WARNING: Summaries are generated by a large language model and may be inaccurate. We suggest that you use the synopsis, short and long summaries only as a loose guide to the topics discussed. The model may attribute to the speaker or other participants views that they do not in fact hold. It may also attribute to the speaker views expressed by other participants, or vice versa. The raw transcript (see the bottom of the page) is likely to be a more accurate representation of the seminar content, except for errors at the level of individual words during transcription.

Technology and disruption are two related concepts that are often misunderstood. This video explores how a science of technology disruption can be developed in order to better understand it. It covers the importance of knowledge in technology, the concept of disruption, and the Siba Technology Disruption Framework. Learn how technology disruption affects investments, city and transportation planning, and climate change scenarios, and how a science of technology disruption can be used to make predictions and accelerate improvement.

Technology is a form of knowledge used to explain phenomena. Science is the search for good explanations and requires evidence. Technology is characterized by widgets and tools, and includes software, hardware and other forms of knowledge. Technological progress is determined by the speed, scale and precision of transformation. Mental tools and software are also included, such as buying an item online. This seminar discussed whether a science of technology disruption can exist, exploring the concepts of science, technology and disruption.

Knowledge is essential to technology, as it is the idea and understanding behind the tool. It is not just understanding the process and functions, but also the execution of them. The speaker compared knowledge and technology to a tree with branches that can rejoin and create fusions, and logic with propositions being knowledge and technologies being psychological formulas. Technology and practical knowledge are used to intentionally transform the state of the universe from a less desirable to a more desirable state, which can be measured by metrics such as speed, scale and precision. Additionally, it can be thought of in terms of error reduction, with better technology achieving the desired transformation with less unintended consequences.

Disruption is a rapid and non-linear shift from an older technology to a newer technology which is superior in its capabilities and cheaper to execute transformations. This type of change is triggered when a new technology or convergence of multiple technologies emerges. However, not all technological change is disruptive, as some technologies remain static or have modest marginal improvements. Moving the universe from a less desirable state to a more desirable state involves setting a goal and trying to solve the problem. It is important to note that goals can be incommensurate with one another, meaning they can be completely orthogonal or even contradictory.

Disruption is a rapid, non-linear change caused by the introduction of a cheaper and more effective technology. This change is often represented as a sigmoid curve, with a lower rate of change followed by an exponential surge and then a decay towards an asymptote. It does not necessarily have to be cheaper, it could be more expensive but still be considered disruptive if it is rapidly adopted. Ancient examples of disruption can be seen in the evolution of tools, such as from rocks to flint to obsidian, which could be considered disruption even without the concept of currency.

Technology disruption is a phenomenon where people make decisions to switch from one technology to another, often due to cost, time, effort, values or fashion. To understand it better, a science of technology disruption is needed which looks at people's decisions and how they are affected. A theoretical framework is needed to allow for analysis and improvement of the explanations, and evidence needs to be gathered and predictions made in order to create a science of technology disruption. The Siba Technology Disruption Framework is a theoretical framework based on historical evidence, and there are few teams that are using the same tools.

Technology disruption is a slow and linear process, but mistakes have been made in investments, city and transportation planning, and climate change scenarios due to a lack of understanding of it. To combat this, a science of technology disruption should be developed to better understand it, such as the AI safety community looking at predicting the rate of technological change and literature focused on accelerating improvement in computation. This has been going on for a long time and is an example of non-linearity in technological improvement, which could have serious consequences for climate change if not taken into account.

Science is the search for good explanations. Karl Popper and David Deutsch have proposed frameworks for thinking scientifically. Popper talks about empiricism, induction, deduction and logical inference. This seminar will discuss whether a science of technology disruption can exist, exploring the concepts of science, technology and disruption in the process. It will be an open-ended discussion, with the aim of determining which direction to go in for future sessions.

Technology is a form of knowledge which provides an explanatory framework for phenomena. It is logical, consistent and parsimonious, making testable and falsifiable claims with useful predictions. Science is the search for good explanations and requires a theory with evidence. Technology is characterized by widgets and tools, but also includes other forms of knowledge.

Technology is a form of knowledge that can be applied to software, hardware, and other tangible objects, as well as intangible institutions and norms. Technological progress is determined by the speed, scale and precision with which we can transform the material universe. It is not just hardware such as hammers and levels, but also mental tools and software. An example of this is buying an item online, where the tool is not the item itself, but the knowledge behind it.

Knowledge is essential to technology, as it is the idea and understanding behind the tool. For example, a hammer is an embodiment of knowledge, as it is used to join things together. Knowledge is more than just understanding the process and functions, it is also about how to execute them. This is shown in the example of a traditional potter, who needs to know the process of creating a pot, but also the execution of it. Knowledge is essential for technology, as it is the idea and understanding of the tool, as well as the process and execution of it.

The speaker discussed the idea of knowledge and technology as a tree, with branches that can rejoin and create fusions. They then compared this to logic, with propositions being knowledge and technologies being psychological formulas. As an example, they used a hammer, with the implication being that a proof is needed to get the hammer, and then the consequence being that it can be used to fasten something.

Technology and practical knowledge are used to intentionally transform the state of the universe from a less desirable to a more desirable state. We can measure this transformation using metrics such as speed, scale and precision. Additionally, we can think of it in terms of error reduction, with better technology achieving the desired transformation with less unintended consequences. Asking questions such as these can help us to further understand the concept of a science of this knowledge.

The speaker discussed the concept of moving the universe from a less desirable state to a more desirable state. This is done by setting a goal and trying to solve the problem in order to transform the world around us. It was noted that there is no way to make objective claims about desirability, and that goals can be incommensurate with one another, meaning they can be completely orthogonal or even contradictory. It is important to be aware of this, even if we cannot solve it.

Disruption is a rapid and non-linear shift from an older technology or suite of technologies to a newer technology or suite of technologies in predominant use. This shift is triggered when a new technology or convergence of multiple technologies emerges which is comparable or superior in its capabilities to execute transformations at a significantly lower cost than the older technologies. Not all technological change is disruptive, as some technologies remain static or have modest marginal improvements that do not trigger a regime shift.

Disruption is a phenomenon where rapid and non-linear change occurs in an industry or sector due to the introduction of a cheaper, more effective technology or suite of technologies. This change is often represented as a sigmoid curve, with a lower rate of change followed by a surge that looks exponential and then a decay towards an asymptote. To create a science of technology disruption, the constituents need to be identified. This disruption usually only occurs when the new technology is substantially cheaper than the existing one, rather than just a little bit better.

Disruption does not necessarily have to be cheaper, it could be more expensive but still be considered disruptive if it is rapidly adopted. Ancient examples of disruption can be seen in the evolution of tools, such as from rocks to flint to obsidian. This could be considered disruption even without the concept of currency, as there would be an opportunity cost in terms of the time taken to create the new tools. To include these ancient examples, the definition of cost would need to be expanded to include things such as opportunity cost.

Technology disruption is a phenomenon where people make decisions to switch from one technology to another. It can be due to cost, time, effort, alignment with values, or fashion and taste. It can be distinguished from other types of disruption, such as those caused by accidents of history, as it is more closely related to knowledge. To understand it better, a science of technology disruption would need to look at people's decisions and how they are affected by cost, time, effort and values.

A theoretical framework of explanations that define the phenomena of technology disruption is needed to allow for analysis and improvement of the explanations. A body of empirical evidence has been gathered, including historical examples, market share, user numbers and sales figures. New evidence needs to be gathered and meaningful predictions need to be made in order to create a science of technology disruption.

Technology disruption is an area of research that has not been undertaken before. There are many technology forecasters, but they do not approach the task in a scientific manner and often the predictions are low quality. Kodak in 1995 is an example of this, as the incentives were not in place for them to make accurate predictions. My team has a theoretical framework, the Siba Technology Disruption Framework, which is based on historical evidence and is not pulled out of thin air. We have yet to find any other teams that are using the same tools as us.

Technology disruption is a thriving science and there is potential value in researching it further. Mistakes have been made in investments, city and transportation planning, and climate change scenarios. These mistakes have been costly and investments in coal power plants have been sub-optimal for at least 10 years. City and transportation planning still assumes cars will never be electric or able to drive themselves. Climate change scenarios are based on the assumption that greenhouse gas emissions will continue on a linear trajectory.

Technology disruption is a slow and linear process over many decades, and failing to understand it has serious consequences for climate change. The speaker suggests developing a science of technology disruption in order to better understand it. The AI safety community looks at predicting the rate of technological change and the literature is focused on accelerating improvement in computation. This has been going on for a long time and it is an example of non-linearity in technological improvement.

i think the big question that i have is um can we have a science of technology disruption that's my question that is the organizing idea behind this seminar this is my big this is my question and it's an open question um so that's what i'd like to talk about i thought i would introduce a few ideas maybe some framing for for how to think through this um and then have a fairly open-ended discussion i hope for for a big chunk of the time that we have here and with luck we'll get a sense of whether which direction to go which directions to go in if we explore further in future sessions and and so forth but this is the big question can we have a science of technology disruption and so to to think about that um uh i think there are there are some sub questions to ask first right we the the um obviously one is what is a science um and we can just kind of deal with that real briefly here a second is what is technology and what is disruption so we have need to have a bit of a sense i think of these three ideas in order to in order to begin approaching the the question of can we have a science of technology disruption so i don't want to turn this into some deep dive into um uh philosophy of science and and that's nowhere near my expertise anyway but my general approach to thinking about um what a science is into thinking scientifically more generally is i like karl popper uh and and his sort of um uh take on things i like uh david deutsch very much who's elaborated quite a bit on um popper's original work and so i like personally and obviously this is an open discussion so so so we can we can um debate it although i recommend we don't get bogged down on this particular question to start with but but where i'm coming from with this is that i i prefer to think of science as the search for good explanations i think that that's a really useful and very powerful lens through which to think about science and scientific method um i i mean empiricism certainly plays a role uh induction and deduction and other logical inference certainly plays a role but i i tend to see things at bedrock being really about explaining explaining phenomenon and um uh uh that framework for its practicality i suppose and its elegance um and uh uh so i think it has a lot to to to commend it there are other ways to think about philosophy of science but i think that's a particularly good one with this in mind um there are sort of two things that popper talks about and um that those of us who who receive uh

postgraduate training in um the social environmental and environmental sciences that's my experience we are we are uh trained to understand that two things are necessary here one is to have theory the second is to have evidence so for in order to in order to be doing science as it were uh which by my lights is be searching for good explanations for phenomena you need a theory which provides you with an a a a an explanatory framework um it's uh we again i don't want to get bogged down in details about model versus theory versus theoretical framework these these are debated um nuances really a theory is the is at least in my mind um is the set of one or more a set of explanations of conjectured explanations or um the processes and dynamics that govern a specific phenomenon um it has some usually has some features like uh as logical rationale it has logical consistency as it's it is uh parsimonious these are these are things that make a theory um uh attractive uh and and they tend to be associated with with uh theories that offer good explanatory power as opposed to you know bad theories like um you know zeus did it or something like that would be probably not a very good theory for most scientific phenomena um uh things that will be familiar to everybody or that decent theory makes testable falsifiable preferably um uh claims and ideally not not always but um ideally uh predictions that are useful or testable as well um and then as far as evidence goes uh this is pretty open-ended um it's basically any observation of the phenomenon we're interested in um uh quantitative data can be helpful but it does it is not truly essential for um for science and lots of great scientific work especially historically done without um much quantification at all um a good example might be you know darwin's work on evolution by natural selection for example wasn't very quantitative but it was it was still superb so that that's my t my super short take on um what is science and uh i'll i'll i'll return to those two things a a theory or rather a uh theoretical framework however you prefer to think about it and a body of evidence um i'll return to that in in a bit when i just talk a little bit more about what we do at rethinkx that's the first question just super quickly second one to ask is what is technology now my personal preferred definition is that technology is a form of knowledge it's a i think it's it tends to be um characterized and quite an a a narrower view than that often um sort of sort of widgets and

gizmos and gadgets kind of thing i like to think of technology in a very broad sense it's any it's it's know-how it's practical knowledge and um so that can be software it can be hardware it can be all sorts of tangible things like the knowledge that's embedded in institutions or um norms and so forth all of these in my mind do generally fit into this uh more more broad definition of technology being uh a form of knowledge not all knowledge is practical so not all knowledge is technology per se i think that that that really gets cuts to the essence of what a technology is um and a tool is perhaps a synonym that's useful if you think if you can think of tools being very broad that way mental tools software as tools beyond just hardware you know hammers and levels and that kind of thing um and to kind of elaborate on that a little bit um a follow-on question to ask or at least to think about is uh what is technological progress like what's the difference between advanced technology and and a primitive technology or a um a a an old poor or sub-optimal technology and a new or better or more optimal technology what is what there seems to be some sort of um uh characteristic there perhaps multi-dimensional that that we can look at maybe we can measure it maybe it might depend but there's there's there is perhaps some important sense in which technology gets better and i don't have a perfect way of operationalizing that it probably has to do with things like um the speed and the scale and the precision with which we can undertake a transformation of the material universe that we find ourselves uh in oh yeah yeah sure sorry i'm i'm not seeing anybody i'm staring at the whiteboard and please feel free to just jump in at any moment i'm i'm again this is actually a seminar rather than lecture kind of style i'm happy to to just be be interrupted at any moment so jump in and ask i wanted to ask about um this definition of technology so could you could you repeat what is it exactly that's the essence is it the um the knowledge like the know-how or is it the tool that kind of comes from that or is this that or something yeah i think of it as the knowledge itself because the the um the tools are are really in so many ways an embodiment of the the knowledge that's behind them and um so it's actually a great question because i'm going to give you a funny example i don't know if you guys and gals have noticed this um so you can you can purchase items uh it could be a literal tool like

a hammer or something like that or it could be just some some object some artifact um you could but you can purchase something that's cheap that's kind of junky and it doesn't quite work you know it doesn't work as well as a high quality version of that same tool or product or whatever would be would work it's almost like well you know it's small differences like like why would you make something junky that falls apart when you can make something that's that's that's better for example and if you think of if you think of any gadgeter or or tool um as the embodiment of knowledge uh and and its functioning then um you know you could say well well the the tool itself could be better or worse it could do its job better or worse but is is is that specific embodiment of it the technology or is the knowledge behind it that the understanding and the explanation for why um that particular embodiment was intended to be able to produce an outcome of a given kind i i think that that's that's an example of why i think um because you can have sort of very worse embodiments of a technology um and we have in our ordinary lives that's kind of why i think uh it's the the knowledge um is really what's essential underneath the technologies themselves so for example with with a hammer i think you mentioned so uh the idea using a hammer shaped object of some quality maybe better or worse i can hit a nail into a piece of wood that is the technology that idea yes yes so the the idea embodied in the hammer is the i is a whole set of of ideas right that that fastener that driven fasteners can be used to join things um and there's a there are better or worse ways to drive fasteners um and uh but that can't be quite right because i can have the idea of an artificial general intelligence but in a sci-fi novel it's not a technology until i have a means of actually achieving it right so actually related to this i was going to ask what about the yeah you guys i think you put a really good uh additional element on to the attached to that knowledge which is that it's it's there's knowledge of of the of processes and functions but also of you know how to execute those so how to how enactment of so like the enactment of the transformation using a tool is separate from the enactment of the creation of the tool i would say that both of those things are um uh constituents of a of practical knowledge of know-how um and i often think of this in very simple terms um so say for example a traditional potter

but he makes pots like spins pot collects the clay and spins pots on a wheel and you would think okay well pottery is a technology what what goes into that well it's it's it's not just how a pot gets used and what it's useful usefulness is also the knowledge about how to make it and and what distinguishes a good pot from a bad pot and and so forth so oh sure yeah and i think we've got we've got i think this i think uh for you i'm sure i'm not the only person in here who's played played civ in the civ series right civilization um but i and other you know other um games uh like that but the idea that technology is a tree um i think is is a reasonably decent metaphor um and that knowledge is a tree as well that you know you have you have um you can trace the lineage of of knowledge and technology back through its its you know pre-precursors you can trace it back through these branches there are yeah they're definitely um uh sort of contingent i guess the difference there is unlike a literal tree is that you have fusions you have you know where the branches can rejoin and so on um which is a little bit more like lineage than like uh um you know a literal tree but um but definitely i guess those chains are there for sure i guess for this audience this conversation made me think of uh logic so okay you can think of maybe propositions as knowledge and in a similar way maybe you can think of technologies as um like psychological formulas and so a hammer well i think a proposition is is more like the idea of an agi right yeah maybe the proof is more like knowledge oh yeah okay but but i was thinking okay you mentioned um well if i go back to the hammer maybe the idea of if i have a hammer then i can do this with it it's kind of like an implication and to actually action that then you need a proof that gets you to actually have the hammer um and then and then you can actually achieve what you wanted which was to fasten something that's the consequence of the implication but you need um you need a and a implies b in order to get b um that may well be a a a powerful uh i'll have to think more about that i don't know whether that is that functions as a metaphor or whether it's good enough to be a sort of almost the literal um realization of it um yeah i'm thinking of it as a metaphor at the moment but yeah so yeah it's interesting though it really is i mean and and one so there are two other things to add to this definition before i press on a little bit here because i don't want to

run out of time um but but but uh one other thing is to is just is to think about you know are there dimensions um along which this you know this knowledge progresses and so thinking of the chain or a tree or something like that might be one way to you know sort of operationalize and measure progress i mentioned speed and the scale and the precision with which we can execute transformations so i often think of of technology and and the practical uh element in practical knowledge practical uh element is um that we are our deliberately transforming uh you know this the state of the universe that we find from a less desirable into a more desirable state and it's the so technology the practical dimension of it is is that intentional transformation we see the world as one way and we wish it were another and so we transform it and we transform it using our knowledge um and you could you could then think okay well uh how would you measure that transformation and you could say well how fast did it occur what was the speed of the transformation what scale of transformations are achievable and with what precision um are these are transformations achievable and maybe or maybe you know certainly a combination of all of those but i think those and there i'm sure there are probably other dimensions to think about but that's that's one thing to add into this picture here is um if if we are talking about making transformations using knowledge um and uh not all transformations are equal we can we can we can operationalize and compare um metrics of those uh transformations so that's one thing to think about and then something that dan um and i talked about uh not long ago just recently was the idea that maybe the you could you could think of it in terms of error like a primitive technology achieves the transformation but but inefficiently with lots of error along the way and a better technology is one that can achieve the transformation that's desired but that has less efficiency less loss uh less you know unintended consequences less error along the way and so there's there's something about that about the the error reduction in that process this also seems to be relevant somehow so um fascinating but i don't want to get us to get too bogged down on that it's just that these are questions that are worth asking if we're really you know trying to dig into this this larger question of can we have a science of of um of this stuff um third question i don't want to i want to keep you too

long here because like you said you you should move on but i i wanted to quickly comment that you mentioned basically uh in your definition of sort of like the something practical that you're moving the universe from a less desirable state to a more desirable state when you say desirable i i'm assuming you're basically meaning kind of by definition like we we just have some goal and that is what defines what it is exactly i'm not sure that that there at this point there's a way to make any objective claims about desirability i'm and certainly that's that's above my pay grade as far as like moral realism or some other element of philosophy there that that i wouldn't dare go near um but just i think just for for purely practical purposes at the level of human experience today um you know we look around and we wish you know we wish the world right around us were different in some way and that's basically how we construct concepts of goals or maybe how we construct concepts of problems and solving those problems is to transform the world that's right around us in some way that moves us towards a something that we des uh would would define as desirable and not uh certainly in any objective sense that i can see but that yeah that goal it seems to come from outside of this theory yes yes i would say so for sure at this point i mean i don't i i and again i think that probably would very quickly take you into territory uh philosophical territory that i would be scared to go anywhere near yeah i think i think kind of in general that it's good to it's good to like keep that explicit because otherwise kind of abdicating responsibility for this kind of issue even though it's really thorny uh can lead to more problems like it can lead to technology being misused for example or something like that um yeah i think it's important to kind of address this even if we're not going to solve it just like be aware of it explicitly sure and and i would i would you i mean again you would you would probably end up having to do an exploration about what goals are and what um solutions even look like um or or what viable transformations even look like and certainly we can imagine i mean all of us know this that uh goals can be incommensurate with one another right so you can have you can have multiple goals and the achievement of one undermines the other directly or partially you know the goals are not congruent they can be completely orthogonal to one another or they can run completely contrary to one another

and all sorts of combinations complexity there so yeah getting into goals and values and and all of that stuff is is definitely uh uh that's getting down into the into the philosophical weeds that that um that's certainly enough to frighten me off um but definitely certainly something worth thinking about and and um uh i suppose a very mature science of technology disruption would need to also be fully engaged with the philosophy of understanding goals and values um but it's a good point okay um wait a minute i got disconnected let me see if i can reconnect here i got booted out of roblox one second oh shoot i'm sorry i didn't i was still for too long um can i you see this on the board okay now i i was able to click on the board from way down here at the deportation gate so i'm not going to waste your time by walking up to the top of the mountain i'll just click from here okay um uh so the next question is what is disruption well um i think this is hopefully a little bit shallower of a discussion than the other ones but the way that my research group defines this is a disruption is is essentially a rapid and non-linear regime shift within um to be within an industry or a sector of the economy or a society as a whole that's that's that's a shift from an older technology or suite of technologies in predominant use to a newer technology or suite of technologies in predominant use and um what we've seen over the last certainly since the industrial revolution and actually many dozens of cases before that although those are difficult to to quantify we've seen that this occurs when a new technology or a convergence of multiple technologies um emerges that is either directly comparable or superior in its capabilities to execute these sorts of transformations that we've talked about um at a significantly lower cost than the older technologies so that seems to be the the the these seem to be the key features and trigger of a technology disruption so not all it's important to know that not all technological change is disruptive some technologies do remain static for for a very very long period of time or they can just have very modest marginal improvements that that do not trigger um a regime shift in an industry or as a sector or uh society um so this is so so technological change writ large is different it is more than just disruption so i don't want anybody to think that we're saying that this happens with every technology that comes along um but what we do see is that for many

many many uh examples of uh technologies and it's quite often um the convergence of of several and into a suite of technologies that when this this new capability uh emerges that's that's cheaper that does trigger a surprisingly swift change and hence the term disruption um and so it this is a it's an interesting phenomenon because it's the the impacts uh of transforming an industry or a sector implications and impacts of that can be substantial and um uh so um yeah that's that's the working definition that that that my group uses for a what a disruption is then you can get more specific if you look at things through you know specifically through an economic or financial lens or the you know the lens of business or entrepreneurialism um or if you look at things through a social lens or something like that so you can you can add nuances but the the big picture is that you get rapid change non-linear and um as far as we can tell virtually always some sort of sigmoid function so the the the the nature of the non-linearity is that it's it it follows the general um uh a lower rate of change followed by a higher rate of change followed by a lower rate of change so uh that if you were to think of that that would just sort of give you a crude bell shaped uh curve over time and the um uh uh you know the the if you measure the cumulative effect of that or the integral or however you want to look at it you could you could see a sigmoid coming out of that for something like the share of an industry or a market that is dominated by the new technology over time so it starts out small suddenly you get a surge that looks exponential and then you get um uh you know some decay towards some asymptote some limit at the top of the sigmoid so uh that is our idea of what disruption is and um uh i think this so that's this anyway so that's one of the three big the three big questions here um and from there you can then say okay well what would uh a science of technology disruption look like what would we what we have what would what would the constituents of that be and so sorry go ahead one question about what is destruction um you mentioned kind of um two things there's kind of this this rapid non-linear change but then there's also you also mentioned maybe this is because it's cheaper or something like that um yeah usually usually substantially cheaper like it has to be we what we don't see are are these sorts of changes occurring when a a new product is sort of a little bit better it has to be quite different

it has to be quite compelling especially compelling economically um so that's separate from the question of what is disruption like the definition of disruption doesn't include cheapness right it uh if you had for some reason something that was actually more expensive but still everyone converted to it rapidly would surely you would still consider that to be disruption yes you're right i mean that's that that uh in principle i suppose yeah it's a very good point yeah so i would what i would love to see would be would be to see uh examples that sort of buck this this general pattern that we tend to see well yeah i mean i think the general pattern is probably like some kind of uh like theory that comes out of the science but to me it seems like you could maybe consider applying this to situations where like economics is not as prominent so i'm i'm actually thinking of something if you go way back to like the stone age where you have the um the evolution of ways that people uh turn stones into tools or the the material the type of stone they're using like from just rocks to flint to obsidian and these kinds of things i mean there's no concept of there's no concept of cheaper in terms of monetary value because these are societies before as far as we can tell currency but um maybe you have some like less effort or more effectiveness of the tools that you make or something like that and you still end up with um maybe not rapid in today's standards but um maybe you could call these disruptions because it seems like after some time this technology kind of takes over yeah so i think that that's i think that that's a great point and i think probably what probably the the uh what would be needed there this is great this is excellent superb insight i mean i know that that the um our body of examples that we have includes ancient ones um including several neolithic neolithic ones i think i mentioned arrowheads as an example to dan in a conversation long ago but i think probably um uh we would have to be a bit more expansive in what we mean by cost to include some of the things you said so something like you know it could even be something like opportunity cost like the time taken that you have to spend doing a new you know like uh um yeah the time to the time it would take for example to chip a new arrowhead um or axe hand axe head or something like that would be you know yeah there would be an opportunity cost there compared to the older technology even if there weren't

anything remotely resembling what we would think of as a monetary or financial or token or currency cost or or trade cost so we probably have to be pretty expansive with how we think about cost there and it might be that all of that can somehow in some way reduce it might be reduced to time or time plus uh error time plus effort you know probably efficiency in some way i mean yeah we need to see if we can reduce that down to really essentials um i think it possibly comes back to just whatever people are making their decisions based on which has got to be some kind of like um perceived effort or like um or sort of alignment with their values because this is all about people making choices between technologies like people switching from one technology they have used to another technology as a choice whether or not that's um because this one costs less or because this one is going to take less time or because this one is going to be more effective yeah that's a good point and then you can i can also think of examples where where they could be counter productive but still the choices are made um so for example yeah what about what about fashion and you know taste in certain things like like you know you can even you can even have something right now i think we can sorry sorry but i think it's important to distinguish between technological disruption and disruption i mean so you could have many kind of transitions in people's behavior that have nothing to do with technology or knowledge and one one shouldn't really expect any scientific description of the phenomena of technological disruption to necessarily also apply to those yeah i think i mean the role of knowledge i mean okay you could consider a disruption which is simply due to one country say discovering a hidden cache of raw materials so that the cost of production of a given good is way lower there or something i mean you could certainly disrupt an economy or have a disruption where one product replaces another simply because there's some accident of cost being lower in that new production method or something nothing to do with knowledge just accidents of history and it doesn't it doesn't really seem like that's what you're interested in right um yeah yeah okay great well um let's see i've got a i've got a few more things to get to here before sort of opening things completely to general discussion um uh so how about we ask this question so so if we've what would a science of technology disruption look like

well maybe it would have a few features i have some suggestions here um we would imagine it would have a theory or a theoretical framework if there's if there's sort of a collection of um explanations a collection of theoretical conjectures that that clearly define um the phenomena in question so we're working on that as you can see um that allows analysis to be undertaken that ultimately explains the phenomenon um and uh so i think we i think that as popper's approach and framing of of of science um says you know the theory is essential here um so we're not talking about an a theoretic approach i think i think having a a uh a set of conjectures um that offer explanation that can be improved that these these explanations of this phenomenon can improve can get better we can we can continue searching for better explanations i think that that's a key piece here um of of what the sci science of um technology disruption would look like second thing of uh of course following from what i mentioned earlier is uh that we would need a body of empirical evidence of existing evidence which that theory can be applied and so we could ask what that might look like but my team has been amassing this body of evidence over the last couple of years and we now have several dozen historical examples throughout the ages and a number of the recent ones have some decent data going with them in terms of metrics parameters anyway like market share number of users sales in a given industry those sorts of things so we've actually got um some of that going back into the mid-1800s when fairly reliable census census censuses um started being taken and um then we would need some viable means of gathering new empirical evidence so i think that that's fairly straightforward we can continue to make observations and gather new data today but that would certainly be i think necessary to have a science of this kind and then um ideally i think we can do this as well we need the potential to make meaningful predictions and by meaningful i you know tend to tend to think of the terms of the predictions need to be useful um they need to be measurable they need to be falsifiable um uh but in any case some meaningful however you want to construe that some meaningful predictions about current or future technology disruptions you know instances of this general phenomena that we're interested in and if we have those ingredients then i think uh that would look an awful lot like a science of technology disruption

so um before i open that to discussion and i think that that would be i think that maybe that is something that we can you know spend some time on um in general discussion let me just add a couple of a couple of key points uh and and so if you guys could just postpone questions for just another couple of minutes um immediately following on from that one should probably ask does a science of technology disruption already exist i mean is this already a thing or is this a silly exercise if this entire field already exists and and the answer as far as i can tell is no i don't see anyone else at the moment or historically doing undertaking you know um uh a a search for good explanations it meets these criteria that we've laid out i don't i don't see any evidence that that's been undertaken there are lots of technology forecasters tons and tons there are you know there's there's tons of prophecying and and um prognosticating and all kinds of you know they're all kinds of industry experts who predict whose whole business is predicting the future of a specific industry um uh industries and even individual companies employ their own sort of outlook and forecasting teams that but these folks don't approach that at undertaking in what i would call a scientific manner um that uh and it's plagued with problems and the quality of the analyses and the uh predictions suffers i think as a result of that um many very low quality predictions happen and then of course you know there's a lot of a lot of conflict of interest right so um an example i like to use is how much stock would you put in the predictions of kodak 1995 the year the first consumer digital camera came out they were probably not doing a great job of predicting the future of digital cameras at least publicly um because the incentives were all upside down for making accurate predictions among other things so um uh i don't think we see that this already exists i certainly don't see anybody else using the tools that my team uses any better than we do so we have a theoretical framework and i get to talk about that uh uh uh i don't want to get bogged down in it too much but we have we have a theoretical framework that one of the founders of my uh organization developed siba technology disruption framework um and i don't see other teams citing historical evidence to make their cases um a lot of it just seems to be uh basically pulled out of thin air so um i don't think that we've have any compelling reason to believe that that

there is a thriving science of technology disruption already um that's that's being executed with any rigor anywhere i just don't see it i could be wrong i i would love to be wrong but i as far as i can tell that's not the case um and then so then the final question to ask here is is it worth doing this is this exercise worthwhile is this something we should even be bothering with um and uh again that's an open question i have my own uh sort of personal biases and and feelings about this obviously uh that's not a neutral not an objective view because i'm you know this is partly how i make my living um doing analyses of this kind but um i think that that there could be a lot of value to this because many mistakes in appear to be being made and those mistakes are consequential let me just give you several examples real quick and then i'll end um the first is uh we some of the investments that we make in industries that look like they are going to be disrupted by new technology um or rather are probably already in the process of being disrupted um a big example is the energy sector so clean energy as far as our analyses suggest clean energy seems to be a textbook example of technology disruption coming for an entire sector and misallocation of investment in the energy sector today already at least according to our analysis is in the trillions t trillion with a t uh trillions of dollars so um it has been sub-optimal to invest in coal power plants for example for at least 10 years now at least because of the how because of the lifetime of those power plants the expected lifetime of a coal power plant is 40 years but we should have been able to see that clean energy was coming and disruption was imminent 10 years ago so the last 10 years any investment in a coal power plant was not a good idea and um so that's just one example in the energy sector another example is city and transportation planning so we still are planning our cities as though cars are never going to be electric and they're never going to be able to drive themselves that's probably not good planning and the final example that i've published on recently is climate change where our scenarios and our planning um are all based based on assumptions that we are going to continue to see greenhouse gas emissions from the energy sector the transportation sector and the food sector which together are about 90 of all global greenhouse gas emissions those three sectors that they are going to continue um on a on a linear trajectory and that

decarbonization is going to be a low sorry is going to be a slow and linear process um taking place over many decades um and if that is incorrect and failing to understand disruption as real has has serious um consequences they're relevant to planning for climate change so i would i guess this is my way of very briefly making the case for why this uh why developing a science of technology disruption would be worthwhile um and so with that i guess i'll just open it up completely to to general conversation and questions um and if anybody has any strong feelings or suggestions about where you might take this seminar going forward i have some ideas but where we could go from here i would love to hear those as well yeah thanks good thanks adam um i won't hog the floor with questions because i can ask adam any time those questions may stimulate other ideas and thinking so don't don't don't withhold all of them oh um you asked for other areas um that might be doing something like this i a couple of things came to my mind so um i'm not sure how rigorous and academic this is but in the um in the ai safety community so this is the group of people who are kind of trying to um predict the future of the development of artificial intelligence and in particular looking out for its kind of worst case consequences um um coming from sort of misuse or accidents from the power of the technology um i think people look at kind of trying to predict the rate of technological change um it's not quite the same as disruption i guess but they do talk about things happening quite rapidly because there are kind of feedback loops that come in when you for example increase artificial intelligence tools and then apply those tools into the development of future artificial intelligence tools so i wonder if you looked into this kind of literature i'm not sure how much of a literature there is but if you've looked into this kind of area yeah it's it's uh we have sort of um we have sort of uh maybe you might call it a love-hate relationship with the um uh with the uh not the community but i would say the literature is focused specifically on the um uh specifically on advances in information technology so the accelerating uh accelerating improvement in um computation per you know per unit cost or something like that i mean this is this is just a stratospheric example of non-linearity in technological improvement and it's it's i mean for goodness sake it's gone on for so long i mean i think um you know it's been