 Hi everyone, I'm Gorn Haft, Technology Evangelist with Red Hat and I'm here today to talk about invention and specifically invention as a collective activity. And the reason I'm talking about this is that this has become very relevant to the open source development model. Now I'd argue that historically, in many people's minds, invention has been sort of thought of in this singular context in terms of an individual. And back in grade school, high school, you know, if you had a quiz about invention, it would be, you know, who invented the steam engine, who invented the refrigerator, who invented the elevator. And maybe the answer would be something like L Labs or Xerox Part, but most likely the answer they'd be looking for would be some specific person. One particular person that epitomizes this is Nikolai Tesla, who's sort of a cult figure, particularly a law of geek communities. And one of the reasons is he's often viewed as this singular inventor, so the inventor of AC, the inventor of the Tesla coil, the inventor of various other things. Now again, I'm not going to say that you don't have singular inventors who really advance the state of the art significantly on their own. And that was certainly the case with Tesla. On the other hand, at the time Tesla was inventing AC electricity, there were actually similar types of invention going on in other places. There was quite a few names that were associated with AC around this time. And of course, they built on other things that AC electricity didn't come out of in air. There certainly were scientists who had come up with some of the theory. And there were also other prior inventions. Now it's certainly true that as I mentioned Bell Labs and Xerox Part, a minute ago, there were other cases where certainly an inventor like Thomas Edison, in the case of Lightbulb, pretty much worked and run large laboratories where many people are doing the inventing, even though one person is maybe most associated with it. But also the case of Lightbulb is illustrative of how inventions can build off of previous inventions. There were actually Lightbulbs of various kinds before Edison invented the Lightbulb that just their filaments didn't last very long. They weren't very practical and therefore Edison is probably rightly credited with coming up with really the first practical Lightbulb. It's also the case that someone like James Watt, who is also often credited with inventing the steam engine, in fact also refined previous designs. Now in Watt's case, what he specifically invented was something called the steam condenser because sort of working on existing steam engines, he recognized that they weren't very efficient and were using a lot of energy, a lot of heat to boil water into steam. And then a lot of that steam was escaping and not really doing anything terribly useful. So Watt came up with this invention, the steam condenser, and was able to make a refined steam engine that was significantly more efficient as predecessors and therefore it was kind of most associated with steam power really becoming a very important economic force. These sort of connections between different types of inventions and building off of what existed in the past was captured wonderfully by James Burke in a BBC show a number of decades ago called Connections and very much encourage anybody listening to this to check out the show. I think most if not all of the episodes are on YouTube and there's also a companion book but Burke is great, he did at least a couple other series but he draws this connection between all these different types of inventions and developments over time and how they kind of led often rather non-obvious ways to the modern world. One example that's a fairly direct connection is Cicard Loom here which was invented essentially to automate textile manufacturing in France originally and it did so by pulling the instructions on to punched cards of various types. Now anybody who is of a certain age and I will admit to having used computer punched cards in school originally and also paper tape which is another variation on the same idea but all of these things were sort of the lineage of how do you package and store a set of instructions and that's of course very core to computers in general today although we use usually various forms of magnetic or maybe optical types of storage today rather than physical media like this is. I'd like to talk now a little bit about some of the theory about how innovation diffuses and I'll give you some historical examples and then I'm going to cap it off with the importance of network effects so not just diffusion but sometimes so much stuff comes together at one time and in one place that has an outsize effect and there's a particularly good historical example of that. So where this sort of theory of diffusion comes from is actually sometimes more this graph in particular this diagram in particular is sometimes associated more with later authors but it actually comes from Everett Rogers in the early 1960s and his theory was that you have these early adopters you have a small number of innovators you have a somewhat larger number of early adopters then you have this majority and then you have these laggards that only adopt things very late in the cycle and if you think about how innovations diffuse it starts out on that left hand side becomes popular and then eventually fills in with the laggards and Clayton Christensen long-time professor at Harvard Business School passed away last year I had a book called the innovators dilemma and I think until I was doing the research for this talk I actually thought he was the one who came up with this curve originally but rather he takes this and he applies this to the idea about how new technologies are adopted and why why that adoption can be challenging for existing firms so for example in context of saying that people here probably familiar with you look around 2000 or so the majority of commercial firms were using Unix there might have been some laggards who were still using mini computers but you know Unix was kind of the big thing in the middle here then there was this little thing called Linux coming along and companies wearing were just starting to use that Linux maybe in the early adopter side and in by around 2000 or so but mostly it was kind of those leading edge people were using it and actually one of the challenges for the mainstream firms who were catering to this early and late majority was that Linux wasn't better than what they had it but it was cheaper it was easier to acquire and it was increasingly good enough for a lot of purposes one other consultant Jeffrey Moore also took up this idea of this you know kind of this curve and what he really identified was this what you call chasm between the early adopters and the early late majority and basically Morris thesis was that companies and projects and technologies can succeed among these innovators and early adopters but it's a very different process to appeal to the majority and a lot of a lot of products projects technologies just don't make it because they're interesting for the this sort of cutting edge type folks but they never make it over that chasm to the majority of users another way that diffusion has happened historically is geographically I mean this is obviously somewhat less relevant today although you still have technologies that for various reasons catch on in particular geographies and don't so much in others at least not for a while but if you look at the old trading routes in oh you know maybe 1200s or so starting something like paper that was invented in China eventually made its way over to Europe and other places by these by these trading routes and that's the way that that kind of technology spread and was improved and changed in various ways something else that was a very important technology called the Norea it's agricultural pumps and that was invented in Roman Syria and it spread east and west and became a rather important innovation in agriculture and gunpowder originally invented in China and it made its way across across the map again various innovations can change this along the way gunpowder could be used for a lot of different purposes you know not even just limited to military and and it and that innovation diffused in various ways and this really happened in all kinds of different technology areas I mean we could end you know next hour or so in these but this is an example from how viticulture growing growing grapes spread from essentially near the hilly flanks where agriculture initially developed into all these other geographies at least where the climate was suitable for growing grapes and here's some uh technology adoption curves for various types of technology consumer technologies that um anybody here is probably could be aware of and yeah you know things happened at different rates so for example the cell phone once it started to really take off really took off and this is I think a somewhat older curve um you know obviously yeah cell phones quite a bit higher up there now and I think it's uh so the shape of these curves vary but the basic idea that you usually have this ramp up period and then it really soars in you know if it crosses the chasm it soars into that majority audience and then eventually you either uh you know hit essentially 100% as in the case of tv or you plateau at some level where uh there's only you know maybe 20 percent of the audience just you know isn't really into it and so you hit some talk line this is uh there's another study that looked at European patents and using patents is kind of a proxy for technology advance and that's and that's kind of an important element here we're not talking just that it takes time for people to decide to use something but rather that the thing that they are using is advancing and uh Robert Allen uh did some of the early studies uh study in this collective invention idea in the early 1980s and what Allen was studying was blast furnaces in the UK so um you know basically smelting smelting furnaces and he was looking at how blast furnaces were improving over time and trying to figure out how these improvements were happening how the information was diffusing and here's some numbers um that you got from some sources and it basically chimneys were getting higher by and large and this he was using this as a proxy for the improvement of blast furnaces you know there were other changes happening as well but the chimney height seemed like a good proxy for the general goodness of the blast furnaces and he drew some conclusions that I think are relevant in open source and elsewhere the first is a lot of this was fairly incremental and you know this this sounds maybe a little bit like DevOps and Agile and all that kind of thing as you these were by and large major game changing advances I mean there were some of those going on in the industrial revolution as well but by and large the pattern was increment was incremental he also found pioneers and followers his idea there were uh or there were you know companies um that were that were the biggest bestest fastest blast furnace and then there were some were willing to take these uh advances in technology and just milk them for the low cost uh and you can imagine the same thing today with something like TVs you know if you could walk basic flat panel TV you can get one really cheap but maybe you will take advantage of technology instead so you can have this great big latest and greatest OLED panel instead and I think the important thing here is he concluded that yeah there was some publication engineering literature but by and large a lot of this was informal disclosure because company companies and individuals found that the friction involved in kind of trying to keep secrets was outweighed by the benefits of sharing and this sounds an awful lot like the open source development model finally I I just love this I just love this graph and Andrew McAfee has also used it in his second machine age book but this is from a book called why the west rules for now by anthropologist Ian Morris and good book it's actually a pretty nice history of the eastern and the western cores it's not really a world history but it's a history of those two cores which were historically the most advanced for various reasons including the fact that you know nod to Jared Diamond done germs and steel they were the two cores that for a variety of geographic reasons sort of had the head start and the really interesting thing about this chart is that you know based he used his social development index which you know very numerically oriented attempt to capture how advanced civilizations were and basically by you know in the grand scheme of things compared to today nothing happened until essentially the industrial revolution or industrial revolutions and then this coming together of technology the steam engine for example uh everything just took off all this stuff came together to make things look like they had before and I think we may see something similar related to open source software today this is the cloud name of computing foundation uh landscape they actually have an interactive landscape uh these days because well you can see this you try and put all in one page uh it's kind of hard to read but I think this idea that you have all this innovation happening and to a degree that innovation is being shared and feeding off each other that's a incredibly powerful thing and therefore I see that we may be in software where the industrial revolution was with technology more generally so thank you for your time and I hope you found this helpful