1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/org/sloman.org	Tue Aug 13 00:47:01 2013 -0400
     1.3 @@ -0,0 +1,936 @@
     1.4 +#+TITLE:Transcript of Aaron Sloman - Artificial Intelligence - Psychology - Oxford Interview
     1.5 +#+AUTHOR:Dylan Holmes
     1.6 +#+EMAIL:
     1.7 +#+STYLE: <link rel="stylesheet" type="text/css" href="../css/sloman.css" /> 
     1.8 +
     1.9 +
    1.10 +#+BEGIN_QUOTE
    1.11 +
    1.12 +
    1.13 +
    1.14 +
    1.15 +
    1.16 +
    1.17 +
    1.18 +
    1.19 +
    1.20 +
    1.21 +
    1.22 +
    1.23 +
    1.24 +
    1.25 +
    1.26 +*Editor's note:* This is a working draft transcript which I made of
    1.27 +[[http://www.youtube.com/watch?feature=player_detailpage&v=iuH8dC7Snno][this nice interview]] of Aaron Sloman. Having just finished one
    1.28 +iteration of transcription, I still need to go in and clean up the
    1.29 +formatting and fix the parts that I misheard, so you can expect the
    1.30 +text to improve significantly in the near future.
    1.31 +
    1.32 +To the extent that this is my work, you have my permission to make
    1.33 +copies of this transcript for your own purposes. Also, feel free to
    1.34 +e-mail me with comments or corrections.
    1.35 +
    1.36 +You can send mail to =transcript@aurellem.org=.
    1.37 +
    1.38 +Cheers,
    1.39 +
    1.40 +---Dylan
    1.41 +#+END_QUOTE
    1.42 +
    1.43 +
    1.44 +
    1.45 +* Introduction
    1.46 +
    1.47 +** Aaron Sloman evolves into a philosopher of AI
    1.48 +[0:09] My name is Aaron Sloman. My first degree many years ago in
    1.49 +Capetown University was in Physics and Mathematics, and I intended to
    1.50 +go and be a mathematician. I came to Oxford and encountered
    1.51 +philosophers --- I had started reading philosophy and discussing
    1.52 +philosophy before then, and then I found that there were philosophers
    1.53 +who said things about mathematics that I thought were wrong, so
    1.54 +gradually got more and more involved in [philosophy] discussions and
    1.55 +switched to doing philosophy DPhil. Then I became a philosophy
    1.56 +lecturer and about six years later, I was introduced to artificial
    1.57 +intelligence when I was a lecturer at Sussex University in philosophy
    1.58 +and I very soon became convinced that the best way to make progress in
    1.59 +both areas of philosophy (including philosophy of mathematics which I
    1.60 +felt i hadn't dealt with adequately in my DPhil) about the philosophy
    1.61 +of mathematics, philosophy of mind, philsophy of language and all
    1.62 +those things---the best way was to try to design and test working
    1.63 +fragments of mind and maybe eventually put them all together but
    1.64 +initially just working fragments that would do various things.
    1.65 +
    1.66 +[1:12] And I learned to program and ~ with various other people
    1.67 +including ~Margaret Boden whom you've interviewed, developed---helped
    1.68 +develop an undergraduate degree in AI and other things and also began
    1.69 +to do research in AI and so on which I thought of as doing philosophy,
    1.70 +primarily.
    1.71 +
    1.72 +[1:29] And then I later moved to the University of Birmingham and I
    1.73 +was there --- I came in 1991 --- and I've been retired for a while but
    1.74 +I'm not interested in golf or gardening so I just go on doing full
    1.75 +time research and my department is happy to keep me on without paying
    1.76 +me and provide space and resources and I come, meeting bright people
    1.77 +at conferences and try to learn and make progress if I can.
    1.78 +
    1.79 +** AI is hard, in part because there are tempting non-problems.
    1.80 +
    1.81 +One of the things I learnt and understood more and more over the many
    1.82 +years --- forty years or so since I first encountered AI --- is how
    1.83 +hard the problems are, and in part that's because it's very often
    1.84 +tempting to /think/ the problem is something different from what it
    1.85 +actually is, and then people design solutions to the non-problems, and
    1.86 +I think of most of my work now as just helping to clarify what the
    1.87 +problems are: what is it that we're trying to explain --- and maybe
    1.88 +this is leading into what you wanted to talk about:
    1.89 +
    1.90 +I now think that one of the ways of getting a deep understanding of
    1.91 +that is to find out what were the problems that biological evolution
    1.92 +solved, because we are a product of /many/ solutions to /many/
    1.93 +problems, and if we just try to go in and work out what the whole
    1.94 +system is doing, we may get it all wrong, or badly wrong.
    1.95 +
    1.96 +
    1.97 +* What problems of intelligence did evolution solve?
    1.98 +
    1.99 +** Intelligence consists of solutions to many evolutionary problems; no single development (e.g. communication) was key to human-level intelligence.
   1.100 +
   1.101 +[2:57] Well, first I would challenge that we are the dominant
   1.102 +species. I know it looks like that but actually if you count biomass,
   1.103 +if you count number of species, if you count number of individuals,
   1.104 +the dominant species are microbes --- maybe not one of them but anyway
   1.105 +they're the ones who dominate in that sense, and furthermore we are
   1.106 +mostly --- we are largely composed of microbes, without which we
   1.107 +wouldn't survive.
   1.108 +
   1.109 +
   1.110 +# ** Many nonlinguistic competences require sophisticated internal representations
   1.111 +[3:27] But there are things that make humans (you could say) best at
   1.112 +those things, or worst at those things, but it's a combination.  And I
   1.113 +think it was a collection of developments of which there isn't any
   1.114 +single one. [] there might be, some people say, human language which
   1.115 +changed everything. By our human language, they mean human
   1.116 +communication in words, but I think that was a later development from
   1.117 +what must have started as the use of /internal/ forms of
   1.118 +representation --- which are there in nest-building birds, in
   1.119 +pre-verbal children, in hunting mammals --- because you can't take in
   1.120 +information about a complex structured environment in which things can
   1.121 +change and you may have to be able to work out what's possible and
   1.122 +what isn't possible, without having some way of representing the
   1.123 +components of the environment, their relationships, the kinds of
   1.124 +things they can and can't do, the kinds of things you might or might
   1.125 +not be able to do --- and /that/ kind of capability needs internal
   1.126 +languages, and I and colleagues [at Birmingham] have been referring to
   1.127 +them as generalized languages because some people object to
   1.128 +referring...to using language to refer to something that isn't used
   1.129 +for communication. But from that viewpoint, not only humans but many
   1.130 +other animals developed abilities to do things to their environment to
   1.131 +make them more friendly to themselves, which depended on being able to
   1.132 +represent possible futures, possible actions, and work out what's the
   1.133 +best thing to do.
   1.134 +
   1.135 +[5:13] And nest-building in corvids for instance---crows, magpies,
   1.136 + [hawks], and so on --- are way beyond what current robots can do, and
   1.137 + in fact I think most humans would be challenged if they had to go and
   1.138 + find a collection of twigs, one at a time, maybe bring them with just
   1.139 + one hand --- or with your mouth --- and assemble them into a
   1.140 + structure that, you know, is shaped like a nest, and is fairly rigid,
   1.141 + and you could trust your eggs in them when wind blows. But they're
   1.142 + doing it, and so ... they're not our evolutionary ancestors, but
   1.143 + they're an indication --- and that example is an indication --- of
   1.144 + what must have evolved in order to provide control over the
   1.145 + environment in /that/ species.
   1.146 +
   1.147 +** Speculation about how communication might have evolved from internal lanagues.
   1.148 +[5:56] And I think hunting mammals, fruit-picking mammals, mammals
   1.149 +that can rearrange parts of the environment, provide shelters, needed
   1.150 +to have .... also needed to have ways of representing possible
   1.151 +futures, not just what's there in the environment. I think at a later
   1.152 +stage, that developed into a form of communication, or rather the
   1.153 +/internal/ forms of representation became usable as a basis for
   1.154 +providing [context] to be communicated. And that happened, I think,
   1.155 +initially through performing actions that expressed intentions, and
   1.156 +probably led to situtations where an action (for instance, moving some
   1.157 +large object) was performed more easily, or more successfully, or more
   1.158 +accurately if it was done collaboratively. So someone who had worked
   1.159 +out what to do might start doing it, and then a conspecific might be
   1.160 +able to work out what the intention is, because that person has the
   1.161 +/same/ forms of representation and can build theories about what's
   1.162 +going on, and might then be able to help.
   1.163 +
   1.164 +[7:11] You can imagine that if that started happening more (a lot of
   1.165 +collaboration based on inferred intentions and plans) then sometimes
   1.166 +the inferences might be obscure and difficult, so the /actions/ might
   1.167 +be enhanced to provide signals as to what the intention is, and what
   1.168 +the best way is to help, and so on.
   1.169 +
   1.170 +[7:35] So, this is all handwaving and wild speculation, but I think
   1.171 +it's consistent with a large collection of facts which one can look at
   1.172 +--- and find if one looks for them, but one won't know if [some]one
   1.173 +doesn't look for them --- about the way children, for instance, who
   1.174 +can't yet talk, communicate, and the things they'll do, like going to
   1.175 +the mother and turning the face to point in the direction where the
   1.176 +child wants it to look and so on; that's an extreme version of action
   1.177 +indicating intention.
   1.178 +
   1.179 +[8:03] Anyway. That's a very long roundabout answer to one conjecture
   1.180 +that the use of communicative language is what gave humans their
   1.181 +unique power to create and destroy and whatever, and I'm saying that
   1.182 +if by that you mean /communicative/ language, then I'm saying there
   1.183 +was something before that which was /non/-communicative language, and I
   1.184 +suspect that noncommunicative language continues to play a deep role
   1.185 +in /all/ human perception ---in mathematical and scientific reasoning, in
   1.186 +problem solving --- and we don't understand very much about it.
   1.187 +
   1.188 +[8:48]
   1.189 +I'm sure there's a lot more to be said about the development of
   1.190 +different kinds of senses, the development of brain structures and
   1.191 +mechanisms is above all that, but perhaps I've droned on long enough
   1.192 +on that question.
   1.193 +
   1.194 +
   1.195 +* How do language and internal states relate to AI?
   1.196 +
   1.197 +[9:09] Well, I think most of the human and animal capabilities that
   1.198 +I've been referring to are not yet to be found in current robots or
   1.199 +[computing] systems, and I think there are two reasons for that: one
   1.200 +is that it's intrinsically very difficult; I think that in particular
   1.201 +it may turn out that the forms of information processing that one can
   1.202 +implement on digital computers as we currently know them may not be as
   1.203 +well suited to performing some of these tasks as other kinds of
   1.204 +computing about which we don't know so much --- for example, I think
   1.205 +there may be important special features about /chemical/ computers
   1.206 +which we might [talk about in a little bit? find out about]. 
   1.207 +
   1.208 +** In AI, false assumptions can lead investigators astray.
   1.209 +[9:57] So, one of the problems then is that the tasks are hard ... but
   1.210 +there's a deeper problem as to why AI hasn't made a great deal of
   1.211 +progress on these problems that I'm talking about, and that is that
   1.212 +most AI researchers assume things---and this is not just AI
   1.213 +researchers, but [also] philsophers, and psychologists, and people
   1.214 +studying animal behavior---make assumptions about what it is that
   1.215 +animals or humans do, for instance make assumptions about what vision
   1.216 +is for, or assumptions about what motivation is and how motivation
   1.217 +works, or assumptions about how learning works, and then they try ---
   1.218 +the AI people try --- to model [or] build systems that perform those
   1.219 +assumed functions. So if you get the /functions/ wrong, then even if
   1.220 +you implement some of the functions that you're trying to implement,
   1.221 +they won't necessarily perform the tasks that the initial objective
   1.222 +was to imitate, for instance the tasks that humans, and nest-building
   1.223 +birds, and monkeys and so on can perform. 
   1.224 +
   1.225 +** Example: Vision is not just about finding surfaces, but about finding affordances.
   1.226 +[11:09] I'll give you a simple example --- well, maybe not so simple,
   1.227 +but --- It's often assumed that the function of vision in humans (and
   1.228 +in other animals with good eyesight and so on) is to take in optical
   1.229 +information that hits the retina, and form into the (maybe changing
   1.230 +--- or, really, in our case definitely changing) patterns of
   1.231 +illumination where there are sensory receptors that detect those
   1.232 +patterns, and then somehow from that information (plus maybe other
   1.233 +information gained from head movement or from comparisons between two
   1.234 +eyes) to work out what there was in the environment that produced
   1.235 +those patterns, and that is often taken to mean \ldquo{}where were the
   1.236 +surfaces off which the light bounced before it came to me\rdquo{}. So
   1.237 +you essentially think of the task of the visual system as being to
   1.238 +reverse the image formation process: so the 3D structure's there, the
   1.239 +lens causes the image to form in the retina, and then the brain goes
   1.240 +back to a model of that 3D structure there. That's a very plausible
   1.241 +theory about vision, and it may be that that's a /subset/ of what
   1.242 +human vision does, but I think James Gibson pointed out that that kind
   1.243 +of thing is not necessarily going to be very useful for an organism,
   1.244 +and it's very unlikely that that's the main function of perception in
   1.245 +general, namely to produce some physical description of what's out
   1.246 +there.
   1.247 +
   1.248 +[12:37] What does an animal /need/? It needs to know what it can do,
   1.249 +what it can't do, what the consequences of its actions will be
   1.250 +.... so, he introduced the word /affordance/, so from his point of
   1.251 +view, the function of vision, perception, are to inform the organism
   1.252 +of what the /affordances/ are for action, where that would mean what
   1.253 +the animal, /given/ its morphology (what it can do with its mouth, its
   1.254 +limbs, and so on, and the ways it can move) what it can do, what its
   1.255 +needs are, what the obstacles are, and how the environment supports or
   1.256 +obstructs those possible actions.
   1.257 +
   1.258 +[13:15] And that's a very different collection of information
   1.259 +structures that you need from, say, \ldquo{}where are all the
   1.260 +surfaces?\rdquo{}: if you've got all the surfaces, /deriving/ the
   1.261 +affordances would still be a major task. So, if you think of the
   1.262 +perceptual system as primarily (for biological organisms) being
   1.263 +devices that provide information about affordances and so on, then the
   1.264 +tasks look very different. And most of the people working, doing
   1.265 +research on computer vision in robots, I think haven't taken all that
   1.266 +on board, so they're trying to get machines to do things which, even
   1.267 +if they were successful, would not make the robots very intelligent
   1.268 +(and in fact, even the ones they're trying to do are not really easy
   1.269 +to do, and they don't succeed very well--- although, there's progress;
   1.270 +I shouldn't disparage it too much.)
   1.271 +
   1.272 +** Online and offline intelligence
   1.273 +
   1.274 +[14:10] It gets more complex as animals get more sophisticated. So, I
   1.275 +like to make a distinction between online intelligence and offline
   1.276 +intelligence. So, for example, if I want to pick something up --- like
   1.277 +this leaf <he plucks a leaf from the table> --- I was able to select
   1.278 +it from all the others in there, and while moving my hand towards it,
   1.279 +I was able to guide its trajectory, making sure it was going roughly
   1.280 +in the right direction --- as opposed to going out there, which
   1.281 +wouldn't have been able to pick it up --- and these two fingers ended
   1.282 +up with a portion of the leaf between them, so that I was able to tell
   1.283 +when I'm ready to do that <he clamps the leaf between two fingers>
   1.284 +and at that point, I clamped my fingers and then I could pick up the
   1.285 +leaf. 
   1.286 +
   1.287 +[14:54] Whereas, --- and that's an example of online intelligence:
   1.288 +during the performance of an action (both from the stage where it's
   1.289 +initiated, and during the intermediate stages, and where it's
   1.290 +completed) I'm taking in information relevant to controlling all those
   1.291 +stages, and that relevant information keeps changing. That means I
   1.292 +need stores of transient information which gets discarded almost
   1.293 +immediately and replaced or something. That's online intelligence. And
   1.294 +there are many forms; that's just one example, and Gibson discussed
   1.295 +quite a lot of examples which I won't try to replicate now.
   1.296 +
   1.297 +[15:30] But in offline intelligence, you're not necessarily actually
   1.298 +/performing/ the actions when you're using your intelligence; you're
   1.299 +thinking about /possible/ actions. So, for instance, I could think
   1.300 +about how fast or by what route I would get back to the lecture room
   1.301 +if I wanted to [get to the next talk] or something. And I know where
   1.302 +the door is, roughly speaking, and I know roughly which route I would
   1.303 +take, when I go out, I should go to the left or to the right, because
   1.304 +I've stored information about where the spaces are, where the
   1.305 +buildings are, where the door was that we came out --- but in using
   1.306 +that information to think about that route, I'm not actually
   1.307 +performing the action. I'm not even /simulating/ it in detail: the
   1.308 +precise details of direction and speed and when to clamp my fingers,
   1.309 +or when to contract my leg muscles when walking, are all irrelevant to
   1.310 +thinking about a good route, or thinking about the potential things
   1.311 +that might happen on the way. Or what would be a good place to meet
   1.312 +someone who I think [for an acquaintance in particular] --- [barber]
   1.313 +or something --- I don't necessarily have to work out exactly /where/
   1.314 +the person's going to stand, or from what angle I would recognize
   1.315 +them, and so on.
   1.316 +
   1.317 +[16:46] So, offline intelligence --- which I think became not just a
   1.318 +human competence; I think there are other animals that have aspects of
   1.319 +it: Squirrels are very impressive as you watch them. Gray squirrels at
   1.320 +any rate, as you watch them defeating squirrel-proof birdfeeders, seem
   1.321 +to have a lot of that [offline intelligence], as well as the online
   1.322 +intelligence when they eventually perform the action they've worked
   1.323 +out [] that will get them to the nuts. 
   1.324 +
   1.325 +[17:16] And I think that what happened during our evolution is that
   1.326 +mechanisms for acquiring and processing and storing and manipulating
   1.327 +information that is more and more remote from the performance of
   1.328 +actions developed. An example is taking in information about where
   1.329 +locations are that you might need to go to infrequently: There's a
   1.330 +store of a particular type of material that's good for building on
   1.331 +roofs of houses or something out around there in some
   1.332 +direction. There's a good place to get water somewhere in another
   1.333 +direction. There are people that you'd like to go and visit in
   1.334 +another place, and so on. 
   1.335 +
   1.336 +[17:59] So taking in information about an extended environment and
   1.337 +building it into a structure that you can make use of for different
   1.338 +purposes is another example of offline intelligence. And when we do
   1.339 +that, we sometimes use only our brains, but in modern times, we also
   1.340 +learned how to make maps on paper and walls and so on. And it's not
   1.341 +clear whether the stuff inside our heads has the same structures as
   1.342 +the maps we make on paper: the maps on paper have a different
   1.343 +function; they may be used to communicate with others, or meant for
   1.344 +/looking/ at, whereas the stuff in your head you don't /look/ at; you
   1.345 +use it in some other way.
   1.346 +
   1.347 +[18:46] So, what I'm getting at is that there's a great deal of human
   1.348 +intelligence (and animal intelligence) which is involved in what's
   1.349 +possible in the future, what exists in distant places, what might have
   1.350 +happened in the past (sometimes you need to know why something is as
   1.351 +it is, because that might be relevant to what you should or shouldn't
   1.352 +do in the future, and so on), and I think there was something about
   1.353 +human evolution that extended that offline intelligence way beyond
   1.354 +that of animals. And I don't think it was /just/ human language, (but
   1.355 +human language had something to do with it) but I think there was
   1.356 +something else that came earlier than language which involves the
   1.357 +ability to use your offline intelligence to discover something that
   1.358 +has a rich mathematical structure. 
   1.359 +
   1.360 +** Example: Even toddlers use sophisticated geometric knowledge
   1.361 +#+<<example-gap>>
   1.362 +[19:44] I'll give you a simple example: if you look through a gap, you
   1.363 +can see something that's on the other side of the gap. Now, you
   1.364 +/might/ see what you want to see, or you might see only part of it. If
   1.365 +you want to see more of it, which way would you move? Well, you could
   1.366 +either move /sideways/, and see through the gap---and see it roughly
   1.367 +the same amount but a different part of it [if it's a ????], or you
   1.368 +could move /towards/ the gap and then your view will widen as you
   1.369 +approach the gap. Now, there's a bit of mathematics in there, insofar
   1.370 +as you are implicitly assuming that information travels in straight
   1.371 +lines, and as you go closer to a gap, the straight lines that you can
   1.372 +draw from where you are through the gap, widen as you approach that
   1.373 +gap. Now, there's a kind of theorem of Euclidean geometry in there
   1.374 +which I'm not going to try to state very precisely (and as far as I
   1.375 +know, wasn't stated explicitly in Euclidean geometry) but it's
   1.376 +something every toddler--- human toddler---learns. (Maybe other
   1.377 +animals also know it, I don't know.) But there are many more things,
   1.378 +actions to perform, to get you more information about things, actions
   1.379 +to perform to conceal information from other people, actions that will
   1.380 +enable you to operate, to act on a rigid object in one place in order
   1.381 +to produce an effect on another place. So, there's a lot of stuff that
   1.382 +involves lines and rotations and angles and speeds and so on that I
   1.383 +think humans (maybe, to a lesser extent, other animals) develop the
   1.384 +ability to think about in a generic way. That means that you could
   1.385 +take out the generalizations from the particular contexts and then
   1.386 +re-use them in a new contexts in ways that I think are not yet
   1.387 +represented at all in AI and in theories of human learning in any []
   1.388 +way --- although some people are trying to study learning of mathematics.
   1.389 +
   1.390 +* Animal intelligence
   1.391 +
   1.392 +** The priority is /cataloguing/ what competences have evolved, not ranking them.
   1.393 +[22:03] I wasn't going to challenge the claim that humans can do more
   1.394 +sophisticated forms of [tracking], just to mention that there are some
   1.395 +things that other animals can do which are in some ways comparable,
   1.396 +and some ways superior to [things] that humans can do. In particular,
   1.397 +there are species of birds and also, I think, some rodents ---
   1.398 +squirrels, or something --- I don't know enough about the variety ---
   1.399 +that can hide nuts and remember where they've hidden them, and go back
   1.400 +to them. And there have been tests which show that some birds are able
   1.401 +to hide tens --- you know, [eighteen] or something nuts --- and to
   1.402 +remember which ones have been taken, which ones haven't, and so
   1.403 +on. And I suspect most humans can't do that. I wouldn't want to say
   1.404 +categorically that maybe we couldn't, because humans are very
   1.405 +[varied], and also [a few] people can develop particular competences
   1.406 +through training. But it's certainly not something I can do.
   1.407 +
   1.408 +
   1.409 +** AI can be used to test philosophical theories
   1.410 +[23:01] But I also would like to say that I am not myself particularly
   1.411 +interested in trying to align animal intelligences according to any
   1.412 +kind of scale of superiority; I'm just trying to understand what it
   1.413 +was that biological evolution produced, and how it works, and I'm
   1.414 +interested in AI /mainly/ because I think that when one comes up with
   1.415 +theories about how these things work, one needs to have some way of
   1.416 +testing the theory. And AI provides ways of implementing and testing
   1.417 +theories that were not previously available: Immanuel Kant was trying
   1.418 +to come up with theories about how minds work, but he didn't have any
   1.419 +kind of a mechanism that he could build to test his theory about the
   1.420 +nature of mathematical knowledge, for instance, or how concepts were
   1.421 +developed from babyhood onward. Whereas now, if we do develop a
   1.422 +theory, we have a criterion of adequacy, namely it should be precise
   1.423 +enough and rich enough and detailed to enable a model to be
   1.424 +built. And then we can see if it works. 
   1.425 +
   1.426 +[24:07] If it works, it doesn't mean we've proved that the theory is
   1.427 +correct; it just shows it's a candidate. And if it doesn't work, then
   1.428 +it's not a candidate as it stands; it would need to be modified in
   1.429 +some way.
   1.430 +
   1.431 +* Is abstract general intelligence feasible?
   1.432 +
   1.433 +** It's misleading to compare the brain and its neurons to a computer made of transistors
   1.434 +[24:27] I think there's a lot of optimism based on false clues:
   1.435 +the...for example, one of the false clues is to count the number of
   1.436 +neurons in the brain, and then talk about the number of transistors
   1.437 +you can fit into a computer or something, and then compare them. It
   1.438 +might turn out that the study of the way synapses work (which leads
   1.439 +some people to say that a typical synapse [] in the human brain has
   1.440 +computational power comparable to the Internet a few years ago,
   1.441 +because of the number of different molecules that are doing things,
   1.442 +the variety of types of things that are being done in those molecular
   1.443 +interactions, and the speed at which they happen, if you somehow count
   1.444 +up the number of operations per second or something, then you get
   1.445 +these comparable figures).
   1.446 +
   1.447 +** For example, brains may rely heavily on chemical information processing
   1.448 +Now even if the details aren't right, there may just be a lot of
   1.449 +information processing that...going on in brains at the /molecular/
   1.450 +level, not the neural level. Then, if that's the case, the processing
   1.451 +units will be orders of magnitude larger in number than the number of
   1.452 +neurons. And it's certainly the case that all the original biological
   1.453 +forms of information processing were chemical; there weren't brains
   1.454 +around, and still aren't in most microbes. And even when humans grow
   1.455 +their brains, the process of starting from a fertilized egg and
   1.456 +producing this rich and complex structure is, for much of the time,
   1.457 +under the control of chemical computations, chemical information
   1.458 +processing---of course combined with physical sorts of materials and
   1.459 +energy and so on as well.
   1.460 +
   1.461 +[26:25] So it would seem very strange if all that capability was
   1.462 +something thrown away when you've got a brain and all the information
   1.463 +processing, the [challenges that were handled in making a brain],
   1.464 +... This is handwaving on my part; I'm just saying that we /might/
   1.465 +learn that what brains do is not what we think they do, and that
   1.466 +problems of replicating them are not what we think they are, solely in
   1.467 +terms of numerical estimate of time scales, the number of components,
   1.468 +and so on.
   1.469 +
   1.470 +** Brain algorithms may simply be optimized for certain kinds of information processing other than bit manipulations
   1.471 +[26:56] But apart from that, the other basis of skepticism concerns
   1.472 +how well we understand what the problems are. I think there are many
   1.473 +people who try to formalize the problems of designing an intelligent
   1.474 +system in terms of streams of information thought of as bit streams or
   1.475 +collections of bit streams, and they think of as the problems of
   1.476 +intelligence as being the construction or detection of patterns in
   1.477 +those, and perhaps not just detection of patterns, but detection of
   1.478 +patterns that are useable for sending /out/ streams to control motors
   1.479 +and so on in order to []. And that way of conceptualizing the problem
   1.480 +may lead on the one hand to oversimplification, so that the things
   1.481 +that /would/ be achieved, if those goals were achieved, maybe much
   1.482 +simpler, in some ways inadequate. Or the replication of human
   1.483 +intelligence, or the matching of human intelligence---or for that
   1.484 +matter, squirrel intelligence---but in another way, it may also make
   1.485 +the problem harder: it may be that some of the kinds of things that
   1.486 +biological evolution has achieved can't be done that way. And one of
   1.487 +the ways that might turn out to be the case is not because it's not
   1.488 +impossible in principle to do some of the information processing on
   1.489 +artificial computers-based-on-transistors and other bit-manipulating
   1.490 +[]---but it may just be that the computational complexity of solving
   1.491 +problems, processes, or finding solutions to complex problems, are
   1.492 +much greater and therefore you might need a much larger universe than
   1.493 +we have available in order to do things.
   1.494 +
   1.495 +** Example: find the shortest path by dangling strings
   1.496 +[28:55] Then if the underlying mechanisms were different, the
   1.497 +information processing mechanisms, they might be better tailored to
   1.498 +particular sorts of computation. There's a [] example, which is
   1.499 +finding the shortest route if you've got a collection of roads, and
   1.500 +they may be curved roads, and lots of tangled routes from A to B to C,
   1.501 +and so on. And if you start at A and you want to get to Z --- a place
   1.502 +somewhere on that map --- the process of finding the shortest route
   1.503 +will involve searching through all these different possibilities and
   1.504 +rejecting some that are longer than others and so on. But if you make
   1.505 +a model of that map out of string, where these strings are all laid
   1.506 +out on the maps and so have the lengths of the routes. Then if you
   1.507 +hold the two knots in the string -- it's a network of string --- which
   1.508 +correspond to the start point and end point, then /pull/, then the
   1.509 +bits of string that you're left with in a straight line will give you
   1.510 +the shortest route, and that process of pulling just gets you the
   1.511 +solution very rapidly in a parallel computation, where all the others
   1.512 +just hang by the wayside, so to speak.
   1.513 +
   1.514 +** In sum, we know surprisingly little about the kinds of problems that evolution solved, and the manner in which they were solved.
   1.515 +[30:15] Now, I'm not saying brains can build networks of string and
   1.516 +pull them or anything like that; that's just an illustration of how if
   1.517 +you have the right representation, correctly implemented---or suitably
   1.518 +implemented---for a problem, then you can avoid very combinatorially
   1.519 +complex searches, which will maybe grow exponentially with the number
   1.520 +of components in your map, whereas with this thing, the time it takes
   1.521 +won't depend on how many strings you've [got on the map]; you just
   1.522 +pull, and it will depend only on the shortest route that exists in
   1.523 +there. Even if that shortest route wasn't obvious on the original map.
   1.524 +
   1.525 +
   1.526 +[30:59] So that's a rather long-winded way of formulating the
   1.527 +conjecture which---of supporting, a roundabout way of supporting the
   1.528 +conjecture that there may be something about the way molecules perform
   1.529 +computations where they have the combination of continuous change as
   1.530 +things move through space and come together and move apart, and
   1.531 +whatever --- and also snap into states that then persist, so [as you
   1.532 +learn from] quantum mechanics, you can have stable molecular
   1.533 +structures which are quite hard to separate, and then in catalytic
   1.534 +processes you can separate them, or extreme temperatures, or strong
   1.535 +forces, but they may nevertheless be able to move very rapidly in some
   1.536 +conditions in order to perform computations.
   1.537 +
   1.538 +[31:49] Now there may be things about that kind of structure that
   1.539 +enable searching for solutions to /certain/ classes of problems to be
   1.540 +done much more efficiently (by brain) than anything we could do with
   1.541 +computers. It's just an open question.
   1.542 +
   1.543 +[32:04] So it /might/ turn out that we need new kinds of technology
   1.544 +that aren't on the horizon in order to replicate the functions that
   1.545 +animal brains perform ---or, it might not. I just don't know. I'm not
   1.546 +claiming that there's strong evidence for that; I'm just saying that
   1.547 +it might turn out that way, partly because I think we know less than
   1.548 +many people think we know about what biological evolution achieved.
   1.549 +
   1.550 +[32:28] There are some other possibilities: we may just find out that
   1.551 +there are shortcuts no one ever thought of, and it will all happen
   1.552 +much more quickly---I have an open mind; I'd be surprised, but it
   1.553 +could turn up. There /is/ something that worries me much more than the
   1.554 +singularity that most people talk about, which is machines achieving
   1.555 +human-level intelligence and perhaps taking over [the] planet or
   1.556 +something. There's what I call the /singularity of cognitive
   1.557 +catch-up/ ...
   1.558 +
   1.559 +* A singularity of cognitive catch-up
   1.560 +
   1.561 +** What if it will take a lifetime to learn enough to make something new?
   1.562 +... SCC, singularity of cognitive catch-up, which I think we're close
   1.563 +to, or maybe have already reached---I'll explain what I mean by
   1.564 +that. One of the products of biological evolution---and this is one of
   1.565 +the answers to your earlier questions which I didn't get on to---is
   1.566 +that humans have not only the ability to make discoveries that none of
   1.567 +their ancestors have ever made, but to shorten the time required for
   1.568 +similar achievements to be reached by their offspring and their
   1.569 +descendants. So once we, for instance, worked out ways of complex
   1.570 +computations, or ways of building houses, or ways of finding our way
   1.571 +around, we don't need...our children don't need to work it out for
   1.572 +themselves by the same lengthy trial and error procedure; we can help
   1.573 +them get there much faster.
   1.574 +
   1.575 +Okay, well, what I've been referring to as the singularity of
   1.576 +cognitive catch-up depends on the fact that---fairly obvious, and it's
   1.577 +often been commented on---that in case of humans, it's not necessary
   1.578 +for each generation to learn what previous generations learned /in the
   1.579 +same way/. And we can speed up learning once something has been
   1.580 +learned, [it is able to] be learned by new people. And that has meant
   1.581 +that the social processes that support that kind of education of the
   1.582 +young can enormously accelerate what would have taken...perhaps
   1.583 +thousands [or] millions of years for evolution to produce, can happen in
   1.584 +a much shorter time. 
   1.585 +
   1.586 +
   1.587 +[34:54] But here's the catch: in order for a new advance to happen ---
   1.588 +so for something new to be discovered that wasn't there before, like
   1.589 +Newtonian mechanics, or the theory of relativity, or Beethoven's music
   1.590 +or [style] or whatever --- the individuals have to have traversed a
   1.591 +significant amount of what their ancestors have learned, even if they
   1.592 +do it much faster than their ancestors, to get to the point where they
   1.593 +can see the gaps, the possibilities for going further than their
   1.594 +ancestors, or their parents or whatever, have done.
   1.595 +
   1.596 +[35:27] Now in the case of knowledge of science, mathematics,
   1.597 +philosophy, engineering and so on, there's been a lot of accumulated
   1.598 +knowledge. And humans are living a /bit/ longer than they used to, but
   1.599 +they're still living for [whatever it is], a hundred years, or for
   1.600 +most people, less than that. So you can imagine that there might come
   1.601 +a time when in a normal human lifespan, it's not possible for anyone
   1.602 +to learn enough to understand the scope and limits of what's already
   1.603 +been achieved in order to see the potential for going beyond it and to
   1.604 +build on what's already been done to make that...those future steps.
   1.605 +
   1.606 +[36:10] So if we reach that stage, we will have reached the
   1.607 +singularity of cognitive catch-up because the process of education
   1.608 +that enables individuals to learn faster than their ancestors did is
   1.609 +the catching-up process, and it may just be that we at some point
   1.610 +reach a point where catching up can only happen within a lifetime of
   1.611 +an individual, and after that they're dead and they can't go
   1.612 +beyond. And I have some evidence that there's a lot of that around
   1.613 +because I see a lot of people coming up with what /they/ think of as
   1.614 +new ideas which they've struggled to come up with, but actually they
   1.615 +just haven't taken in some of what was...some of what was done [] by
   1.616 +other people, in other places before them. And I think that despite
   1.617 +the availability of search engines which make it /easier/ for people
   1.618 +to get the information---for instance, when I was a student, if I
   1.619 +wanted to find out what other people had done in the field, it was a
   1.620 +laborious process---going to the library, getting books, and
   1.621 +---whereas now, I can often do things in seconds that would have taken
   1.622 +hours. So that means that if seconds [are needed] for that kind of
   1.623 +work, my lifespan has been extended by a factor of ten or
   1.624 +something. So maybe that /delays/ the singularity, but it may not
   1.625 +delay it enough. But that's an open question; I don't know. And it may
   1.626 +just be that in some areas, this is more of a problem than others. For
   1.627 +instance, it may be that in some kinds of engineering, we're handing
   1.628 +over more and more of the work to machines anyways and they can go on
   1.629 +doing it.  So for instance, most of the production of computers now is
   1.630 +done by a computer-controlled machine---although some of the design
   1.631 +work is done by humans--- a lot of /detail/ of the design is done by
   1.632 +computers, and they produce the next generation, which then produces
   1.633 +the next generation, and so on.
   1.634 +
   1.635 +[37:57] I don't know if humans can go on having major advances, so
   1.636 +it'll be kind of sad if we can't.
   1.637 +
   1.638 +* Spatial reasoning: a difficult problem
   1.639 +
   1.640 +[38:15] Okay, well, there are different problems [ ] mathematics, and
   1.641 +they have to do with properties. So for instance a lot of mathematics
   1.642 +that can be expressed in terms of logical structures or algebraic
   1.643 +structures and those are pretty well suited for manipulation and...on
   1.644 +computers, and if a problem can be specified using the
   1.645 +logical/algebraic notation, and the solution method requires creating
   1.646 +something in that sort of notation, then computers are pretty good,
   1.647 +and there are lots of mathematical tools around---there are theorem
   1.648 +provers and theorem checkers, and all kinds of things, which couldn't
   1.649 +have existed fifty, sixty years ago, and they will continue getting
   1.650 +better.
   1.651 +
   1.652 +
   1.653 +But there was something that I was [[example-gap][alluding to earlier]] when I gave the
   1.654 +example of how you can reason about what you will see by changing your
   1.655 +position in relation to a door, where what you are doing is using your
   1.656 +grasp of spatial structures and how as one spatial relationship
   1.657 +changes namely you come closer to the door or move sideways and
   1.658 +parallel to the wall or whatever, other spatial relationships change
   1.659 +in parallel, so the lines from your eyes through to other parts of
   1.660 +the...parts of the room on the other side of the doorway change,
   1.661 +spread out more as you go towards  the doorway, and as you move
   1.662 +sideways, they don't spread out differently, but focus on different
   1.663 +parts of the internal ... that they access different parts of the
   1.664 +... of the room.
   1.665 +
   1.666 +Now, those are examples of ways of thinking about relationships and
   1.667 +changing relationships which are not the same as thinking about what
   1.668 +happens if I replace this symbol with that symbol, or if I substitute
   1.669 +this expression in that expression in a logical formula.  And at the
   1.670 +moment, I do not believe that there is anything in AI amongst the
   1.671 +mathematical reasoning community, the theorem-proving community, that
   1.672 +can model the processes that go on when a young child starts learning
   1.673 +to do Euclidean geometry and is taught things about---for instance, I
   1.674 +can give you a proof that the angles of any triangle add up to a
   1.675 +straight line, 180 degrees. 
   1.676 +
   1.677 +** Example: Spatial proof that the angles of any triangle add up to a half-circle
   1.678 +There are standard proofs which involves starting with one triangle,
   1.679 +then adding a line parallel to the base one of my former students,
   1.680 +Mary Pardoe, came up with which I will demonstrate with this <he holds
   1.681 +up a pen> --- can you see it? If I have a triangle here that's got
   1.682 +three sides, if I put this thing on it, on one side --- let's say the
   1.683 +bottom---I can rotate it until it lies along the second...another
   1.684 +side, and then maybe move it up to the other end ~. Then I can rotate
   1.685 +it again, until it lies on the third side, and move it back to the
   1.686 +other end. And then I'll rotate it again and it'll eventually end up
   1.687 +on the original side, but it will have changed the direction it's
   1.688 +pointing in --- and it won't have crossed over itself so it will have
   1.689 +gone through a half-circle, and that says that the three angles of a
   1.690 +triangle add up to the rotations of half a circle, which is a
   1.691 +beautiful kind of proof and almost anyone can understand it. Some
   1.692 +mathematicians don't like it, because they say it hides some of the
   1.693 +assumptions, but nevertheless, as far as I'm concerned, it's an
   1.694 +example of a human ability to do reasoning which, once you've
   1.695 +understood it, you can see will apply to any triangle --- it's got to
   1.696 +be a planar triangle --- not a triangle on a globe, because then the
   1.697 +angles can add up to more than ... you can have three /right/ angles
   1.698 +if you have an equator...a line on the equator, and a line going up to
   1.699 +to the north pole of the earth, and then you have a right angle and
   1.700 +then another line going down to the equator, and you have a right
   1.701 +angle, right angle, right angle, and they add up to more than a
   1.702 +straight line. But that's because the triangle isn't in the plane,
   1.703 +it's on a curved surface. In fact, that's one of the
   1.704 +differences...definitional differences you can take between planar and
   1.705 +curved surfaces: how much the angles of a triangle add up to. But our
   1.706 +ability to /visualize/ and notice the generality in that process, and
   1.707 +see that you're going to be able to do the same thing using triangles
   1.708 +that stretch in all sorts of ways, or if it's a million times as
   1.709 +large, or if it's made...you know, written on, on...if it's drawn in
   1.710 +different colors or whatever --- none of that's going to make any
   1.711 +difference to the essence of that process. And that ability to see
   1.712 +the commonality in a spatial structure which enables you to draw some
   1.713 +conclusions with complete certainty---subject to the possibility that
   1.714 +sometimes you make mistakes, but when you make mistakes, you can
   1.715 +discover them, as has happened in the history of geometrical theorem
   1.716 +proving. Imre Lakatos had a wonderful book called [[http://en.wikipedia.org/wiki/Proofs_and_Refutations][/Proofs and
   1.717 +Refutations/]] --- which I won't try to summarize --- but he has
   1.718 +examples: mistakes were made; that was because people didn't always
   1.719 +realize there were subtle subcases which had slightly different
   1.720 +properties, and they didn't take account of that. But once they're
   1.721 +noticed, you rectify that. 
   1.722 +
   1.723 +** Geometric results are fundamentally different than experimental results in chemistry or physics.
   1.724 +[43:28] But it's not the same as doing experiments in chemistry and
   1.725 +physics, where you can't be sure it'll be the same on [] or at a high
   1.726 +temperature, or in a very strong magnetic field --- with geometric
   1.727 +reasoning, in some sense you've got the full information in front of
   1.728 +you; even if you don't always notice an important part of it. So, that
   1.729 +kind of reasoning (as far as I know) is not implemented anywhere in a
   1.730 +computer. And most people who do research on trying to model
   1.731 +mathematical reasoning, don't pay any attention to that, because of
   1.732 +... they just don't think about it. They start from somewhere else,
   1.733 +maybe because of how they were educated. I was taught Euclidean
   1.734 +geometry at school. Were you?
   1.735 +
   1.736 +(Adam ford: Yeah)
   1.737 +
   1.738 +Many people are not now. Instead they're taught set theory, and
   1.739 +logic, and arithmetic, and [algebra], and so on. And so they don't use
   1.740 +that bit of their brains, without which we wouldn't have built any of
   1.741 +the cathedrals, and all sorts of things we now depend on.
   1.742 +
   1.743 +* Is near-term artificial general intelligence likely? 
   1.744 +
   1.745 +** Two interpretations: a single mechanism for all problems, or many mechanisms unified in one program.
   1.746 +
   1.747 +[44:35] Well, this relates to what's meant by general. And when I
   1.748 +first encountered the AGI community, I thought that what they all
   1.749 +meant by general intelligence was /uniform/ intelligence ---
   1.750 +intelligence based on some common simple (maybe not so simple, but)
   1.751 +single powerful mechanism or principle of inference. And there are
   1.752 +some people in the community who are trying to produce things like
   1.753 +that, often in connection with algorithmic information theory and
   1.754 +computability of information, and so on. But there's another sense of
   1.755 +general which means that the system of general intelligence can do
   1.756 +lots of different things, like perceive things, understand language,
   1.757 +move around, make things, and so on --- perhaps even enjoy a joke;
   1.758 +that's something that's not nearly on the horizon, as far as I
   1.759 +know. Enjoying a joke isn't the same as being able to make laughing
   1.760 +noises. 
   1.761 +
   1.762 +Given, then, that there are these two notions of general
   1.763 +intelligence---there's one that looks for one uniform, possibly
   1.764 +simple, mechanism or collection of ideas and notations and algorithms,
   1.765 +that will deal with any problem that's solvable --- and the other
   1.766 +that's general in the sense that it can do lots of different things
   1.767 +that are combined into an integrated architecture (which raises lots
   1.768 +of questions about how you combine these things and make them work
   1.769 +together) and we humans, certainly, are of the second kind: we do all
   1.770 +sorts of different things, and other animals also seem to be of the
   1.771 +second kind, perhaps not as general as humans. Now, it may turn out
   1.772 +that in some near future time, who knows---decades, a few
   1.773 +decades---you'll be able to get machines that are capable of solving
   1.774 +in a time that will depend on the nature of the problem, but any
   1.775 +problem that is solvable, and they will be able to do it in some sort
   1.776 +of tractable time --- of course, there are some problems that are
   1.777 +solvable that would require a larger universe and a longer history
   1.778 +than the history of the universe, but apart from that constraint,
   1.779 +these machines will be able to do anything [].  But to be able to do
   1.780 +some of the kinds of things that humans can do, like the kinds of
   1.781 +geometrical reasoning where you look at the shape and you abstract
   1.782 +away from the precise angles and sizes and shapes and so on, and
   1.783 +realize there's something general here, as must have happened when our
   1.784 +ancestors first made the discoveries that eventually put together in
   1.785 +Euclidean geometry. 
   1.786 +
   1.787 +It may be that that requires mechanisms of a kind that we don't know
   1.788 +anything about at the moment. Maybe brains are using molecules and
   1.789 +rearranging molecules in some way that supports that kind of
   1.790 +reasoning. I'm not saying they are --- I don't know, I just don't see
   1.791 +any simple...any obvious way to map that kind of reasoning capability
   1.792 +onto what we currently do on computers. There is---and I just
   1.793 +mentioned this briefly beforehand---there is a kind of thing that's
   1.794 +sometimes thought of as a major step in that direction, namely you can
   1.795 +build a machine (or a software system) that can represent some
   1.796 +geometrical structure, and then be told about some change that's going
   1.797 +to happen to it, and it can predict in great detail what'll
   1.798 +happen. And this happens for instance in game engines, where you say
   1.799 +we have all these blocks on the table and I'll drop one other block,
   1.800 +and then [the thing] uses Newton's laws and properties of rigidity of
   1.801 +the parts and the elasticity and also stuff about geometries and space
   1.802 +and so on, to give you a very accurate representation of what'll
   1.803 +happen when this brick lands on this pile of things, [it'll bounce and
   1.804 +go off, and so on]. And you just, with more memory and more CPU power,
   1.805 +you can increase the accuracy--- but that's totally different than
   1.806 +looking at /one/ example, and working out what will happen in a whole
   1.807 +/range/ of cases at a higher level of abstraction, whereas the game
   1.808 +engine does it in great detail for /just/ this case, with /just/ those
   1.809 +precise things, and it won't even know what the generalizations are
   1.810 +that it's using that would apply to others []. So, in that sense, [we]
   1.811 +may get AGI --- artificial general intelligence --- pretty soon, but
   1.812 +it'll be limited in what it can do. And the other kind of general
   1.813 +intelligence which combines all sorts of different things, including
   1.814 +human spatial geometrical reasoning, and maybe other things, like the
   1.815 +ability to find things funny, and to appreciate artistic features and
   1.816 +other things may need forms of pattern-mechanism, and I have an open
   1.817 +mind about that.
   1.818 +
   1.819 +* Abstract General Intelligence impacts
   1.820 +
   1.821 +[49:53] Well, as far as the first type's concerned, it could be useful
   1.822 +for all kinds of applications --- there are people who worry about
   1.823 +where there's a system that has that type of intelligence, might in
   1.824 +some sense take over control of the planet. Well, humans often do
   1.825 +stupid things, and they might do something stupid that would lead to
   1.826 +disaster, but I think it's more likely that there would be other
   1.827 +things [] lead to disaster--- population problems, using up all the
   1.828 +resources, destroying ecosystems, and whatever. But certainly it would
   1.829 +go on being useful to have these calculating devices. Now, as for the
   1.830 +second kind of them, I don't know---if we succeeded at putting
   1.831 +together all the parts that we find in humans, we might just make an
   1.832 +artificial human, and then we might have some of them as your friends,
   1.833 +and some of them we might not like, and some of them might become
   1.834 +teachers or whatever, composers --- but that raises a question: could
   1.835 +they, in some sense, be superior to us, in their learning
   1.836 +capabilities, their understanding of human nature, or maybe their
   1.837 +wickedness or whatever --- these are all issues in which I expect the
   1.838 +best science fiction writers would give better answers than anything I
   1.839 +could do, but I did once fantasize when I [back] in 1978, that perhaps
   1.840 +if we achieved that kind of thing, that they would be wise, and gentle
   1.841 +and kind, and realize that humans are an inferior species that, you
   1.842 +know, have some good features, so they'd keep us in some kind of
   1.843 +secluded...restrictive kind of environment, keep us away from
   1.844 +dangerous weapons, and so on. And find ways of cohabitating with
   1.845 +us. But that's just fantasy.
   1.846 +
   1.847 +Adam Ford: Awesome. Yeah, there's an interesting story /With Folded
   1.848 +Hands/ where [the computers] want to take care of us and want to
   1.849 +reduce suffering and end up lobotomizing everybody [but] keeping them
   1.850 +alive so as to reduce the suffering. 
   1.851 +
   1.852 +Aaron Sloman: Not all that different from /Brave New World/, where it
   1.853 +was done with drugs and so on, but different humans are given
   1.854 +different roles in that system, yeah.
   1.855 +
   1.856 +There's also /The Time Machine/, H.G. Wells, where the ... in the
   1.857 +distant future, humans have split in two: the Eloi, I think they were
   1.858 +called, they lived underground, they were the [] ones, and then---no,
   1.859 +the Morlocks lived underground; Eloi lived on the planet; they were
   1.860 +pleasant and pretty but not very bright, and so on, and they were fed
   1.861 +on by ...
   1.862 +
   1.863 +Adam Ford: [] in the future.
   1.864 +
   1.865 +Aaron Sloman: As I was saying, if you ask science fiction writers,
   1.866 +you'll probably come up with a wide variety of interesting answers. 
   1.867 +
   1.868 +Adam Ford: I certainly have; I've spoken to [] of Birmingham, and
   1.869 +Sean Williams, ... who else? 
   1.870 +
   1.871 +Aaron Sloman: Did you ever read a story by E.M. Forrester called /The
   1.872 +Machine Stops/ --- very short story, it's [[http://archive.ncsa.illinois.edu/prajlich/forster.html][on the Internet somewhere]]
   1.873 +--- it's about a time when people sitting ... and this was written in
   1.874 +about [1914 ] so it's about...over a hundred years ago ... people are
   1.875 +in their rooms, they sit in front of screens, and they type things,
   1.876 +and they communicate with one another that way, and they don't meet;
   1.877 +they have debates, and they give lectures to their audiences that way,
   1.878 +and then there's a woman whose son says \ldquo{}I'd like to see
   1.879 +you\rdquo{} and she says \ldquo{}What's the point? You've got me at
   1.880 +this point \rdquo{} but he wants to come and talk to her --- I won't
   1.881 +tell you how it ends, but.
   1.882 +
   1.883 +Adam Ford: Reminds me of the Internet.
   1.884 +
   1.885 +Aaron Sloman: Well, yes; he invented ... it was just extraordinary
   1.886 +that he was able to do that, before most of the components that we
   1.887 +need for it existed.
   1.888 +
   1.889 +Adam Ford: [Another person who did that] was Vernor Vinge [] /True
   1.890 +Names/. 
   1.891 +
   1.892 +Aaron Sloman: When was that written?
   1.893 +
   1.894 +Adam Ford: The seventies.
   1.895 +
   1.896 +Aaron Sloman: Okay, well a lot of the technology was already around
   1.897 +then. The original bits of internet were working, in about 1973, I was
   1.898 +sitting ... 1974, I was sitting at Sussex University trying to
   1.899 +use...learn LOGO, the programming language, to decide whether it was
   1.900 +going to be useful for teaching AI, and I was sitting [] paper
   1.901 +teletype, there was paper coming out, transmitting ten characters a
   1.902 +second from Sussex to UCL computer lab by telegraph cable, from there
   1.903 +to somewhere in Norway via another cable, from there by satellite to
   1.904 +California to a computer Xerox [] research center where they had
   1.905 +implemented a computer with a LOGO system on it, with someone I had
   1.906 +met previously in Edinburgh, Danny Bobrow, and he allowed me to have
   1.907 +access to this sytem. So there I was typing. And furthermore, it was
   1.908 +duplex typing, so every character I typed didn't show up on my
   1.909 +terminal until it had gone all the way there and echoed back, so I
   1.910 +would type, and the characters would come back four seconds later.
   1.911 +
   1.912 +[55:26] But that was the Internet, and I think Vernor Vinge was
   1.913 +writing after that kind of thing had already started, but I don't
   1.914 +know. Anyway.
   1.915 +
   1.916 +[55:41] Another...I mentioned H.G. Wells, /The Time Machine/. I
   1.917 +recently discovered, because [[http://en.wikipedia.org/wiki/David_Lodge_(author)][David Lodge]] had written a sort of
   1.918 +semi-novel about him, that he had invented Wikipedia, in advance--- he
   1.919 +had this notion of an encyclopedia that was free to everybody, and
   1.920 +everybody could contribute and [collaborate on it]. So, go to the
   1.921 +science fiction writers to find out the future --- well, a range of
   1.922 +possible futures.
   1.923 +
   1.924 +Adam Ford: Well the thing is with science fiction writers, they have
   1.925 +to maintain some sort of interest for their readers, after all the
   1.926 +science fiction which reaches us is the stuff that publishers want to
   1.927 +sell, and so there's a little bit of a ... a bias towards making a
   1.928 +plot device there, and so the dramatic sort of appeals to our
   1.929 +amygdala, our lizard brain; we'll sort of stay there obviously to some
   1.930 +extent. But I think that they do come up with sort of amazing ideas; I
   1.931 +think it's worth trying to make these predictions; I think that we
   1.932 +should more time on strategic forecasting, I mean take that seriously.
   1.933 +
   1.934 +Aaron Sloman: Well, I'm happy to leave that to others; I just want to
   1.935 +try to understand these problems that bother me about how things
   1.936 +work. And it may be that some would say that's irresponsible if I
   1.937 +don't think about what the implications will be. Well, understanding
   1.938 +how humans work /might/ enable us to make [] humans --- I suspect it
   1.939 +wont happen in this century; I think it's going to be too difficult.