot of awards there she also investigates crowdsourcing open source open online communities wikipedia hackathons make-a-thons her research has received the prestigious inspired grant from the national science foundation so i consider us extremely lucky to have hila here she's also a very good friend i feel very lucky to have her here today and i'm going to hand it off to her in one second i just want to remind everybody we will have a q a session at the end of hila's talk so go ahead and drop your questions into the q a box at the bottom of your screen and thank you for joining us hila without further ado thank you so much for the wonderful introduction it is an honor to be introduced by professor melissa schilling and everything is heard well and seen well looks good great so we'll start so uh i will kick us off with a big question for this beautiful conference today about the future of innovation processes so that's the question i want us to think about together i'm going to share with you new research and i welcome questions in the q a part s melissa suggested please engage with us tell us what you think this is new research i will try to answer in the q a and if not this is my email this is my twitter account you can find me on linkedin on via email and i would love to engage and answer your questions so what is the focus of today i would i like us to rethink what is the role of experts crowds and ai for innovation and what starts with experts since historically this is how we start if you think about the traditional innovation processes they're all built on the assumption that we need domain experts this is why i am here this is why professor schilling is here we assume that if we get phds and we become in-depth expert in a specific domain then we have the ability to innovate to create new knowledge this is how universities are built this is how r d organizations are built incentivizing people for publications for patents giving noble awards this is all built on this assumption and this has started way back in the 15th century if you think about it this is the lone inventor myth and of course it has been updated not fully uh but it has been updated and we have been talking in the last 200 300 years let's say about the fact that it's not just one person who does the innovation we need multiple people preferably cross-disciplinary innovation coming from different experts and then more recently a lot of talk about the collaborative diverse way of organizing for innovation as the recommended way but still within the circle of experts right and within the boundaries of organizations organizing with experts in them to innovate and then a decade ago here comes everybody from everywhere right the web has emerged open source open sign citizen projects collective intelligence scholars academics practitioners are saying why do we have this model we don't need it anymore anyone can innovate from the tip of their fingers with their phone from anywhere let's forget about that old model and we don't need it anymore let's bring the crowds in so some people talked about it as crowdsourcing others about democratizing innovation it could be experts but not necessarily domain experts it can be anyone from anywhere super minds the collective intelligence power the main point here was both about everybody instead of experts and the second thing is everywhere which is the topic of today's conference virtually we don't need to be co-located this beautiful image of sitting around the table getting the brainstorming archaic that's what most of these people said let's just be connected on an online platform an online community and that can work what i want to suggest and to remind us that many times when we have these new ways of organizing we kind of say that the last way is actually not that good right and in that study that melissa described that i've i was conducting for almost three years at nasa this was one of the underlying assumption that i realized that actually took place so this was a longitudinal study of three years before throughout and after their experimentation with this virtual distributed online innovation this is published of course and what i want us to take from that study briefly is that it did produce the scientific breakthrough and a lot of innovation in speed neck amazing speed three months for you know an r d project that usually takes five to fifteen years a couple of three thousand thousand dollars instead of eight hundred thousand dollars so really big difference cost and time-wise but it also created a lot of resistance and those that wanted to adopt this new way of working the experts that wanted to transition had no clear path they had to struggle and to go through a deep transformation that i discussed in the paper but what i want us to discuss for today is why did they had to go through this deep transformation it was because there was this underlying assumption that they could be substituted by the crowds they are no longer needed because voila this person solved a nasa challenge that the nasa experts could not solve in three months in 30 000 so why do we need you so no one said that explicitly but the open innovation online platforms that was the message and the expert because no one said it explicitly we're on their own to interpret and i'm saying today let's bring those assumptions explicitly and start talking about them and try to think in a way that will augment will enhance the overall process of innovation instead of having all these you know hints of threats every time we move to a new type of way of innovating and what is of course the new type now why do we need humans here comes ai right so 10 years later 2021 who needs humans to innovate let's delete us all and the machine can innovate instead of us and it may sound a bit humorous but this is serious so of course uh we can think about it and talk about it and debate this academics but this is already happening companies are already investing so the image i brought for you from you with this kind of you can see the robot looking at mathematics this is from an ai mathematics project that there is now a paper in mathematics claiming a discovery 2ai ai in pharma in drugs new drugs will be going through fda in the next couple of years that were invented by ai we have art exhibitions and sales around creativity on the left-hand side the artificial inventor project i would recommend to check last week in virginia a court ruled that we cannot have an inventor a patent inventor listed as a.i there has to be a human but in australia it's different so there's a global debate even if ai can be listed on a patent as an inventor and then we have alpha thought that we we cannot ignore the largest biggest most significant scientific innovation that took place in the area i would say of science in the last few years made by a uh and now that their way of organizing it is not to have a patent for it or not to have it for themselves but to open it through embell and other organizations to have a the largest uh data repository open for anyone so different ways to organize it but still i want us to ask the question on the underlying assumption so are we again bringing the same story of automation substitution or can we think about it differently and maybe think about how to augment and not to threaten the existence so this is what i suggest for today i suggest asking what each agent entity actor brings to the innovation process experts what do they bring crowds and ai we have a first a paper on the topic that i've authored with nikki couture daphna shahaf bob kraut and many other great scholars from hci human computer interaction discipline and computer science discipline a cross-disciplinary project yourself with many phd students discussing how can we scale up innovation based on the analogical perspective of innovation with crowds nai what we suggest there is to look at the main attributes each of those actors bring humans are the experts such as me can bring complexity the in-depth knowledge they've acquired but they are fixated sometime on how they think right we know that from research so we can get that open different way open that fixation from the crowds from different individuals experts or professionals or not professionals even that are from different domains and look at the problem in a fresh new way such as we saw at the nasa breakthrough and then comes ai that has the ability to offer scalability on a whole different dimension that humans can if we think about ai and drugs etc and new materials the ability to scan and read literature in an unprecedented volume so let me tell you one study that we've conducted in that spirit in that mindset of how can we augment experts instead of replacing them for ideation this is a study with nikki couture bob kraut and felicia nang from carnegie mellon in collaboration with industrial research institute which is a great umbrella institute for many companies so these are some of the companies that tightly collaborated with us bosch hershey's and i invite companies that are in the audience if they want to participate in such field experiments to approach to us so we started the study by listening to the experts so we talked to them for a year in their meetings in their conferences and what we heard is that they are thirsty for something that will help them break their cognitive fixation in our language in their language they talked about this solutionism this short sighted way of looking for a solution that is the closest to what you've done and they want to think differently this is why they wanted to be innovators and to work in r d but it's very hard their own way of thinking has been habituated and the organizational pressure that they're constantly in and then when we asked them what could you imagine that with today's technology can change that they said we don't know because all these enterprise systems that they bring us are so convoluted they're not intuitive we want something that works the way you know our mobile phone apps work something very simple very intuitive that can help us maybe break this cognitive fixation and since my collaborators to the projects are hci scholars which means they know how to design things in a way that makes it easy for humans to work with what we came up with was what if experts had an inspiration engine something that would augment the expert in their ideation process that will enable them to think maybe differently but will be something intuitive and easy on their mobile app you can call it tinder for ideation if you will so this is was kind of the way we designed it which was maybe they should get inspiration for a problem that they're dealing with from crowds or ai but they would get it individually they will not get threatened it will not be taking their credit away or their reputation it's something that they can use whenever they want however they want they can swipe if it's good or if it's bad if it's helpful the machine can learn from them and this way they get some analogical thinking from other domains which is very hard to do for experts so an analogical inspiration engine so this is what we did and we designed such an augmented process and tested it with real r d experts and on a real problem the for the first study was done with crowds from amazon mechanical tours because the ai was not worthy and not doing it fast enough and well enough this was a couple of years ago and now of course we have ai and we're working on testing it with ai i'll share with you the results so we chose from one of the organizations that was willing to participate with us a real r d problem that they tried to solve internally unsatisfying solutions came up and they put it before on the open source kind of model as well so they tried the first two models that i showed you internally with r d experts externally online markets and they said we're still not satisfied so we said let's try this new way that we take the expert but we augment them we don't replace them so the problem is a water efficient laundry solution a known problem since we all know there is water drought and laundries are really inefficient if you look at the numbers it's crazy how much an average american family wastes are just for laundry so there was a whole problem uh formulation of course that was taken from the real organization the bottom line i'll just share with you was how would you get clothes cleaned up without putting them through the wash this is what our experts saw and the study was a field experiment so we had real experts sitting in a room in their conference for a limited period of time working on this problem they were randomly assigned as they entered the room they didn't know they got different numbers some expert had the regular the baseline as i introduced you today the r d experts working alone r t experts solo condition we'll call it right so they had the time to think about it they had 10 minutes where they could search online on public domains not something proprietary but they had to bring us a solution a prime of course this is not a fully developed solution this was an experiment but this wasn't a real problem and some sketches and some solutions then the second condition was augmented process the crowd brought some inspiration for the r d experts to solve so this is kind of a tinder for ideas a situation you can see some of those inspirations that they brought so there were uv lights instead of using water other ideas for instance was to have the closed design to begin with with stainless materials and these were sent to the r d experts in 10 minutes they could see they could swipe they could choose maybe if something was helpful for them and do whatever they found as suitable to do with it and then they wrote the solution they proposed and these are just an example of how they got the inspiration they had an image a link if they wanted to explore further and a small short text explaining from the crowd we mapped afterwards the solution space to see how was the search of the experts different with this augmented with this inspiration engine or not and just to show you so this is it was a huge map of uh more than 150 different ideas but a snapshot of it is that you can see that there are different areas when they search the most immediate area is the area of the problem as we defined it to clean without washing machine and then there were other a bit more remote areas of washing machine solution but maybe a different type of washing machine and the most distant one was to reduce the need to clean with why do we even need to clean clothes to begin with so that is a bit more if we think about creativity and distant search this is a bit more distant and creative the results were that when these experts were augmented and got these 10 minutes of an intervention of those different inspiration their search was impacted and you can see that 71 in the solo brought all of their solutions in the same area of clean without a washing machine but once we gave them this augmented intervention if you will or this inspiration engine less than a half the other half brought different figures so it does do something to the cognitive fixation and the disillusion exploration we also had another study i won't show you that actually compared the experts versus crowd and i can tell you that even for the full solution uh the experts had more novel and more useful solution than the crowd could do because they had that in-depth knowledge and experience they built really beautiful solutions and and new ones solution based on that inspiration so it's not like like the copy paste we have it all text analyzed so we could see whatever they've done so this is just a taste just one study one step to show that this can be done if we design the process differently very differently from how it's designed today not instead of the expert but augmenting the expert and trying to think of what is hurting what is problematic for the expert how can we help and then the next stage of course is what if any way could be the role of ai so some people say it's good document it could scale some people say don't even try because it will replace us and the only thing we're good at being creative but as i said this is already taking place so i prefer to try to think what it could uh do instead of waiting and hoping that maybe you know well nothing will change i i enjoy looking at the change in the society and in technology so what we think that it could do is around scalability and the question is what exactly it could do so this is exactly at the study that we are now uh developing and what we are trying give me a second because it skipped one slide so what we are trying to do now is basically to develop an ai based inspiration engine based on all of the patents database that we have using gpt3 to try to understand the hierarchy of each of the patents and looking across different domains you can imagine this being especially helpful in domains that are more prone to analogical design or analogical thinking such as mechanical engineering medical device engineering so it works less actually for the pharma and molecules that work on just chemicals and numbers this is especially good for patents publication for text so this is an nlp approach a text-based approach how will you use it as a practitioner and academic so if you are an engineering professor and you're developing new things would you be interested in looking if similar solutions or similar or different mechanisms to a similar problem are maybe developed somewhere else if you put your problem in this analogical search engine inspiration engine it could help you see other panthers other publication other academics that are working on it and maybe that could bring us back to that collaboration that we talked about this cross-disciplinary collaboration that we believe helps innovation that's one way for academics for companies you could imagine that reinventing the wheel is costly and many times we do that so one thing that it will definitely create is the ability to search more effectively what we do today with technology scouting etc another way is to just a specific tool this augmented tool for your engineer or for your scientists that is thinking about it and giving them all of the sudden the ability to quickly see they don't need to know completely other fields because this can help them search for other mechanisms that are similar or answering similar problems underlying problems but in a very different approach from other domains so it can open their mind in the same way that we did in our experiment this is the hope that it can open the cognitive fixation of experts today to think differently and to bring different mechanisms to solve the problem they're trying to solve right now so this is in collaboration with daphna shahaf a professor for hebrew you who claims that computers can be creative they can have a sense of humor and many other provocative things and with ben wolfson a phd student at stern at nyu working on the project right now and designing the model what was interesting as we're designing the model i would like to share with you that one of the most interesting things that happened is that we re realize our own underlying assumption about how innovation actually happens and that's what i would like us uh to kind of think about together as well so one thing from today's talk is experts crowds and ai what is the role of every each of those actors agents entities the second thing i want us to think about together is what do we each of us have as an underlying assumption of what really makes innovation happen and i can show you how each of them will lead to a different model so if we take the original uh assumption of you know the thinker the lone inventor the the experts the r d experts right so all we will be doing is we're designing the ai model would be to enhance the cognitive abilities of that expert that's it and if we are able to do that that's success if we think about the more collaborative diverse group way of thinking social innovation or even network perspective of innovation which is a very popular and strong one then we will be thinking of who should we have in the table how do we connect the right people so then we might be even developing a very different ai model we should be developing a connector ai for innovation that finds the way to connect those that need to be connected and today are not connected that maybe are working on similar things that if we bring them together then we will get this aha right it's like the manhattan project but then they thought who should we bring for such but without knowing those people right with really having the whole world and databases and anyone who wants to put their work in those databases saying this is what i can do and then the ai is trying to connect people so this is more ai based for innovation assuming that the underlying theory of innovation is network this is what we assume we need to do we just put the right people in the room and this will happen and then as we said earlier on who needs to be in a room right some other people are claiming it's all about the connections and the knowledge but they don't need to be you know sitting together they could and actually maybe it's even better they're more free to think because they don't have this kind of groupthink if they're all sitting together let's have the virtual model work here create an ai that is an online platform and it's underlying that's the engine of the online platform and people can innovate and work together on the same project like they do at cern or others but they don't actually need to meet they don't need to see each other everything can happen online so these are for instance three different uh ai models that will be built differently based on those different models that i share with you a fourth one that i recently also heard and saw is that it's just a matter of pouring money into the problem uh and then what does it mean on the ai model that we should just try to build ai models for innovation that help predict and minimize the risk of investing in the right things so you can see now a lot of vcs are trying to build uh ai models to predict and to learn from their investments so this is a different direction it made sounds a bit uh you know less deep but actually thomas edison himself says the first thing you do from his biography uh the first criteria you need this for innovation and as an inventor is to pour money at the problem so in order to experiment more and more so maybe it's not that far so where with that being said i would like to uh open it and to en to q a and to invite academic and industry collaborations as i said in questions for anyone and and that's it for today thank you very much that was terrific gila thank you so much i'm going to exercise the moderator's prerogative take the first question if you don't mind so when you were talking about these different models of how we might open it up and use ai i was thinking about the research and psychology that shows sometimes there's this risk that when we have groups of people working together on a problem they end up coalescing around this like really mediocre compromise because they know they're going to have to get buy-in from others and that's the easiest way to get buy-in is to sort of converge around the sort of compromised position and then it made me think well wait a minute could you use a i to sort of you could tune ai right you could tune ai to to present really novel provocative solutions that are very different from current solutions or you could tune the ai to provide solutions we think are extremely executable because they're based on our current way of doing things and they could leverage our existing intellectual property and i wanted to hear what you thought about that could we could we like treat ai is almost like something we could parameterize in terms of the kind of innovation we want to get that's beautiful question so i i definitely think you know um that if we go one by one on all the faults and the problems that we know happen in creativity and from the team's uh experience and literature right we have the first idea bias people get the first idea and they jump on it or we have this group thing or we have group polarization not enough voice given to all the people you know in the room so you can imagine an ai design as an ai instigator consultant facilitator in the room imagine you have an alexa ai sitting in that middle of the table and saying by the way we haven't heard enough from x you know the voice distribution is not equally on the table or uh excuse me folks this is the first idea don't get stuck here continue to move and this will for instance help with the first idea bias so i think that is definitely if that's why i'm saying it's so important to think about our theory of innovation as practitioners or academics why do we think innovation really happens and then this is how we should design the ai and i don't think that's necessarily the way we you know the industry is going for today but i do think uh it is something that will show up so consulting companies for innovation that is something that could be very natural for them to do to start to design such a thing so the ai could become a really smart moderator of the conversation could nudge us in particular directions because it would be able to identify sort of what buckets were falling into yes in my class i i do this exercise where the students we try to develop kind of ai for diversity and one of the ideas that the students brought was actually about having that alexa for team discussions with not even just for creativity to to enhance the diversity around the table which we know diversity leads to more creativity so that could be something really neat to do that's awesome we have a question from the audience i want to throw at you here uh from jimmy delanto he asks how do we protect the intellectual property in each of these kinds of processes when we're opening up the you know across using ai and crowdsourcing what what what are people thinking there that is a a whole talk by itself so first of all i know that melissa you're an expert in the lecture property so feel free to tell me what you think as well uh there is a complicated problem so let's start with each of the processes so experts that's the traditional model and that's why we have patents and all of these proprietary for experts has that been suffered from for innovation debated opinions let's say it this way right we patents have become more or less a mechanism for legal uh actions and much less uh are fruitful for innovation depending on which which researcher and what do you read and what do you think so it's definitely not the best and not the only way to get to innovation intellectual property when you look at the second model that i introduced the virtual distributed many cases there sometimes you have a you know very clear intellectual property but sometimes you are very clear but very open right so creative commons license open science look at the open source so indeed open source has some litigation in some cases but it's not like we imagined right so you would imagine when it started people say who will own it how will it happen companies will take it so there are a few cases but it is something that is protectable there are ways and of course with blockchain and other technologies there are ways to really track whatever you know someone is doing or someone is writing more and more so i think it is solvable from a technological perspective it's not as scary as it seems but it's definitely the first people you know who get scared from all of this are the legal folks and i can tell you that there are some legal solutions and then about ai as i show you i think it is an important topic to debate it's not clear at all there is a global debate about whether ai can be listed as an inventor and what is the role of ai i only recommend to check that the artificial inventor project it's it's pretty fascinating there is specific you know they have like team the name of the people who are working on it and then they have the name of the ai kind of in its debuts and and they're claiming the ws deserves uh to be written to be listed as a patent you know owner so that is an open question i think melissa do you have any any thoughts on that since you you work on that i want to ask you i know that you've looked at hackathons and makeathons when someone participates in one of these hackathons is the agreement to participate essentially a license agreement where the solution will be owned by the sponsor of the hackathon that's a great question so there are two models the first one when it's voluntary usually for social problems you're doing it pro bono and then people don't want the intellectual property they really want people to use it so they post it afterwards on open source on open source hardware or software and you can see that being done on a huge level today with uh alpha go as i said the the the biggest the most important scientific breakthrough is about to be open they're not trying to take it proprietary right so it's not just something when people do it voluntary it can also happen on a large scale the other model is that once you get an award you give the ips so if you look at a lot of the online platforms innocent if top quarter uh most of them yet to come when you put your solution out there and the company that wanted it gives you money you give up many times when they win the ansari x prize they keep the idea i forgot about that so you can either you know give the ip away or you can start the negotiation so each platform has a very clear contract so it's not that in the air like it's very clear these platforms have been working for a decade so they have clear i.p regimes that's super cool thank you uh there's another question here from isabella goldmans she asks we found that when using experts innovation crowdsourcing innovation and using ai tools the main barriers are at the interfaces even when experts and organizations are very receptive to crowdsourcing the ideas don't go through the interface efficiently there's too many ideas or they're not understood etc any ideas on how to improve this yes so as you saw in my kind of van diagram i think we need to put more attention on on that overlap on that interface today these models have not been designed with the overlap in mind they were designed with the substitution mindset of it's either or and that's why i think the interface looks the way it is i've been in the crowdsourcing world for like a decade now and i can tell you it's not been designed for this it's been designed to replace the experts but if it were to be designed in a way how do we integrate it how do we work back and forth there's not even a back and forth it's a one-sided thing you put the problem like you get the solution goodbye like that should not be the case if you're designed for an interaction right you should have a dialogue in a way to continue because innovation doesn't stop at the solution we all know it it's a journey but it's not designed as a journey today so i think it's a matter of uh starting to discuss more the underlying assumptions explicitly and designing the process so for instance with hackathons every company every organization that i you know talk to about it i always say we cannot finish there like the timeline cannot be the day of the hackathon the solutions award you know hooray it has to be a year later like what happens let's talk about the whole process so it's even the timeline of the process makes a difference so i think it's all about the process design cool uh harriet wallerstein has a very provocative question that i really liked she asks uh i appreciate the discussion i wonder about the following human error can be useful what about a.i error being
useful the attitudes about these processes can be very optimistic even idealized and there's no process that is like that what about the foresight with human efforts or crowd efforts or ai efforts i love this idea about error being and it's really important we see that a lot when we study when we study inventors as case studies we see how important error actually is so what are your what is your thinking about that and ai that's beautiful i haven't thought about it you know so i'll i'll think on the go but i will think more about it so i definitely think it's like the penitent those many stories right but human error how would we translate it to ai would it be the noise would we be trying to look for something in the noise will we train ai to look for the outliers so i can tell you a research that i've seen recently from james evans from michigan that i highly appreciate uh doing science of science and that's what he's trying to to do to agree but it's still ai so it's not full error right it's not he's not looking for the error but he's saying there is something about looking for the outliers not looking for the pattern so he's designing the ai to try to predict new materials that will be found by looking for the outliers in the publication so he's doing a network mapping and looking at those that are least connected not the ones that are the most cited and from there he's thinking about it so this is the closest to what i know but i haven't seen anyone work in or i haven't thought yet about this error concept i think it's beautiful and worth investigating as i said this is preliminary starting you know it's like we've been in this only a couple of years i think there's room for so many different approaches uh to work on uh it's really like a new path for research and we'll have an nsf kind of workshop this year if anyone is interested and more like there are many people i think that you know could work with their different theories of innovation about it yeah it's a super cool idea especially when we think about like there's a lot of evolutionary models of innovation and creativity and an evolution as you know mutation is often the the element that's going to lead to progress and that mutation is kind of like error right and think about cognition mutation is a little bit like error you can also have directed change i suppose but it's a it's a nice sort of um random element to put in the mix and i reckon we could program ai to make errors right if we wanted if we wanted error we could actually program ai to give certain kinds of error even um maybe it could be oxymoron and it could be like what is an error right is an error like that's some people say we are also like you know an error could be predictable like what do we mean by an error in humans some people think it's something really unpredictable so it's a question that's a good point we should get a doctoral student to start this start working on this excellent you know i noticed one thing about your talk um that i found interesting and that is that a lot of times when i see these um crowdsourcing models a lot of them are based on software but you gave a specifically a hardware example well what i thought of as a hardware example how do i clean my clothes without a washing machine do you think this crowdsourcing model or incorporating ai is particularly more useful for hardware or software innovations because it seems to me like some hardware innovations are very hard to crowd source because of the sort of equipment or scale or investment that would be needed what are your problems uh that's a great question so indeed and especially because the topic of today's innovative virtually i do think that empirically we've seen the success of virtual distributed innovation uh much higher in software than in hardware and there are you know materialistic reasons why i was hoping and i still am hoping that the 3d printing kind of revolution and the advancing technology is there and grab cad so if people check grabcad as a platform this is where those platforms have open hardware to the i would say the strongest degree they are now open hardware medical devices so i would say it's in a lag of at least eight years after uh software and because it is indeed harder but if we imagine that 3d printing will become much more common and much more distributed globally then i do think we can overcome some of it but it's not the same right building a boeing building a cr an aircraft some things will not be uh the same but i you know i am working on an open source hardware ventilator project uh right now for coffee then i'm seeing that it is more possible than i thought it is so to be to be tested to be seen wow that's super cool i can't wait to see what you come up with all right well we're actually a tiny bit ahead of schedule but i think we should thank our speaker i i gotta started a little ahead of schedule so now we'll probably just run ahead of schedule until someone goes a little long so thank you so much hilah that was a fantastic talk i really loved it there's some comments for you that we're going to save from the q a that were just intended for you alone and i really appreciate you being here thanks for doing such great work thank you everyone see you soon [Music] you
2021-10-08