Home / Teaching / Decolonising Education: From Theory to Practice / Interview with Dr Hemma Philamore on decolonisation of robotics and artificial intelligence
Interview with Dr Hemma Philamore on decolonisation of robotics and artificial intelligence
In this interview Dr Hemma Philamore covers the decolonisation of robotics and artificial intelligence.
i<00:00:10.880> think<00:00:11.120> it<00:00:11.200> means<00:00:11.440> sort<00:00:11.599> of<00:00:11.840> improving<00:00:12.240> the i think it means sort of improving the i think it means sort of improving the accessibility accessibility
accessibility of<00:00:14.080> robots<00:00:14.719> and<00:00:15.280> ai<00:00:15.920> and<00:00:16.320> particularly of robots and ai and particularly
of robots and ai and particularly thinking<00:00:18.320> about<00:00:19.199> these<00:00:19.439> as<00:00:19.600> things<00:00:20.000> that<00:00:20.160> are thinking about these as things that are thinking about these as things that are very very
very integrated<00:00:22.240> with<00:00:22.480> human<00:00:22.800> users<00:00:23.119> so<00:00:23.279> we're integrated with human users so we’re
integrated with human users so we’re starting<00:00:23.680> to<00:00:23.840> see<00:00:24.000> robots<00:00:24.400> sort<00:00:24.640> of<00:00:24.800> out<00:00:25.039> in starting to see robots sort of out in
starting to see robots sort of out in in<00:00:26.080> the<00:00:26.160> real<00:00:26.400> world<00:00:26.720> and<00:00:26.960> sort<00:00:27.119> of in the real world and sort of
in the real world and sort of interfacing<00:00:27.920> with<00:00:28.080> people<00:00:28.240> that<00:00:28.400> don't<00:00:28.640> build interfacing with people that don’t build
interfacing with people that don’t build and<00:00:29.119> operate and operate
and operate robots<00:00:30.800> so<00:00:31.760> in<00:00:32.160> decolonizing<00:00:32.880> ai<00:00:33.200> and robots so in decolonizing ai and
robots so in decolonizing ai and robotics<00:00:33.920> i<00:00:34.000> think robotics i think
robotics i think it’s<00:00:35.200> about<00:00:35.520> trying<00:00:35.760> to<00:00:35.920> remove<00:00:36.960> the<00:00:37.200> sort<00:00:37.360> of it’s about trying to remove the sort of
it’s about trying to remove the sort of inherent<00:00:39.200> bias<00:00:40.079> you<00:00:40.239> know<00:00:40.320> that<00:00:40.480> comes<00:00:40.879> from inherent bias you know that comes from
inherent bias you know that comes from colonialism<00:00:43.200> but<00:00:43.440> sort<00:00:43.600> of<00:00:43.760> in<00:00:43.840> the<00:00:43.920> human colonialism but sort of in the human
colonialism but sort of in the human creators<00:00:44.960> of creators of
creators of robots<00:00:46.000> and<00:00:46.320> ai<00:00:46.719> and<00:00:46.879> trying<00:00:47.200> to<00:00:47.440> sort<00:00:47.680> of robots and ai and trying to sort of
robots and ai and trying to sort of remove<00:00:48.480> those<00:00:48.719> from remove those from
remove those from systems<00:00:49.680> that<00:00:49.760> we're<00:00:49.920> hoping<00:00:50.239> will<00:00:50.960> go<00:00:51.199> out systems that we’re hoping will go out
systems that we’re hoping will go out and<00:00:51.600> be<00:00:51.840> used<00:00:52.160> by and be used by
and be used by people<00:00:52.960> who<00:00:53.199> are<00:00:53.280> not<00:00:53.440> the<00:00:53.520> developers people who are not the developers
people who are not the developers themselves<00:00:54.480> but<00:00:54.640> are<00:00:54.800> there<00:00:55.199> the<00:00:55.360> people themselves but are there the people
themselves but are there the people um<00:00:57.199> that<00:00:57.600> sort<00:00:57.760> of<00:00:58.079> from<00:00:58.239> now<00:00:58.399> onwards<00:00:58.800> will<00:00:58.879> be um that sort of from now onwards will be
um that sort of from now onwards will be using<00:01:00.719> ai<00:01:01.039> and<00:01:01.120> robotics<00:01:01.600> products these<00:01:11.119> are<00:01:11.439> systems<00:01:11.840> that<00:01:12.000> are<00:01:12.159> designed<00:01:12.640> to these are systems that are designed to these are systems that are designed to work work
work autonomously<00:01:15.040> so<00:01:15.119> they're<00:01:15.280> designed<00:01:15.600> to<00:01:15.759> work autonomously so they’re designed to work
autonomously so they’re designed to work without<00:01:16.560> any<00:01:16.880> sort<00:01:17.040> of<00:01:17.200> human without any sort of human
without any sort of human uh<00:01:18.560> supervisory<00:01:19.280> input<00:01:19.680> or<00:01:19.840> minimal<00:01:20.560> human uh supervisory input or minimal human
uh supervisory input or minimal human supervisory<00:01:21.439> input<00:01:21.600> that's<00:01:21.840> like<00:01:22.000> so<00:01:22.159> the supervisory input that’s like so the
supervisory input that’s like so the sort<00:01:22.479> of<00:01:22.560> ideal<00:01:22.880> that<00:01:22.960> you're<00:01:23.119> working sort of ideal that you’re working
sort of ideal that you’re working towards<00:01:23.759> with<00:01:23.920> something<00:01:24.320> where towards with something where
towards with something where it<00:01:25.280> is<00:01:25.759> um<00:01:26.640> you<00:01:26.799> know<00:01:27.040> sort<00:01:27.200> of<00:01:27.360> deemed<00:01:27.680> or it is um you know sort of deemed or
it is um you know sort of deemed or developed<00:01:28.240> to<00:01:28.320> have<00:01:28.479> intelligent<00:01:28.960> behavior developed to have intelligent behavior
developed to have intelligent behavior so<00:01:30.000> in<00:01:30.159> that<00:01:30.960> that<00:01:31.200> system<00:01:31.600> that<00:01:31.680> you<00:01:31.840> then<00:01:32.079> put so in that that system that you then put
so in that that system that you then put out<00:01:32.400> into<00:01:32.560> the<00:01:32.640> world<00:01:32.960> needs<00:01:33.200> to<00:01:33.360> be<00:01:33.439> able<00:01:33.600> to out into the world needs to be able to
out into the world needs to be able to interact<00:01:34.400> with<00:01:34.640> anyone<00:01:35.040> and<00:01:35.200> everyone<00:01:35.520> that interact with anyone and everyone that
interact with anyone and everyone that it<00:01:35.840> encounters<00:01:36.400> without<00:01:36.880> some<00:01:37.119> sort<00:01:37.280> of<00:01:37.439> human it encounters without some sort of human
it encounters without some sort of human in<00:01:37.920> the<00:01:38.079> loop in the loop
in the loop to<00:01:38.880> sort<00:01:39.040> of<00:01:39.119> correct<00:01:39.439> its<00:01:39.600> behavior<00:01:40.799> so<00:01:41.360> i to sort of correct its behavior so i
to sort of correct its behavior so i think<00:01:41.840> in<00:01:42.000> designing<00:01:42.560> systems<00:01:42.960> that<00:01:43.280> are think in designing systems that are
think in designing systems that are biased<00:01:44.479> in<00:01:44.560> their<00:01:44.720> effectiveness<00:01:45.439> towards biased in their effectiveness towards
biased in their effectiveness towards one<00:01:46.159> group<00:01:46.479> or<00:01:46.640> another one group or another
one group or another basically<00:01:48.399> means<00:01:48.640> that<00:01:48.880> you're<00:01:49.200> you're<00:01:49.759> you basically means that you’re you’re you
basically means that you’re you’re you know<00:01:50.000> increasing<00:01:50.399> the<00:01:50.479> probability<00:01:50.960> of<00:01:51.040> your know increasing the probability of your
know increasing the probability of your system<00:01:51.520> failing<00:01:51.840> once<00:01:52.079> it's<00:01:52.320> out<00:01:52.560> in system failing once it’s out in
system failing once it’s out in in<00:01:52.960> the<00:01:53.119> world<00:01:54.000> so<00:01:54.240> i<00:01:54.320> think<00:01:54.640> in<00:01:55.280> trying<00:01:55.600> to in the world so i think in trying to
in the world so i think in trying to remove<00:01:57.600> sources<00:01:58.000> of<00:01:58.159> inherent<00:01:58.640> bias<00:01:58.960> or remove sources of inherent bias or
remove sources of inherent bias or remove<00:01:59.360> the<00:01:59.439> effects<00:01:59.840> of<00:01:59.920> inherent<00:02:00.320> bias remove the effects of inherent bias
remove the effects of inherent bias from<00:02:02.240> you<00:02:02.399> know<00:02:02.560> developing<00:02:02.960> robots from you know developing robots
from you know developing robots developing<00:02:03.759> ai<00:02:04.079> we<00:02:04.240> were<00:02:04.399> increasing<00:02:04.799> the developing ai we were increasing the developing ai we were increasing the success success
success of<00:02:06.159> the<00:02:06.320> probability<00:02:06.719> of<00:02:06.880> it<00:02:06.960> being<00:02:07.759> a of the probability of it being a of the probability of it being a successful successful
successful uh<00:02:10.000> technology<00:02:11.599> once<00:02:11.920> it's<00:02:12.080> out<00:02:12.319> of<00:02:12.480> the<00:02:12.560> hands uh technology once it’s out of the hands
uh technology once it’s out of the hands of<00:02:12.959> the<00:02:13.200> the<00:02:13.360> people<00:02:13.599> that<00:02:13.840> develop<00:02:14.160> it<00:02:14.239> the of the the people that develop it the
of the the people that develop it the people<00:02:14.560> that<00:02:14.640> create<00:02:14.959> it<00:02:15.040> and<00:02:15.200> it's<00:02:15.360> put people that create it and it’s put
people that create it and it’s put through<00:02:15.840> that<00:02:16.080> sort<00:02:16.239> of<00:02:16.400> rigorous through that sort of rigorous
through that sort of rigorous um<00:02:18.640> test<00:02:18.879> that<00:02:19.120> is<00:02:19.280> having<00:02:19.599> to<00:02:19.760> work<00:02:20.239> in um test that is having to work in
um test that is having to work in a<00:02:21.360> sort<00:02:21.520> of<00:02:21.760> stochastic<00:02:22.720> unpredictable<00:02:23.680> real a sort of stochastic unpredictable real a sort of stochastic unpredictable real world world world setting i<00:02:35.040> think<00:02:35.280> the<00:02:35.440> biggest<00:02:35.760> one<00:02:36.720> is<00:02:36.879> probably i think the biggest one is probably
i think the biggest one is probably the<00:02:39.360> fact<00:02:39.519> that<00:02:39.680> we<00:02:39.920> often<00:02:40.400> forget<00:02:41.680> the<00:02:41.840> sort the fact that we often forget the sort the fact that we often forget the sort of of
of human<00:02:44.000> agency<00:02:44.480> in<00:02:44.640> developing<00:02:45.200> things<00:02:45.519> that human agency in developing things that
human agency in developing things that are<00:02:45.840> supposed<00:02:46.239> to<00:02:46.319> be<00:02:46.480> kind<00:02:46.640> of<00:02:46.879> standalone are supposed to be kind of standalone
are supposed to be kind of standalone agents<00:02:49.200> by<00:02:49.440> themselves<00:02:49.840> so<00:02:50.000> there's<00:02:50.160> often<00:02:50.400> a agents by themselves so there’s often a
agents by themselves so there’s often a perception<00:02:50.959> of perception of
perception of artificial<00:02:52.640> intelligence<00:02:53.280> that<00:02:53.440> it's<00:02:53.599> a<00:02:53.680> sort artificial intelligence that it’s a sort
artificial intelligence that it’s a sort of<00:02:54.160> black<00:02:54.560> box of black box
of black box and<00:02:56.239> that<00:02:56.480> it<00:02:56.959> it's<00:02:57.280> some<00:02:57.440> kind<00:02:57.680> of<00:02:57.760> sort<00:02:57.920> of and that it it’s some kind of sort of
and that it it’s some kind of sort of magic<00:02:58.560> that<00:02:58.720> does<00:02:58.959> its<00:02:59.200> own magic that does its own
magic that does its own its<00:03:00.239> own<00:03:00.480> thing<00:03:01.680> but<00:03:02.000> there's<00:03:02.480> the<00:03:02.720> human<00:03:03.040> that its own thing but there’s the human that
its own thing but there’s the human that developed<00:03:03.680> that developed that
developed that algorithm<00:03:05.360> and<00:03:06.000> all<00:03:06.319> of<00:03:06.480> their algorithm and all of their
algorithm and all of their you<00:03:07.840> know<00:03:08.239> and<00:03:08.800> sometimes<00:03:09.120> unintentional you know and sometimes unintentional
you know and sometimes unintentional bias<00:03:10.080> goes<00:03:10.319> into<00:03:10.480> the<00:03:10.560> creation<00:03:11.200> of<00:03:11.360> that bias goes into the creation of that bias goes into the creation of that algorithm algorithm
algorithm um<00:03:13.519> and<00:03:13.760> so<00:03:14.000> i<00:03:14.159> think<00:03:15.440> the<00:03:15.599> biggest um and so i think the biggest
um and so i think the biggest problem<00:03:18.400> or<00:03:18.560> one<00:03:18.720> of<00:03:18.800> the<00:03:18.879> greatest problem or one of the greatest problem or one of the greatest challenges challenges
challenges with<00:03:20.560> um<00:03:21.280> you<00:03:21.440> know<00:03:21.599> trying<00:03:21.840> to<00:03:22.000> implement with um you know trying to implement
with um you know trying to implement this<00:03:23.760> sort<00:03:23.920> of<00:03:24.000> removal<00:03:24.400> of<00:03:24.480> biases<00:03:25.040> it's<00:03:25.120> so this sort of removal of biases it’s so
this sort of removal of biases it’s so ingrained<00:03:25.840> sometimes<00:03:26.239> unconsciously<00:03:26.799> in<00:03:26.959> the ingrained sometimes unconsciously in the
ingrained sometimes unconsciously in the people<00:03:27.360> that<00:03:27.599> are people that are
people that are building<00:03:29.200> and<00:03:29.360> developing<00:03:29.840> new<00:03:30.000> systems<00:03:30.959> so building and developing new systems so
building and developing new systems so you<00:03:31.280> can<00:03:31.440> try<00:03:31.599> to<00:03:31.840> build<00:03:32.159> something you can try to build something
you can try to build something as<00:03:33.599> you<00:03:33.680> know<00:03:35.360> inclusive as you know inclusive
as you know inclusive to<00:03:37.200> everyone<00:03:37.680> around<00:03:38.000> you<00:03:38.400> as<00:03:38.560> you<00:03:38.799> like<00:03:39.120> but to everyone around you as you like but
to everyone around you as you like but that’s<00:03:39.680> very that’s very
that’s very much<00:03:40.560> influenced<00:03:41.040> by<00:03:41.200> your<00:03:41.360> own<00:03:41.519> perspective much influenced by your own perspective
much influenced by your own perspective of<00:03:42.159> who<00:03:42.720> everyone<00:03:43.200> is of who everyone is
of who everyone is so<00:03:44.799> people<00:03:45.120> outside<00:03:45.440> of<00:03:45.519> your<00:03:45.680> immediate so people outside of your immediate
so people outside of your immediate circle<00:03:46.879> of circle of
circle of the<00:03:47.280> people<00:03:47.440> that<00:03:47.599> you<00:03:47.680> operate<00:03:48.000> where<00:03:48.159> the the people that you operate where the
the people that you operate where the people<00:03:48.480> that<00:03:48.560> you<00:03:48.720> interact<00:03:49.120> with<00:03:49.840> might<00:03:50.000> not people that you interact with might not
people that you interact with might not even<00:03:50.480> occur<00:03:50.720> to<00:03:50.879> you even occur to you
even occur to you in<00:03:51.680> developing<00:03:52.159> something<00:03:52.560> and<00:03:52.720> so<00:03:53.200> your<00:03:53.519> own in developing something and so your own
in developing something and so your own sort<00:03:54.400> of<00:03:54.480> limited<00:03:54.879> sphere<00:03:55.280> of sort of limited sphere of
sort of limited sphere of interaction<00:03:57.840> can<00:03:58.799> if<00:03:58.879> you're<00:03:59.200> you<00:03:59.360> know interaction can if you’re you know
interaction can if you’re you know developing<00:03:59.840> one<00:04:00.000> of<00:04:00.080> these<00:04:00.239> things<00:04:00.560> can<00:04:00.720> then developing one of these things can then
developing one of these things can then sort<00:04:01.760> of<00:04:01.840> feed<00:04:02.080> forward<00:04:02.400> into<00:04:02.640> the<00:04:02.720> thing<00:04:02.879> that sort of feed forward into the thing that
sort of feed forward into the thing that you<00:04:03.200> develop<00:04:03.599> and<00:04:04.159> and<00:04:04.400> sort<00:04:04.560> of you develop and and sort of
you develop and and sort of i<00:04:05.120> don't<00:04:05.280> know<00:04:05.519> impart<00:04:05.920> your<00:04:06.159> bias<00:04:06.560> into<00:04:06.959> into that uh<00:04:16.959> there's<00:04:17.199> sort<00:04:17.359> of<00:04:17.440> the<00:04:17.519> teaching<00:04:17.919> parts<00:04:18.239> of uh there’s sort of the teaching parts of
uh there’s sort of the teaching parts of that<00:04:18.479> and<00:04:18.560> then<00:04:18.880> there's<00:04:19.120> the<00:04:19.199> kind<00:04:19.359> of<00:04:19.440> the that and then there’s the kind of the
that and then there’s the kind of the research<00:04:20.000> part research part
research part to<00:04:20.799> it<00:04:21.359> um<00:04:22.400> and<00:04:22.560> the<00:04:22.639> research<00:04:23.040> side<00:04:23.280> of<00:04:23.360> things to it um and the research side of things
to it um and the research side of things i<00:04:23.759> think<00:04:23.919> there's i think there’s
i think there’s a<00:04:24.639> lot<00:04:24.800> of<00:04:24.880> things<00:04:25.120> that<00:04:25.199> we're<00:04:25.360> doing<00:04:26.000> revolve a lot of things that we’re doing revolve
a lot of things that we’re doing revolve around<00:04:27.280> the<00:04:27.440> idea<00:04:27.680> of<00:04:27.759> co-authorship around the idea of co-authorship
around the idea of co-authorship and<00:04:28.639> co-creation<00:04:29.360> and<00:04:29.440> sort<00:04:29.680> of and co-creation and sort of
and co-creation and sort of participatory<00:04:30.639> research<00:04:31.199> where participatory research where
participatory research where we<00:04:32.000> try<00:04:32.160> to<00:04:32.400> involve<00:04:33.120> people<00:04:33.840> or<00:04:34.000> we<00:04:34.160> do we try to involve people or we do
we try to involve people or we do actively<00:04:34.880> involve<00:04:35.280> people actively involve people
actively involve people who<00:04:36.400> are<00:04:36.560> maybe<00:04:36.960> outside<00:04:37.680> of<00:04:38.080> the<00:04:38.240> typical who are maybe outside of the typical
who are maybe outside of the typical groups<00:04:39.040> that<00:04:39.280> we<00:04:39.600> would groups that we would
groups that we would test<00:04:40.320> things<00:04:40.720> on<00:04:41.440> in<00:04:41.600> the<00:04:41.759> lab<00:04:42.000> partly<00:04:42.400> because test things on in the lab partly because
test things on in the lab partly because test<00:04:42.880> participants<00:04:43.360> tend<00:04:43.520> to<00:04:43.600> be<00:04:43.680> whoever test participants tend to be whoever
test participants tend to be whoever else<00:04:44.240> it<00:04:44.400> is<00:04:44.479> in<00:04:44.560> the<00:04:44.639> lab<00:04:44.880> at<00:04:44.960> the<00:04:45.040> time<00:04:45.280> so else it is in the lab at the time so
else it is in the lab at the time so they’re<00:04:45.600> people<00:04:46.000> who they’re people who
they’re people who maybe<00:04:46.960> already<00:04:47.199> work<00:04:47.360> in<00:04:47.440> robotics<00:04:48.160> or<00:04:48.479> who maybe already work in robotics or who
maybe already work in robotics or who already<00:04:49.440> work<00:04:49.600> in<00:04:49.680> the<00:04:49.759> university<00:04:50.240> they're already work in the university they’re
already work in the university they’re not<00:04:50.560> they're<00:04:50.720> only<00:04:50.960> so<00:04:51.120> many<00:04:51.280> degrees<00:04:51.600> of not they’re only so many degrees of
not they’re only so many degrees of separation<00:04:52.240> from<00:04:52.479> who separation from who
separation from who we<00:04:53.280> are<00:04:53.440> as<00:04:53.520> the<00:04:53.840> developers<00:04:54.320> of<00:04:54.400> these<00:04:54.639> things we are as the developers of these things we are as the developers of these things so so
so taking<00:04:57.040> an<00:04:57.199> example<00:04:57.840> of<00:04:58.240> something<00:04:58.560> that taking an example of something that
taking an example of something that we’re<00:04:58.800> doing<00:04:59.520> recently<00:05:00.000> is<00:05:00.160> looking<00:05:00.479> at<00:05:00.639> it's we’re doing recently is looking at it’s
we’re doing recently is looking at it’s not<00:05:01.280> sort<00:05:01.440> of not sort of
not sort of directly<00:05:02.240> related<00:05:02.800> to<00:05:03.600> decolonization<00:05:04.479> this directly related to decolonization this
directly related to decolonization this particular<00:05:05.039> example<00:05:05.440> but<00:05:05.680> it's particular example but it’s
particular example but it’s you<00:05:06.080> know<00:05:06.320> it<00:05:06.400> could<00:05:06.560> be<00:05:06.720> sort<00:05:06.880> of<00:05:07.039> um<00:05:07.919> extended you know it could be sort of um extended you know it could be sort of um extended uh uh
uh in<00:05:09.440> that<00:05:09.600> direction<00:05:10.320> but<00:05:10.880> running in that direction but running
in that direction but running co-creation<00:05:12.000> workshops<00:05:12.560> for<00:05:12.720> development<00:05:13.360> of co-creation workshops for development of
co-creation workshops for development of uh<00:05:14.560> soft<00:05:14.800> robots<00:05:15.520> who<00:05:16.080> with<00:05:16.240> the<00:05:16.400> aim<00:05:16.560> of uh soft robots who with the aim of
uh soft robots who with the aim of having<00:05:17.039> these<00:05:17.280> robots<00:05:17.759> has having these robots has
having these robots has sort<00:05:18.240> of<00:05:18.400> robot<00:05:18.720> avatars<00:05:19.280> for<00:05:19.520> people<00:05:19.840> who sort of robot avatars for people who sort of robot avatars for people who can’t can’t
can’t be<00:05:22.400> in<00:05:22.560> a<00:05:22.639> physical<00:05:22.960> space<00:05:23.280> because<00:05:23.520> they<00:05:23.680> are be in a physical space because they are
be in a physical space because they are socially<00:05:24.560> isolating<00:05:25.120> because<00:05:25.280> they're socially isolating because they’re
socially isolating because they’re isolating<00:05:25.840> for<00:05:26.000> another<00:05:26.400> another<00:05:26.720> reason isolating for another another reason
isolating for another another reason and<00:05:28.240> in<00:05:28.560> that<00:05:28.720> we're<00:05:28.880> sort<00:05:29.039> of<00:05:29.120> mostly<00:05:29.440> looking and in that we’re sort of mostly looking
and in that we’re sort of mostly looking at<00:05:29.919> elderly<00:05:30.320> people at elderly people
at elderly people but<00:05:31.120> there's<00:05:31.360> no<00:05:31.520> reason<00:05:31.759> that<00:05:31.919> that<00:05:32.160> couldn't but there’s no reason that that couldn’t
but there’s no reason that that couldn’t be<00:05:32.880> another<00:05:33.919> another<00:05:34.240> group be another another group
be another another group and<00:05:35.120> actually<00:05:35.360> there's<00:05:35.520> been<00:05:35.759> a<00:05:35.840> few<00:05:36.479> similar and actually there’s been a few similar
and actually there’s been a few similar workshops<00:05:37.360> working<00:05:37.600> with<00:05:37.759> other<00:05:38.240> people<00:05:38.479> for workshops working with other people for
workshops working with other people for the<00:05:38.720> for<00:05:38.880> the<00:05:38.960> perspective<00:05:39.360> of<00:05:39.520> different the for the perspective of different
the for the perspective of different things<00:05:39.919> in<00:05:40.000> the<00:05:40.080> robotics<00:05:40.639> lab things in the robotics lab
things in the robotics lab so<00:05:41.280> the<00:05:41.440> ideas<00:05:41.759> of<00:05:42.000> smart<00:05:42.320> clothing<00:05:42.720> were<00:05:42.880> also so the ideas of smart clothing were also
so the ideas of smart clothing were also done<00:05:43.680> in<00:05:43.759> our<00:05:43.919> group<00:05:44.080> and<00:05:44.240> a<00:05:44.240> lot<00:05:44.400> of<00:05:44.479> those done in our group and a lot of those
done in our group and a lot of those ideas<00:05:45.120> came<00:05:45.360> from ideas came from
ideas came from sort<00:05:47.440> of<00:05:47.680> workshop<00:05:48.240> ideas<00:05:48.560> with<00:05:48.720> the sort of workshop ideas with the
sort of workshop ideas with the potential<00:05:49.199> end<00:05:49.440> users<00:05:49.840> of<00:05:50.160> this<00:05:50.320> type<00:05:50.479> of potential end users of this type of potential end users of this type of thing thing
thing so<00:05:51.520> there's<00:05:51.680> an<00:05:51.840> element<00:05:52.160> of<00:05:52.240> trying<00:05:52.479> to<00:05:52.560> bring so there’s an element of trying to bring
so there’s an element of trying to bring people<00:05:53.039> in<00:05:53.120> who<00:05:53.280> are<00:05:53.360> going<00:05:53.440> to<00:05:53.520> use<00:05:53.680> this people in who are going to use this
people in who are going to use this technology<00:05:54.320> in<00:05:54.400> an<00:05:54.560> early<00:05:54.800> stage<00:05:55.039> in<00:05:55.120> its technology in an early stage in its technology in an early stage in its development development
development on<00:05:56.560> in<00:05:57.039> research<00:05:58.240> and<00:05:58.319> then<00:05:58.560> in<00:05:58.720> teaching on in research and then in teaching
on in research and then in teaching um<00:05:59.919> so<00:06:00.160> for<00:06:00.319> example<00:06:00.720> the<00:06:00.800> module<00:06:01.120> that<00:06:01.360> i<00:06:01.759> i um so for example the module that i i
um so for example the module that i i teach<00:06:02.240> on teach on
teach on mathematical<00:06:03.199> data<00:06:03.440> modeling<00:06:04.479> so<00:06:05.280> before<00:06:05.600> i mathematical data modeling so before i
mathematical data modeling so before i talk<00:06:05.919> about<00:06:06.160> that talk about that
talk about that i<00:06:07.120> think<00:06:07.360> it<00:06:07.520> can<00:06:07.759> be<00:06:07.919> really<00:06:08.160> difficult<00:06:08.639> to i think it can be really difficult to
i think it can be really difficult to try<00:06:09.039> and<00:06:09.199> identify<00:06:09.919> how<00:06:10.160> to try and identify how to
try and identify how to decolonize<00:06:11.520> taurt<00:06:11.840> subjects<00:06:12.479> in<00:06:12.800> stem decolonize taurt subjects in stem decolonize taurt subjects in stem particularly particularly
particularly partly<00:06:14.160> because<00:06:14.400> there<00:06:14.560> is<00:06:14.720> so<00:06:14.880> much<00:06:15.120> emphasis partly because there is so much emphasis
partly because there is so much emphasis on<00:06:16.639> highly<00:06:17.039> or<00:06:17.199> there's<00:06:17.440> so<00:06:17.600> much<00:06:17.759> focus on highly or there’s so much focus
on highly or there’s so much focus necessarily<00:06:19.039> on<00:06:19.280> highly<00:06:19.600> technical<00:06:20.800> subjects necessarily on highly technical subjects
necessarily on highly technical subjects and<00:06:21.520> topics and topics
and topics and<00:06:22.800> the<00:06:22.960> teaching<00:06:23.840> is<00:06:24.000> sometimes and the teaching is sometimes
and the teaching is sometimes you<00:06:25.759> know<00:06:25.919> potentially<00:06:26.319> sort<00:06:26.400> of<00:06:26.560> agnostic<00:06:27.280> to you know potentially sort of agnostic to
you know potentially sort of agnostic to um<00:06:28.240> to um to
um to social<00:06:30.639> influence<00:06:31.120> or<00:06:31.199> social<00:06:31.520> issues social influence or social issues
social influence or social issues but<00:06:32.880> in<00:06:33.199> subjects<00:06:33.759> like<00:06:34.319> uh<00:06:34.639> mathematical but in subjects like uh mathematical
but in subjects like uh mathematical data<00:06:35.520> modeling<00:06:35.840> which<00:06:36.000> is<00:06:36.160> a<00:06:36.240> unit<00:06:36.479> that<00:06:36.560> i data modeling which is a unit that i
data modeling which is a unit that i teach<00:06:36.960> on teach on
teach on at<00:06:37.759> bristol<00:06:38.800> in<00:06:38.960> that<00:06:39.120> case<00:06:39.919> we're<00:06:40.080> looking<00:06:40.400> at at bristol in that case we’re looking at
at bristol in that case we’re looking at some<00:06:41.680> constant<00:06:42.080> focus<00:06:42.479> of<00:06:42.800> the<00:06:42.960> units<00:06:43.280> to some constant focus of the units to
some constant focus of the units to bring<00:06:43.680> in<00:06:44.240> problems<00:06:44.639> from<00:06:44.880> outside<00:06:45.280> of bring in problems from outside of
bring in problems from outside of engineering<00:06:46.720> maths<00:06:46.960> and<00:06:47.039> then<00:06:47.280> in<00:06:47.360> the<00:06:47.440> best engineering maths and then in the best
engineering maths and then in the best case<00:06:47.759> study<00:06:48.080> outside<00:06:48.319> of<00:06:48.400> the<00:06:48.560> university case study outside of the university
case study outside of the university completely<00:06:50.479> um<00:06:50.720> so<00:06:50.880> these<00:06:51.039> are<00:06:51.199> kind<00:06:51.280> of completely um so these are kind of completely um so these are kind of data-driven data-driven
data-driven problems<00:06:53.120> either<00:06:53.360> on<00:06:53.520> sort<00:06:53.680> of<00:06:53.759> physical problems either on sort of physical
problems either on sort of physical modeling<00:06:54.319> or<00:06:54.479> data<00:06:54.720> modeling<00:06:55.360> um<00:06:55.599> or<00:06:55.840> other modeling or data modeling um or other
modeling or data modeling um or other types<00:06:57.199> of<00:06:57.840> mathematical<00:06:58.319> modeling<00:06:58.560> but types of mathematical modeling but
types of mathematical modeling but looking<00:06:58.960> at<00:06:59.120> translating<00:06:59.680> a<00:06:59.759> real<00:06:59.919> world looking at translating a real world looking at translating a real world issue issue
issue into<00:07:01.680> a<00:07:01.759> model<00:07:02.080> to<00:07:02.240> try<00:07:02.479> and<00:07:02.800> answer<00:07:03.120> some<00:07:03.759> some into a model to try and answer some some
into a model to try and answer some some open<00:07:04.240> questions<00:07:04.960> about<00:07:05.280> it open questions about it
open questions about it so<00:07:05.759> in<00:07:05.840> that<00:07:06.080> we're<00:07:06.240> trying<00:07:06.479> to<00:07:06.639> get<00:07:06.880> students so in that we’re trying to get students
so in that we’re trying to get students to<00:07:07.759> well<00:07:07.919> first<00:07:08.160> of<00:07:08.240> all<00:07:08.319> making<00:07:08.560> this<00:07:08.720> an to well first of all making this an
to well first of all making this an assessed<00:07:09.280> part<00:07:09.520> of<00:07:09.599> the assessed part of the
assessed part of the course<00:07:10.560> so<00:07:10.720> we're<00:07:10.960> making<00:07:11.199> it<00:07:11.280> compulsory<00:07:11.840> for course so we’re making it compulsory for
course so we’re making it compulsory for students<00:07:12.479> to students to
students to try<00:07:13.759> to<00:07:13.919> identify<00:07:14.800> and<00:07:15.199> make<00:07:15.440> recommendations try to identify and make recommendations
try to identify and make recommendations to<00:07:16.160> mitigate<00:07:16.639> those<00:07:16.800> sources<00:07:17.199> of<00:07:17.280> bias to mitigate those sources of bias
to mitigate those sources of bias in<00:07:18.319> the<00:07:18.400> data<00:07:18.720> that<00:07:18.800> they're<00:07:18.960> working<00:07:19.280> with in the data that they’re working with in the data that they’re working with and and
and so<00:07:21.039> in<00:07:21.199> that<00:07:21.360> you<00:07:21.440> know<00:07:21.759> it's<00:07:22.000> it's<00:07:22.160> not<00:07:22.479> a<00:07:23.120> sort so in that you know it’s it’s not a sort so in that you know it’s it’s not a sort of of
of trivial<00:07:24.319> thing<00:07:24.560> at<00:07:24.720> all<00:07:24.880> that's<00:07:25.120> a<00:07:25.280> very trivial thing at all that’s a very trivial thing at all that’s a very significant significant
significant thing<00:07:28.160> that's<00:07:28.319> going<00:07:28.479> to<00:07:28.560> have<00:07:28.720> this<00:07:28.880> sort<00:07:29.039> of thing that’s going to have this sort of
thing that’s going to have this sort of measurable<00:07:29.599> impact<00:07:30.080> on<00:07:30.639> not<00:07:30.800> only<00:07:31.039> the measurable impact on not only the
measurable impact on not only the the<00:07:31.759> success<00:07:32.160> of<00:07:32.240> the<00:07:32.400> students<00:07:32.800> project<00:07:33.280> in the success of the students project in
the success of the students project in that<00:07:33.680> course<00:07:34.080> but<00:07:34.240> then that course but then
that course but then you<00:07:34.720> know<00:07:34.880> they're<00:07:35.039> going<00:07:35.120> to<00:07:35.199> be<00:07:35.919> a<00:07:36.000> data you know they’re going to be a data
you know they’re going to be a data engineer<00:07:36.720> for<00:07:36.960> example engineer for example
engineer for example in<00:07:38.319> in<00:07:38.400> their<00:07:38.560> future<00:07:38.880> careers<00:07:39.199> of<00:07:39.280> the<00:07:39.360> future in in their future careers of the future
in in their future careers of the future products<00:07:39.919> that<00:07:40.000> they're<00:07:40.160> developing<00:07:40.639> in<00:07:40.720> the products that they’re developing in the
products that they’re developing in the you<00:07:41.120> know<00:07:41.280> future<00:07:41.599> work<00:07:41.840> that<00:07:41.919> they're<00:07:42.080> doing you know future work that they’re doing
you know future work that they’re doing so<00:07:43.280> i<00:07:43.440> think so i think
so i think there’s<00:07:44.960> these<00:07:45.199> two<00:07:45.680> those<00:07:45.919> are<00:07:46.479> two<00:07:47.199> two there’s these two those are two two
there’s these two those are two two ongoing<00:07:48.400> things<00:07:48.639> that<00:07:48.720> i<00:07:48.800> think<00:07:49.039> are<00:07:49.120> working ongoing things that i think are working
ongoing things that i think are working quite<00:07:49.520> successfully<00:07:50.080> at<00:07:50.160> the<00:07:50.240> moment<00:07:50.639> is quite successfully at the moment is
quite successfully at the moment is you<00:07:51.199> know<00:07:51.360> these<00:07:51.520> two<00:07:51.759> ideas<00:07:52.080> of<00:07:52.160> bringing<00:07:52.479> in you know these two ideas of bringing in
you know these two ideas of bringing in people<00:07:53.039> who<00:07:53.199> are<00:07:53.360> going<00:07:53.520> to<00:07:53.680> use<00:07:53.919> this people who are going to use this people who are going to use this technology technology
technology at<00:07:55.360> an<00:07:55.440> early<00:07:55.680> stage<00:07:56.000> who<00:07:56.160> are<00:07:56.240> outside<00:07:56.639> of<00:07:56.800> i at an early stage who are outside of i
at an early stage who are outside of i direct<00:07:57.759> you<00:07:57.919> know direct you know
direct you know sort<00:07:58.639> of<00:07:59.919> direct<00:08:00.240> networks<00:08:01.199> and<00:08:01.280> then sort of direct networks and then
sort of direct networks and then building<00:08:02.400> in<00:08:03.199> sort<00:08:03.360> of<00:08:03.840> building<00:08:04.240> into<00:08:04.400> the building in sort of building into the building in sort of building into the curriculum curriculum
curriculum non-trivial<00:08:06.879> elements<00:08:07.440> that<00:08:08.000> make<00:08:08.240> students non-trivial elements that make students
non-trivial elements that make students think<00:08:08.800> about think about
think about why<00:08:09.680> bias<00:08:10.240> by<00:08:10.560> why<00:08:10.720> removing<00:08:11.199> sources<00:08:11.520> of<00:08:11.599> bias why bias by why removing sources of bias
why bias by why removing sources of bias is<00:08:12.639> important<00:08:13.840> not<00:08:14.000> only<00:08:14.160> for<00:08:14.319> the<00:08:14.400> purpose<00:08:14.720> of is important not only for the purpose of
is important not only for the purpose of getting<00:08:15.120> a<00:08:15.199> good<00:08:15.440> mark<00:08:15.680> or<00:08:15.840> being<00:08:16.000> a<00:08:16.080> good getting a good mark or being a good getting a good mark or being a good person person
person but<00:08:17.520> for<00:08:18.080> you<00:08:18.160> know<00:08:18.319> the<00:08:18.479> success<00:08:18.879> of<00:08:18.960> what but for you know the success of what
but for you know the success of what they’re<00:08:19.280> doing<00:08:19.599> and<00:08:19.680> how<00:08:20.240> that they’re doing and how that
they’re doing and how that works<00:08:21.440> most<00:08:21.680> effectively<00:08:22.240> in<00:08:22.639> the<00:08:22.720> real<00:08:24.840> world yeah<00:08:32.240> i<00:08:32.320> think<00:08:32.479> it's<00:08:32.560> a<00:08:32.640> good<00:08:32.880> question<00:08:33.279> it's yeah i think it’s a good question it’s
yeah i think it’s a good question it’s something<00:08:33.760> that<00:08:33.919> was<00:08:34.320> um something that was um
something that was um talked<00:08:35.599> about<00:08:35.919> in<00:08:36.000> a<00:08:36.159> paper<00:08:36.479> that<00:08:36.640> you<00:08:36.719> shared talked about in a paper that you shared
talked about in a paper that you shared with<00:08:37.120> me<00:08:37.760> uh<00:08:38.159> not<00:08:38.320> very<00:08:38.560> long<00:08:38.719> ago<00:08:38.959> and<00:08:39.039> it's<00:08:39.200> a with me uh not very long ago and it’s a
with me uh not very long ago and it’s a really<00:08:39.519> interesting really interesting
really interesting um<00:08:42.080> and<00:08:42.240> quite<00:08:42.479> difficult<00:08:42.959> question<00:08:43.279> of<00:08:43.440> if um and quite difficult question of if
um and quite difficult question of if you<00:08:43.760> give you give
you give a<00:08:44.800> robot<00:08:45.519> sort<00:08:45.680> of<00:08:46.240> an<00:08:46.399> implied<00:08:47.360> human a robot sort of an implied human
a robot sort of an implied human race<00:08:49.279> to<00:08:49.440> the<00:08:49.519> potential<00:08:50.080> sort<00:08:50.240> of<00:08:50.320> social race to the potential sort of social race to the potential sort of social implications implications
implications of<00:08:52.720> you<00:08:52.800> know<00:08:52.959> that<00:08:53.120> apply<00:08:53.360> to<00:08:53.440> humans<00:08:53.839> then of you know that apply to humans then
of you know that apply to humans then then<00:08:54.480> apply<00:08:54.959> applies<00:08:55.200> to<00:08:55.279> the<00:08:55.360> robot<00:08:55.839> so then apply applies to the robot so
then apply applies to the robot so a<00:08:56.480> lot<00:08:56.640> of<00:08:56.800> robots<00:08:57.440> humanoid<00:08:57.839> robots<00:08:58.480> are<00:08:58.800> sort a lot of robots humanoid robots are sort a lot of robots humanoid robots are sort of of
of made<00:09:00.240> to<00:09:00.399> look<00:09:00.959> made<00:09:01.200> to<00:09:01.279> look<00:09:01.519> white<00:09:02.399> um<00:09:02.880> and made to look made to look white um and
made to look made to look white um and there’s<00:09:03.200> a<00:09:03.279> lot<00:09:03.440> of<00:09:03.519> you<00:09:03.600> know<00:09:03.680> that's<00:09:03.920> that's there’s a lot of you know that’s that’s
there’s a lot of you know that’s that’s problematic<00:09:04.640> for<00:09:04.800> a<00:09:04.880> lot<00:09:04.959> of<00:09:05.040> different problematic for a lot of different
problematic for a lot of different reasons<00:09:05.760> but<00:09:05.839> there's<00:09:06.080> also<00:09:06.839> this reasons but there’s also this
reasons but there’s also this thing<00:09:08.880> in<00:09:09.360> robotics<00:09:09.920> the<00:09:10.000> uncanny<00:09:10.480> valley thing in robotics the uncanny valley
thing in robotics the uncanny valley where<00:09:10.959> people<00:09:11.279> have<00:09:11.519> this<00:09:11.760> kind<00:09:12.000> of where people have this kind of
where people have this kind of rejection<00:09:13.440> of<00:09:13.600> things<00:09:13.839> that<00:09:13.920> look<00:09:14.160> too rejection of things that look too
rejection of things that look too similar<00:09:14.720> to<00:09:14.880> humans<00:09:15.279> this<00:09:15.440> sort<00:09:15.600> of similar to humans this sort of
similar to humans this sort of where<00:09:16.240> things<00:09:16.640> become<00:09:17.040> you<00:09:17.120> know<00:09:17.279> more<00:09:17.440> and where things become you know more and
where things become you know more and more<00:09:17.760> and<00:09:17.839> more<00:09:18.000> familiar<00:09:18.720> and<00:09:18.880> people<00:09:19.200> sort more and more familiar and people sort
more and more familiar and people sort of<00:09:19.440> warms<00:09:19.680> them<00:09:19.839> and<00:09:19.920> then<00:09:20.080> this<00:09:20.320> massive of warms them and then this massive
of warms them and then this massive drop-off<00:09:20.959> where<00:09:21.120> things<00:09:21.360> become<00:09:21.680> so<00:09:21.920> close<00:09:22.320> to drop-off where things become so close to
drop-off where things become so close to actually<00:09:23.920> resembling<00:09:24.560> humans<00:09:24.959> that<00:09:25.040> they actually resembling humans that they
actually resembling humans that they become<00:09:25.440> creepy<00:09:25.839> and<00:09:26.000> people<00:09:26.160> don't<00:09:26.399> actually become creepy and people don’t actually
become creepy and people don’t actually want<00:09:26.880> to want to
want to to<00:09:27.279> interact<00:09:27.680> with<00:09:27.760> them<00:09:28.399> so<00:09:29.040> i<00:09:29.200> don't<00:09:29.680> think to interact with them so i don’t think
to interact with them so i don’t think there<00:09:30.160> is<00:09:30.320> necessarily there is necessarily
there is necessarily a<00:09:31.920> need<00:09:32.560> for<00:09:33.680> robots a need for robots
a need for robots i<00:09:35.040> mean<00:09:35.279> for<00:09:35.440> robots<00:09:35.839> to<00:09:36.000> resemble<00:09:37.120> humans i mean for robots to resemble humans
i mean for robots to resemble humans necessarily<00:09:38.480> in<00:09:38.640> a<00:09:38.720> lot<00:09:38.880> of<00:09:39.040> different<00:09:39.360> tasks necessarily in a lot of different tasks
necessarily in a lot of different tasks and<00:09:39.839> there's<00:09:39.920> a<00:09:40.000> lot<00:09:40.080> of<00:09:40.160> different<00:09:40.399> tasks and there’s a lot of different tasks
and there’s a lot of different tasks where<00:09:40.800> a<00:09:40.880> humanoid<00:09:41.279> robot<00:09:41.680> is<00:09:41.760> not where a humanoid robot is not
where a humanoid robot is not necessarily<00:09:42.399> the<00:09:42.640> best necessarily the best
necessarily the best suited<00:09:44.080> you<00:09:44.160> know<00:09:44.399> morphology<00:09:45.040> to<00:09:45.519> to<00:09:45.760> achieve suited you know morphology to to achieve
suited you know morphology to to achieve whatever<00:09:46.399> task<00:09:46.640> it's<00:09:46.800> trying<00:09:46.959> to<00:09:47.120> do whatever task it’s trying to do
whatever task it’s trying to do i<00:09:48.480> don't<00:09:48.640> think<00:09:48.800> we<00:09:48.959> necessarily<00:09:49.440> need<00:09:49.600> to i don’t think we necessarily need to
i don’t think we necessarily need to build<00:09:49.920> in<00:09:50.080> race<00:09:50.320> as<00:09:50.399> a<00:09:50.480> human<00:09:50.720> character build in race as a human character build in race as a human character characteristic characteristic
characteristic into<00:09:52.720> robots<00:09:53.519> the<00:09:53.680> same<00:09:53.920> way<00:09:54.160> that<00:09:54.480> there's<00:09:54.640> a into robots the same way that there’s a
into robots the same way that there’s a lot<00:09:54.880> of<00:09:55.040> other<00:09:55.200> human<00:09:55.440> characteristics<00:09:56.240> that lot of other human characteristics that
lot of other human characteristics that are<00:09:57.519> potentially<00:09:58.000> becoming<00:09:58.480> more<00:09:58.640> and<00:09:58.800> more are potentially becoming more and more are potentially becoming more and more unnecessary unnecessary
unnecessary in<00:10:00.320> in<00:10:00.480> robots<00:10:00.959> as<00:10:01.120> we<00:10:01.920> focus<00:10:02.320> on in in robots as we focus on
in in robots as we focus on the<00:10:03.760> things<00:10:04.000> that<00:10:04.160> make<00:10:04.320> them<00:10:04.480> best<00:10:04.800> suited<00:10:05.279> to the things that make them best suited to
the things that make them best suited to to<00:10:06.000> a<00:10:06.079> task to a task
to a task if<00:10:06.720> we<00:10:06.880> look<00:10:07.040> at<00:10:07.120> caring<00:10:07.440> robots<00:10:07.920> they<00:10:08.800> don't if we look at caring robots they don’t
if we look at caring robots they don’t necessarily<00:10:09.680> look necessarily look
necessarily look exactly<00:10:10.720> like<00:10:10.959> humans<00:10:11.360> in<00:10:11.519> fact<00:10:11.680> they<00:10:11.839> look exactly like humans in fact they look
exactly like humans in fact they look kind<00:10:12.320> of<00:10:12.959> far<00:10:13.200> from<00:10:13.440> it kind of far from it
kind of far from it and<00:10:15.279> if<00:10:15.519> we<00:10:15.680> look<00:10:15.839> at<00:10:16.000> things<00:10:16.480> like and if we look at things like
and if we look at things like robots<00:10:17.360> that<00:10:17.440> are<00:10:17.600> used<00:10:18.240> to<00:10:18.800> try<00:10:19.120> to robots that are used to try to
robots that are used to try to have<00:10:21.519> a<00:10:21.600> sort<00:10:21.839> of<00:10:22.880> a<00:10:23.040> calming<00:10:23.519> or<00:10:23.680> a<00:10:23.760> friendly have a sort of a calming or a friendly
have a sort of a calming or a friendly or<00:10:24.320> a<00:10:24.480> kind<00:10:24.640> of or a kind of
or a kind of companion<00:10:26.079> type<00:10:26.320> relationship<00:10:26.880> with<00:10:26.959> a<00:10:27.040> human companion type relationship with a human
companion type relationship with a human they’re<00:10:27.440> often<00:10:27.600> sort<00:10:27.760> of<00:10:28.160> animal they’re often sort of animal
they’re often sort of animal robots<00:10:29.600> animatronic<00:10:30.079> robots<00:10:30.399> that<00:10:30.480> look<00:10:30.720> more robots animatronic robots that look more
robots animatronic robots that look more like<00:10:31.360> a<00:10:31.440> pet<00:10:31.680> for<00:10:31.839> example<00:10:32.240> than like a pet for example than
like a pet for example than a<00:10:33.519> human<00:10:38.279> being
[See bottom of page for full transcript of this video]
Mitigation for inherent bias in robotics and AI, including bias with its origin in colonialism is an important part of teaching and research because of its tangible and measurable effect on the success of robots and autonomous systems deployed in the real world. The agency of the human developer and the influence of any bias they may hold, intentional or unintentional, on the systems they work on is often overlooked. Failing to identify and act on this can negatively affect the success of robots and algorithms when deployed in the real world and can result in outcomes that discriminate against certain users.
Ongoing strategies to counter this at University of Bristol include initiatives in both research and teaching. In robotics research for example, we are engaging communities that are often overlooked in engineering in participatory research to co-author robot behaviours early on in the development process. In teaching we are introducing components of assignments that require students to identify and suggest mitigations for sources of discriminatory bias in their mathematical modelling and algorithms. Perhaps most importantly, we recognise that we are not coming from the perspective of a system that has got it right. By stimulating students to consider sources of bias in their work, our goal is to nurture the development of the next generation of engineers as individuals capable of developing solutions to current problems of discrimination in robotics and AI.
This article is from the online course:
Decolonising Education: From Theory to Practice
Created by
This article is from the free online
Decolonising Education: From Theory to Practice
Reach your personal and professional goals
Unlock access to hundreds of expert online courses and degrees from top universities and educators to gain accredited qualifications and professional CV-building certificates.
Join over 18 million learners to launch, switch or build upon your career, all at your own pace, across a wide range of topic areas.
