Kind: captions
Language: en
what is Wolfram language in terms of
sort of I mean I can answer the question
for you but is it basically not the
philosophical deep to profound the
impact of it I'm talking about in terms
of tools in terms of things you can
download yeah you can play with what is
it what what does it fit into the
infrastructure what are the different
ways to interact with it right so I mean
that the two big things that people have
sort of perhaps heard of that come from
open language one is Mathematica the
other is both NAFA so Mathematica first
came out 1988 it's this system that is
basically a instance of Wolfram language
and it's used to do computations
particularly in sort of technical areas
and the typical thing you're doing is
you're you're typing little pieces of
computational language and you're
getting computations done it's very kind
of there's like as symbolic yeah it's a
symbolic language so symbolic language I
mean I don't know how to clean they
express that but that makes a very
distinct from what how we think about
sort of I don't know programming in a
ling like python or something right but
so so the point is that in a traditional
programming language the raw material of
the programming language it's just stuff
that computers intrinsically do and the
point of often language is that what the
language is talking about is things that
exist in the world or things that we can
imagine and construct not it's not it's
not sort of it's it's aimed to be an
abstract language from the beginning and
so for example one feature it has is
that it's a symbolic language which
means that you know you think all you
have an X you just type in X and what
why would you just say oh that's X it
won't say error undefined thing you know
I don't know what it is computation you
know for the put in terms of the in
terms of computer now that X could
perfectly well be you know the city of
Boston that's a thing that's a symbolic
thing or it could perfectly well be the
you know the trajectory of some
spacecraft represented as a symbolic
thing and that idea that one can work
with sort of computationally work with
these different these kinds of things
that that exist in the world or describe
the world that's really powerful and
that's what I mean you know when I
started designing well I designed the
predecessor of what's now often language
the thing called SMP which was my first
computer language I am I kind of wanted
to have this the sort of infrastructure
for computation which was as fundamental
as possible I mean this is what I got
for having bit of physicists and tried
to find you know fundamental components
of things and wound up with this kind of
idea of transformation rules for
symbolic expressions as being sort of
the underlying stuff from which
computation would be built and that's
what we've been building from in Wolfram
language and you know operationally what
happens it's I would say by far the
highest level computer language that
exists and it's really been built in a
very different direction from other
languages so other languages have been
about there is a liqueur language it
really is kind of wrapped around the
operations that a computer intrinsically
does maybe people add libraries for this
or that that but the goal of Wolfram
language is to have the language itself
be able to cover this sort of very broad
range of things that show up in the
world and that means that you know there
are 6,000 primitive functions in the
Wolfram language that cover things you
know I could probably pick a a random
here I'm gonna pick just because just
for fun I'll pick them let's take a
random sample of them of all the things
that we have here so let's just say
random sample of 10 of them and let's
see what we get Wow ok so these are
really different things from functions
these are all functions boolean converts
ok that's the thing for converting
between different types of boolean
expressions so for people are just
listening human type 10 random sample
names sampling from all functionally how
many you said there might face thousand
six thousand six thousand ten of them
and
there's a hilarious variety of them yeah
right well we've got things about dollar
request or a dress that has to do with
interacting with the the world of the of
the cloud and so on discrete wavelet
data it's for ROI graphical to the
window yeah yeah we know moveable that's
the user interface kind of thing I want
to pick another ten cuz I think this is
some okay so yeah there's a lot of
infrastructure stuff here that you see
if you if you just start sampling at
random there's a lot of kind of
infrastructural things if you're more
you know if you more look at the some of
the exciting machine learning stuff is
shut off is that also in this pool oh
yeah yeah I mean you know so one of
those functions is like image identify
as a function here we just say image
identify you know it's always good to
let's do this let's say current image
and let's pick up an image hopefully
just an image accessing the webcam to
picture yourself anyway we can say image
identify open square brackets and then
we just paste that picture in their
image identify function of running comes
in picture low and says oh wow it says
look I look like a plunger because I got
this great big thing behind me classify
so this image identify classifies the
most likely object in in the image in it
so there's a wonder okay that's that's a
bit embarrassing let's see what it does
let's pick the top 10 um okay well it
thinks there's oh it thinks it's pretty
unlikely that it's a primary two hominid
a puss eight percent probability yeah
that's that's five seven it's a plunger
yeah well so if we will not give you an
existential crisis and then uh eight
percent or not I should say percent but
no that's a scent that it's a hominid um
and yeah okay it's really I mean I'm
gonna do another one of these just
because I'm embarrassed that it there we
go let's try that let's see what that
did um we took a picture a little bit a
little bit more of me and not just my
bald head so to speak okay
eighty-nine percent problem is it's a
person so that so then I would um but
you know so this is image identify as an
example of one of just
just one function and that's the heart
of the that's like a part of the
language yes I mean you know something
like um I could say I don't know let's
find the geo nearest what could we find
let's find the nearest volcano
um let's find the ten I wonder where it
thinks here is let's try finding the ten
volcano's nearest here okay give us your
nearest volcano here 10 years volcanoes
right let's find out where those oh we
can now we got a list of volcanoes out
and I can say geo list plot that and
hopefully ok so there we go so there's a
map that shows the positions of those
ten volcanoes of the East Coast and the
Midwest density well no we're okay
okay there's no it's not too bad yeah
they're not very close to us we could we
could measure how far away they are but
you know the fact that right in the
language it knows about all the
volcanoes in the world that knows you
know computing what the nearest ones are
it knows all the maps of the world and
so on fundamentally different idea of
what a language is yeah right that's
that's why I like to talk about is you
know a full scale computational language
that's that's what we've tried to do and
just if you can comment briefly I mean
this kind of with the Wolfram language
along with the Wolfram Alpha represents
kind of what the dream of what AI is
supposed to be there's now a sort of a
craze of learning kind of idea that we
can take raw data and from that
extracted the different hierarchies of
abstractions and in order to be able to
under the kind of things that well from
language operates with but we're very
far from learning systems being able to
form that but like what was the context
of history of AI if you could just
comment on there is a you said
computation X and there's just some
sense where in the 80's and 90's sort of
expert systems represented a very
particular computation ax yes all right
and there's a kind of notion that those
efforts didn't pan out right but then
out of that emerges kind of Wolfram
language Wolfram Alpha which is the
I mean yeah I think those are in some
sense those efforts were too modest
they're nice they were they were looking
at particular areas and you actually
can't do it with a particular area I
mean like like even a problem like
natural language understanding it's
critical to have broad knowledge of the
world if you want to do good natural
language understanding and you kind of
have to bite off the whole problem if
you if you say work is gonna do the
block's world over here so to speak you
don't really it's it's it's actually
it's one of these cases where it's
easier to do the whole thing than it is
to do some piece of it you know what one
comment to make about so the
relationship between what we've tried to
do and sort of the learning side of AI
you know in a sense if you look at the
development of knowledge in our
civilization as a whole there was kind
of this notion three three hundred years
ago or so now you want to figure
something out about the world you can
reason it out you can do things which
would just use raw human thought and
then along came sort of modern
mathematical science and we found ways
to just sort of blast through that by in
that case writing down equations now we
also know we can do that with
computation and so on um and so that was
kind of a different thing so when we
look at how do we sort of encode
knowledge and figure things out one way
we could do it is start from scratch
learn everything it's just a neuron that
figuring everything out but in a sense
that denies the sort of knowledge-based
achievements of our civilization because
in our civilization we have learnt lots
of stuff we've surveyed all the
volcanoes in the world we've done you
know we've figured out lots of
algorithms for this or that those are
things that we can encode
computationally and that's what we've
tried to do and we're not saying just
you don't have to start everything from
scratch so in a sense a big part of what
we've done is to try and sort of capture
the knowledge of the world in
computational form in computable form
now there's also some pieces which which
were for a long time undoable by
computers like image identification
where there's a really really useful
module that we can add that is those
things which actually were pretty easy
for humans to do that had been hard for
computers to do I think the thing that's
interesting that's a merger
now is the interplay between these
things between this kind of knowledge of
the world that is in a sense very
symbolic and this kind of sort of much
more statistical kind of things like
image identification and so on and
putting those together by having this
sort of symbolic representation of image
identification that that's where things
get really interesting and where you can
kind of symbolically represent patterns
of things and images and so on um I
think that's you know that's kind of a
part of the path forward so to speak
yeah so the dream of so the machine
learning is not when in my view I think
the view of many people is not any more
close to building the kind of wide world
of computable knowledge that Wolfram
don't think we should build but because
you have a kind of you've you've done
the incredibly hard work of building
this world now machine learning too can
be serviced tools to help you explore
that world yeah and that's what you've
added I mean right now with the version
12 oh yeah if you all seeing some demos
it looks amazing right I mean I think
you know this it's sort of interesting
to see the this sort of the once it's
computable once it's in there it's
running in sort of a very efficient
computational way but then there's sort
of things like the interface of how do
you get there you know how do you do
natural language understanding to get
there how do you how do you pick out
entities in a big piece of text or
something that's I mean actually a good
example right now is our NLP NL
which is we've done a lot of stuff
natural language understanding using
essentially not learning based methods
using a lot of you know a little
algorithmic methods human curation
methods and so on and so on people try
to enter a query and then converting so
the process of converting NLU defined
beautifully as converting their query
into computation come into a
computational language which is a very
well first of all super practical
definition a very useful definition and
then also a very clear definition right
right right so I mean a different thing
is natural language processing where
it's like here's a big lump
next go pick out all the cities in that
text for example and so a good example
of you know so we do that we're using
using modern machine learning techniques
um and it's actually kind of kind of an
interesting process that's going on
right now it's this loop between what do
we pick up with NLP you using machine
learning versus what do we pick up with
our more kind of precise computational
methods in natural language
understanding and so we've got this kind
of loop going between those which is
improving both of them yeah I think you
have some of the state-of-the-art
transforms okay have Bert in there I
think oh you know so Josie of you're
integrating all the models I mean this
is the hybrid thing that people have
always dreamed about are talking about
that makes she's just surprised frankly
that Wolfram language is not more
popular than already it already is you
know that's that's a it's a it's a
complicated issue because it's like it
involves you know it involves ideas and
ideas are absorbed absorbed slowly in
the world I mean I think then there's
sort of like we're talking about there's
egos and personalities and and some of
the the absorption absorption mechanisms
of ideas have to do with personalities
and the students of personalities and
and then a little social network so it's
it's interesting how the spread of ideas
works you know what's funny with Wolfram
language is that we are if you say you
know what market sort of market
penetration if you look at the I would
say very high-end of Rd and sort of the
the people where you say wow that's a
really you know impressive smart person
there very often uses of or from
language very very often if you look at
the more sort of it's a funny thing if
you look at the more kind of I would say
people who are like oh we're just
plodding away doing what we do they're
often not yet Wolfram language users and
that dynamic it's kind of odd that there
hasn't been more rapid trickle down
because we really you know the high-end
we've really been very successful in for
a long time and it's it's some but was
you know that's partly I think a
consequence of my fault in a sense
because it's
kind of you know I have a company which
is really emphasizes sort of creating
products and building a sort of the best
possible technical tower we can rather
than sort of doing the commercial side
of things and pumping it out and so yeah
most effective what and there's an
interesting idea that you know perhaps
you can make more popular by opening
everything everything up sort of the
github bottle but there's an interesting
I think I've heard you discussed this
that that turns out not to work in a lot
of cases like in this particular case
that you want it you know that when you
deeply care about the integrity the
quality of the knowledge that you're
building that unfortunately you can't
you can't distribute that effort yeah
it's not the nature of how things work I
mean you know what we're trying to do is
a thing that for better or worse
requires leadership and it requires kind
of maintaining a coherent vision over a
long period of time and doing not only
the cool vision related work but also
the kind of mundane in the trenches make
the thing actually work well work how do
you build the knowledge because that's
the fascinating thing that's the mundane
the fascinating in the mundane as well
building the knowledge they're adding
integrating more data yeah I mean that's
probably not the most stunning that the
things like get it to work in all these
different cloud environments and so on
that's pretty you know it's very
practical stuff you know have the user
interface be smooth and you know have
there be take on him you know a fraction
of a millisecond to do this or that
that's a lot of work and it's some it's
it's but you know I think my it's an
interesting thing over the period of
time you know often language has existed
basically for more than half of the
total amount of time that any language
any computer language has existed that
is computer language maybe 60 years old
you know give or take um and both
languages 33 years old so it's it's kind
of a.m. and I think I was realizing
recently there's been more innovation
in the distribution of software than
probably than in the structure of
programming languages over that period
of time and we you know we've been sort
of trying to do our best to adapt to it
and the good news is that we have you
know because I have a simple private
company and so on that doesn't have you
know a bunch of investors you know
telling us we're gonna do this so that
they have lots of freedom in what we can
do and so for example we're able to oh I
don't know we have this free Wolfram
engine for developers which is a free
version for developers and we've been
you know we've they're a site licenses
for for mathematical more from language
basically all major universities
certainly in the u.s. by now so it's
effectively free to people and all the
universities in effect and you know
we've been doing a progression of things
I mean different things like Wolfram
Alpha for example the main website is
just a free website
what is Wolfram Alpha okay both now for
is a system for answering questions
where you ask in question with natural
language and it'll try and generate a
report telling you the answer to that
question so the question could be
something like you know what's the
population of Boston divided by New York
compared to New York and it'll take
those words and give you an answer and
that have been verts the words into
computable and into into Wolfram
language a common language in the
additional language and then could use
the points in underlying knowledge
belongs to Wolfram Alpha to the Wolfram
language what's the let's call it the
Wolfram knowledge base knowledge base I
mean it's it's been a that's been a big
effort over the decades to collect all
that stuff and you know more of it flows
in every second so can you just pause on
that for a second like that's the one of
the most incredible things of course in
the long term were from language itself
is the fundamental thing but in the
amazing sort of short term the the
knowledge base is kind of incredible so
what's the process of building in that
knowledge base the fact that you first
of all from the very beginning that
you're brave enough to start to take on
the general knowledge base
and how do you go from zero to the
incredible knowledge base that you have
now well yeah it was kind of scary at
some level I mean I had I had wondered
about doing something like this since I
was a kid so it wasn't like I hadn't
thought about it for a while but most of
us most of the brilliant dreamers give
up such a such a difficult engineering
notion at some point right right well
the thing that happened with me which
was kind of it's a it's a live your own
paradigm kind of theory so basically
what happened is I had assumed that to
build something like wolf malphur would
require sort of solving the general AI
problem that's what I had assumed and so
I kept on thinking about that and I
thought I don't really know how to do
that so I don't do anything then I
worked on my new kind of science project
instead of exploring the computational
universe and came up with things like
this principle of computational
equivalence which say there is no bright
line between the intelligence and the
milli computational so I thought look
that's this paradigm I've built you know
now it's you know now I have to eat that
dog food myself so to speak you know
I've been thinking about doing this
thing with computable knowledge forever
and you know let me actually try and do
it and so it was you know if my if my
paradigm is right then this should be
possible but the beginning was certainly
you know is a bit daunting I remember I
took that the the the early team to a
big reference library and we like
looking at this reference library and
it's like you know my basic statement is
our goal over the next year or two is to
ingest everything that's in here and
that's you know it seemed very daunting
but but in a sense I was well aware of
the fact that it's finite you know the
fact you can walk into the reference
library it's a big big thing with lots
of reference books all over the place
but it is finite you know there's not an
infinite you know it's not the infinite
corridor of so to speak of a reference
library it's not truly infinite so to
speak but but no I mean and then then
what happened was sort of interesting
there was from a methodology point of
view was I didn't start off saying let
me have a grand theory for how all this
knowledge works it was like let's you
know
implement this area this area this area
of
hundred areas and so on it's long work I
also found that you know i-i've been
fortunate in that our products get used
by sort of the world's experts and lots
of areas and so that really helped
because we were able to ask people you
know the world expert on this or that
and were able to ask them for input and
so on and I found that my general
principle was that any area where there
wasn't some expert who helped us figure
out what to do wouldn't be right and you
know because our goal was to kind of get
to the point where we had sort of true
expert level knowledge about everything
and so that you know that the ultimate
goal is if there's a question that can
be answered on the basis of general
knowledge and a civilization
make it be automatic to be able to
answer that question and you know and
now what Walton I forgot used in Syria
from the very beginning and it's now as
you know like sir and so it's people are
kind of getting more of the you know
they get more of the sense of this is
what should be possible to do I mean in
a sense the question-answering problem
was viewed as one of the sort of core AI
problems for a long time I had kind of
an interesting experience I had a friend
Marvin Minsky who was a well-known a AI
person from from right around here and I
remember when my morph novel was coming
out um as a few weeks before I came out
I think I happened to see Marvin and I
said I should show you this thing we
have you know it's a question answering
system and he was like okay type
something and it's like okay fine and
then he's talking about something
different
I said no Marvin you know this time it
actually works you know look at this it
actually works these types and a few
more things there's maybe ten more
things of course we have a record of
what he's typed in which is kind of
interesting but and they do I can you
share where his mind was in the testing
space like what whoa all kinds of random
things he's trying random stuff you know
medical stuff and you know chemistry
stuff and you know astronomy and so on I
think was like like you know after a few
minutes he was like oh my god it
actually works
and the the but that was kind of told
you something about the state you know
what what happened in AI because people
had you know in a sense by trying to
solve the bigger problem we were able to
actually make something that would work
now
to be fair you know we had a bunch of
completely unfair advantages for example
we already built a bunch of often
language which was you know very high
level symbolic language we had you know
I had the practical experience of
building big systems I have the sort of
intellectual confidence to not just sort
of give up and doing something like this
I think that the you know it is a it's
always a funny thing you know I've
worked on a bunch of big projects in my
life and I would say that the you know
you mention ego I would also mention
optimism so does very careful I mean in
you know if somebody said this project
is gonna take 30 years its I you know it
would be hard to sell me on that you
know I'm always in the in the well I can
kind of see a few years you know
something's gonna happen a few years and
and usually does something happens in a
few years but the whole the tale can be
decades long and that's a that's a you
know from a personal point of view or is
the challenges you end up with these
projects that have infinite tales and
the question is do the tales kind of do
you just drown in kind of dealing with
all of the tales of these projects and
that's that's an interesting sort of
personal challenge and like my efforts
now to work on fundamental theory of
physics which I've just started doing
and I'm having a lot of fun with it but
it's kind of you know it's it's kind of
making a bet that I can I can kind of
like you know I can do that as well as
doing the incredibly energetic things
that I'm trying to do with all from
language and so on I mean vision yeah
and underlying that I mean I just talked
for the second time with Elon Musk and
that you you to share that quality a
little bit of that optimism of taking on
basics we do the daunting what most
people call impossible and he knew
take it on out of you can call it ego
you can call it naivety you can call it
optimism whatever the heck it is but
that's how you solve the impossible
things yeah I mean look at what happens
and I don't know you know in my own case
I know it's been I progressed oligo a
bit more confident and progressively
able to you know decide that these
projects aren't crazy but then the other
thing is the other the other trap the
one can end up with is oh I've done
these projects and they're big let me
never do a project that's any smaller
than any project I've done so far and
that's yeah you know and that can be a
trap and and often these projects are of
completely unknown you know that their
depth and significance is actually very
hard to know yeah I'm the sort of
building this giant knowledge base is
behind well from language WolframAlpha
what do you think about the internet
what do you think about for example
Wikipedia these large aggregations of
text that's not converted into
computable knowledge do you think you
would if you look at Wolfram language
Wolfram Alpha 20 30 maybe 50 years down
the line do you hope to store all of the
sort of Google's dream is to make all
information searchable accessible but
that's really as defined it's it's a it
doesn't include the understanding of
information right do you hope to make
all of knowledge represented with the
hope so that's what we're trying to do
it hard is that problem they could
closing that gap well it depends on the
use cases I mean so if it's a question
of answering general knowledge questions
about the world we're in pretty good
shape on that right now
if it's a question of representing like
an area that we're going into right now
is computational contracts being able to
take something which would be written in
legalese it might even be the
specifications for you know what should
the self-driving car do when it
encounters the so that or the other what
should the you know
whatever they you know write that in a
computational language and be able to
express things about the world you know
if the creature that you see running
across the road is a you know thing at
this point in the you know Tree of Life
then it's worth this way otherwise don't
those kinds of things are there ethical
components when you start to get to some
of the messy human things are those in
encoder well into computable knowledge
well I think that it is a necessary
feature of attempting to automate more
in the world that we encode more and
more of ethics in a way that gets sort
of quickly you know is able to be dealt
with by computer I mean I've been
involved recently I sort of got backed
into being involved in the question of
automated content selection on the
Internet
so you know the Facebook's Google's
Twitter's you know what how do they rank
the stuff they feed to us humans so to
speak and the question of what are you
know what should never be fed to us what
should be blocked forever what should be
up ranked you know and what is the what
are the current principles behind that
and what I kind of well a bunch of
different things I realized about that
but one thing that's interesting is
being able you know in effect you're
building sort of an AI ethics you have
to build an AI ethics module in effect
to decide is this thing so shocking I'm
not going to show it to people is this
thing so whatever and and I did realize
in thinking about that that you know
there's not gonna be one of these things
it's not possible to decide or it might
be possible but it would be really bad
for the future of our species if we just
decided there's this one AI FX module
and it's going to determine the the the
practices of everything in the world so
to speak and I kind of realized one has
to sort of break it up and that's an
that's an interesting societal problem
of how one does that and how one sort of
has people sort of self-identify for you
know I'm buying in in the case of just
content selection it's sort of easier
because it's like an individual or an
individual it's not something that
cuts across sort of societal boundaries
but it's a really interesting notion of
I heard you'd describe I really like it
sort of maybe in the sort of have
different AI systems that have a certain
kind of brand that they represent
essentially you could have like I don't
know whether it's conserve conservative
or liberal and then libertarian and
there's an R and E an Objectivist like
system a different ethical and Co I mean
it's almost encoding some of the
ideologies which we've been struggling I
come from the Soviet Union that didn't
work out so well with the ideologies
they worked out there is so you you have
but they also everybody purchased that
particular ethic system indeed and in
the same I suppose could be done encoded
that that system could be encoded into
computational knowledge and allow us to
explore in the realm of in the digital
space as that's the right exciting
possibility are you playing with those
ideas and or from language yeah yeah I
mean the the the you know that's we
often language has sort of the best
opportunity to kind of express those
essentially computational contracts
about what to do now there's a bunch
more work to be done to do it in
practice for you know deciding the is
this a credible news story what does
that mean or whatever whatever else do
kind of pick I think that that's um you
know that's the the question of well
exactly what we get to do with that is
you know for me it's kind of a
complicated thing because there are
these big projects that I think about
like you know find the fundamental
theory of physics okay that's possible
one right bucks number two you know
solve the IIx problem in the case of you
know figure out how you rank all content
so to speak and decide what people see
that's that's kind of a box number two
so to speak
these are big projects and and I think
for anything is more important the the
fundamental nature of reality or depends
who you ask it's one of these things
that's exactly like you know what's the
ranking right it's the it's the ranking
system now it's like who's who's module
do you use to rank that if you
and I think come having multiple modules
is really compelling notion to us humans
in a world where there's not clear that
there's a right answer it perhaps you
have systems that operate under
different how would you say it I mean
it's different value systems based
different value systems I mean I think
you know in a sense the I mean I'm not
really a politics oriented person but
but you know in the kind of
totalitarianism it's kind of like you're
gonna have this this system and that's
the way it is I mean kind of the you
know the concept was sort of a
market-based system where you have okay
I as a human I'm going to pick this
system I is another human I'm going to
pick this system I mean that's in a
sense this case of automated content
selection is a non-trivial but it is
probably the easiest of the AI ethics
situations because it is each person
gets to pick for themselves and there's
not a huge interplay between what
different people pick by the time you're
dealing with other societal things like
you know what should the policy of the
central bank could be or something or
healthcare says allow this kind of
centralized kind of things right well I
mean healthcare again has the feature
that that at some level each person can
pick for themselves so to speak I mean
whereas there are other things where
there's a necessary Public Health that's
one example well that's not where that
doesn't get to be you know something
which people can what they pick for
themselves
they may impose on other people and then
it becomes a more non-trivial piece of
sort of political philosophy of course
the central banking system some would
argue we would move we need to move away
into digital currency and so on and
Bitcoin and Ledger's and so on so yes
there's a lot of we've been quite
involved in that and that's it that's
where that's sort of the motivation for
computational contracts in part comes
out of you know this idea oh we can just
have this autonomously executing smart
contract the idea of a computational
contract is just to say you know have
something where all of the conditions of
the contract are represented in
computational form so in principle it's
automatic text secured the contract and
I think that's you
that will surely be the future of you
know the idea of legal contracts written
in English or legalese or whatever and
where people have to argue about what
goes on is it surely not you know we
have a much more streamlined process if
everything can be represented
computationally and the computers can
kind of decide what to do
I mean ironically enough you know old
gottfried leibniz back in the you know
1600s was saying exactly the same thing
but he had you know his pinnacle of
technical achievement was this brass for
function mechanical calculator thing
that never really worked properly
actually um and you know so he was like
300 years too early for that idea but
now that idea is pretty realistic I
think and you know you asked how much
more difficult is it than what we have
now I'm often language to express I call
it symbolic discourse language being
able to express sort of everything in
the world in kind of computational
symbolic form um I I think it is
absolutely within reach I mean I think
it's a you know I don't know maybe I'm
just too much of an optimist but I think
it's a it's a limited number of years to
have a pretty well built out version of
that that will allow one to encode the
kinds of things that are relevant to
typical legal contracts and and these
kinds of things the idea of symbolic
discourse language can you try to define
the scope of what of what it is so we're
having a conversation it's a natural
language can we have a representation of
these sort of actionable parts of that
conversation in a precise computable
form so that a computer could go do it
and not just contracts but really sort
of some of the things we think of as
common sense essentially even just like
basic notions of human life well I mean
things like you know I am I'm getting
hungry and want to eat something right
right that that's something we don't
have a representation you know in wolfen
language right now if I was like I'm
eating blueberries and raspberries and
things like that and I'm eating this
amounts of them we know all about those
kinds of fruits and plants and nutrition
content and all that kind of thing but
the I want to eat them part of it is not
covered
yet um and that you know you need to do
that in order to have a complete
symbolic discourse language to be else I
have a natural language conversation
right right to be able to express the
kinds of things that say you know if
it's a legal contract it's you know the
parties desire to have this and that and
that's you know that's a thing like I
want to eat a bras berry or something
that that's isn't that day isn't this
just throwing you said it's centuries
old this dream yes but it's also the
more near-term the dream of touring in
four million a touring test yes so do
you do you hope do you think that's the
ultimate test of creating something
special we said I tell I think my
special look if the test is does it walk
and talk like a human well that's just
the talking like a human but um the
answer is it's an okay test if you say
is it a test of intelligence you know
people have attached wolf alpha the wolf
now for API - you know Turing test BOTS
and those BOTS just lose immediately
because all you have to do is ask you
five questions that you know are about
really obscure weird pieces of knowledge
and it's just drop them right out and
you say that's not a human ID it's it's
a it's a different thing it's achieving
a different right now but it's yeah I
would argue not I would argue it's not a
different thing it's actually
legitimately Wolfram Alpha is
legitimately
languor Wolfram language only is
legitimately trying to solve the torrent
Dean tent of the Turing test perhaps the
intent yeah perhaps the intent I mean
it's actually kind of fun you know I'm
touring trying to work out he's thought
about taking encyclopedia britannica and
you know making it computational in some
way and he estimated how much work it
would be and actually I have to say he
was a bit more pessimistic than the
reality we did it more efficiently but
to him that represent so I mean he was
that he was on the fighting mental tasks
yeah right he believes that had the same
idea I mean it was you know we were able
to do it more efficiently because we had
a lot we had layers of automation
that he I think hadn't you know it's
it's hard to imagine those layers of
abstraction um that end up being being
built up but to him it represented like
an impossible task essentially well he
saw it was difficult he thought it was
you know maybe if he'd live another 50
years he would have been able to do it I
don't know
you