Select text to annotate, Click play in YouTube to begin
00:00:04
[Music] hello out there thinkers linkers and mappers we are really happy to host a series of conversations around the topic of tools for thinking our longer term goal is to spark a diverse connected
00:00:21
shared memory that will help us make important decisions together as humans our near-term goal with these podcasts is to blow more oxygen on the growing tools for thinking sector addressing key issues and talking with the people who
00:00:33
are doing the work this podcast is created by Beta Works a New York city-based startup Studio I'm Jerry mikulski your interlocutor and obsessive mind mapper our topic today is maps and creativity
00:00:45
our guests are Anne Laura la conf the founder of Nest labs and Lorenzo bernaskina the founder of gems notes and we have a juicy wonderful Topic in front
00:00:58
of us I would love if you could maybe introduce yourselves sort of your relation to the topic and then we'll head on in on Laura if you would go first that'd be great yes sir thanks for having us
00:01:12
um so as you said I'm the founder of Nest labs and on our website we've published I think now more than 50 interviews with Founders and creators of
00:01:23
tools for thought so I haven't built my own like Lorenzo did but I I know a little bit about them and some of the ones I like the most and I find the most
00:01:36
powerful are the kinds of tools that allow you to easily connect your ideas in a visual way so that's why I really like this topic and I'm super excited to talk about it more today love that thank you uh Lorenzo yeah
00:01:49
thank you hat for having me as well and I'm Lorenzo I'm the founder of gems which is a self-organizing whiteboard to Cluster and connect ideas with the help of artificial intelligence and it's actually interesting because I think
00:02:01
that the work of unlore and Nest Labs really influenced this since the very beginning so it's exciting to be here and discussing about this together today I love that Lonzo can you just pick up
00:02:13
on that what you just said and and talk about the influence and and how you thought about what you're creating and just bring us into uh what gems notes is doing sure I think the trigger wasn't the
00:02:26
during the pandemic so I started I was working as a software engineer and I was doing a master in machine learning and artificial intelligence um and I had to pick the topic for the
00:02:39
for the thesis in the middle of the pandemic and of course I thought about something uh helpful in in that context right and so I built another thing to basically have researchers find meaningful information there was a
00:02:52
lot of research and stuff we didn't really know anything about the disease at the time so we're really like in a rush to really understand what was going on with the disease and there were a lot of papers and knowledge shared every day and so there was clearly a problem to
00:03:05
figure out you know like what's going on here what are the the relevant information to really move our understanding forward and so it builds a tool in that space and at the time I realized there was so much noise right it was hard to actually
00:03:17
get to to the insides to the gems that's where the name of the tool comes from um and so that was the kind of trigger for me and I remember at the time there was this course that on Laura did about collectors to creators really sounded
00:03:31
interesting because you know it was this concept of moving from this kind of like yes we have a lot of interesting stuff and and things but how can we actually make sense of them and and do something with them understand or create something
00:03:44
so that was the trigger I got into the course was amazing met a bunch of interesting people and that that was the start how I started thinking about this thing and interacting with people about the topic yeah love that
00:03:57
um and Laura how did the pandemic affect what you were thinking about or doing in this space I think it's amazing that you took the course and that was one of your sources of inspiration I didn't know that so that's great which is also why
00:04:10
it's funny that we crossed path during that course which was in 2020 because I started teaching online during the pandemic also because so I had started a newsletter uh six months before the
00:04:23
pandemic started and it was very much a one-way type of broadcasting relationship that I had with my readers every week I would just send someone something they would read it and that was basically it I did have at the end
00:04:36
of my newsletter a little note saying please reply uh hit reply and say hello if you have anything that you want to share back with me and in March 2020 all of a sudden I got lots of replies a lot
00:04:48
more than I used to have before and people were telling me how lonely they were feeling how confused they were feeling how overwhelmed they were with all of that contradictory information
00:05:00
around them with the fact that they didn't have any mechanism to keep on hanging out with colleagues brainstorm ideas figuring out things together that
00:05:13
it felt very lonely and isolated and so this is why at the very beginning of the pandemic I figured okay what if I try to shift from this one-way broadcasting relationship towards something where I
00:05:26
become more of a facilitator allowing people to connect together and learn from each other how do I ensure that I'm not a barrier preventing them from growing and I you know kind of like yeah
00:05:40
create as behave as a connector basically so I created The Nest Labs community and I then decided to turn a lot of the articles that I wrote into an online course which was called as
00:05:52
Lorenzo said from collector to creator basically teaching people who were naturally curious already who are collecting a lot of ideas but not doing much with them how to generate their own
00:06:04
insights from data and information that they had gathered through their reading or listening to podcasts and all of these kind of things and becoming creators themselves and starting contributing to the public discourse themselves um love that love also how how sort of
00:06:19
our paths all were changed and maybe intersected more at the start of pandemic um because coincident with the start of pandemic I got into a couple of conversations was introduced to a cons guy who runs a consultancy in Boston we
00:06:32
started talking and that turned into US inviting our communities into a series of calls that we call open Global mind which is now a community of practice and has a bunch of a bunch of different standing calls every week where some one
00:06:43
where Geeks two were Geeks meet and one where the whole Community sort of comes together and also there's been just this Rich sharing of of information and methods and what we're seeing and what we're doing and it's lovely
00:06:57
um so I don't know sometimes there's Silver Linings to disasters and and if the pandemic has some Silver Linings maybe it's that including the pressure and the need to start identifying causes and solutions to the pandemic itself
00:07:09
like you know comparing different countries uh policy responses to it and so forth um uh let me take a go sort of take a step back for a second and um and Laura
00:07:22
you wrote a post a while ago called thinking in maps from the Lasko caves to knowledge graphs in which one of the things you said was that thinking in Maps is really different from thinking in sentences can you sort of add some
00:07:36
water to that that thought from back when and bring it back into this conversation for us yes and it's uh kind of relevant to what you were saying about uh making sense of a crisis and all of that contradictory
00:07:48
information when we think in sentences we tend to think in a linear way we go from a and then to B and then to C and we kind of think in the way that fills the most intuitive and logical in the
00:08:02
moment and that leads you very often to results that sound like Common Sense the problem is that and we've seen that during the pandemic Common Sense doesn't mean that you're right it's not because
00:08:15
a lot of people think the same thing and it feels like the most obvious answer that it is the correct answer when you think in maps you are not trying to get from A to B as quickly as
00:08:27
possible to get an answer as quickly as possible you're basically building a playground for emergence you're adding nodes and you're not trying to find an answer by looking at the interesting
00:08:40
quantities or the information in each nodes but you're trying to look at how they work together how they connect together how they influence each other and in this way you can start seeing
00:08:52
patterns that you may not be able to see if you're right to your nose just in front of that piece of information and not seeing it in context you're not going to be able to get the same kind of answers as if you're looking at the
00:09:03
whole map so thinking in maps is slower in general though I'm super curious to hear more about what Lorenzo is doing with AI because one of the reasons why it's much slower at the moment it's
00:09:16
because as a human you do have to go all through all of these steps during all of the pattern recognition it's not that easy to do but you do get better ideas and better insights by doing this
00:09:28
um so that's what thinking in that and that's why it's different from thinking in sentences um thank you Lorenzo do you want to how do you connect to that yeah I I think really like we have always like thought
00:09:41
like mobs are a very powerful tools for thought right we have used them since the forever and so when we talk about the space about tools for thought I think it would be very interesting uh I
00:09:54
thought at the time when it started it would be very interesting to um find a way to augment them right uh and and see how to actually increase our cognitive Powers with them
00:10:08
um and I think that AI is um you know today is starting to getting powerful enough to support this kind of ambitious processes and the challenge there is really come up with an
00:10:21
interface that helps allows us to interact with it in a meaningful way and that sounds to be very uh very interesting as a as a as an interface to do that because again it's something that it's very used to we use
00:10:34
but um you know there are some challenges as on lore said it's really like manual and it's hard so definitely the goal here I would say is use AI to augment this process and help you find those
00:10:46
relationships and clusters faster it's really interesting because one of the things I've made explicit and talked about is people ask me like I map using the brain software which I I was on their first press tour 25 years ago in
00:11:00
another couple weeks it'll be December and it'll actually be a quarter Century that I've been feeding one mind map and one of my observations from doing so is that whenever I add something to this map which I do probably 50 times a day
00:11:12
it throws me into system two thinking from kahneman's book Thinking Fast and Slow where usually and I have this feeling that when we're busy retweeting things that we see in the Tweet stream or a light gesture of yeah this is cool
00:11:26
I should forward this to my people is system one it's very reflexive sort of thinking but but you know very explicitly when I when I put something in my brain is it worth remembering yes where do I put it what is it similar to
00:11:40
like what do I hang it on because I don't I don't just drop things in this space I'm it's like hanging ornaments on a Christmas tree I'm always building context right so where does it go what should I call it um what else is it connected to what
00:11:53
more can I learn right and I do that Loop really quickly 50 times a day just to sort of to feed the brain but then there's a different thing that's going on which is the visual context side of it which is
00:12:04
different from system one system two thinking there's something else that happens when you start to see evidence near near each other it's a little bit it's the reason why when we do crime movies we have murder boards where they put like pictures of all the suspects up
00:12:16
and you have pins in them with like yarn connecting them and all that um and I only learned recently that those are called Murder boards which is kind of a little eerie but cool um but there's something else that
00:12:28
happens with the visual thinking and then Lorenzo I'm really interested in understanding better if something else happens now when you bring machine learning and other other kinds of intelligence into that context into that setting and there's lots of different
00:12:40
things that that AIML could be doing what where are you starting and how does it sort of help in the mapping yeah so first of all I have some couple of those kind of users that use it as a murder board which is kind of
00:12:53
interesting uh and yeah I think I'm really focusing for now uh on spotting relationships faster that was the main focus
00:13:05
um and seeing how things come together by clustering them by topic so right now these are my two main focuses but focuses I think of course in the future the idea is to add uh you know many different filters to help you look at
00:13:18
the knowledge and how the thing comes together from many different angles uh but right now I would say the the primary primary focuses are on connections and and grouping stuff uh
00:13:30
because ultimately you know when it comes to creativity combinatorial creativity you basically make comparisons uh associations um you want to see how the things come together what are the relationships the
00:13:43
similarities what's different and make those sort of comparisons so that's pretty like the basic thing the first thing you you normally do with maps right and then of course on top of it you can build a bunch of other things
00:13:55
and think about more creative ways but what that was the kind of basic that I'm covering right now and Laura like what do you wish you could do with these tools or what are you doing with these sorts of tools now where does this take
00:14:08
you I think something that I've really enjoyed in the past couple of years is how I can I personally use Rome uh which is not as visual but something that I've
00:14:21
really enjoyed which is similar to the way you're using the brain yourself is that I used to have different tools for different pieces of information which I think I don't even think I'm I'm sure was
00:14:33
hindering my creativity because all of these different ideas these different notes were living in silos and in different environments where I could not connect them together and right now I think that my kind of thinking
00:14:46
playground looks very messy from the outside because there's everything my personal stuff is in there my professional stuff is in there I have tasks that I want to work on but I have ideas that I want to think about as well
00:15:00
I have things that are very practical very conceptual I have notes about people that I hang out with I have all of these things together in one place and though this the tool I'm using right
00:15:12
now is not really good when it comes to a visual map there is still all of this underlying Network between those ideas those are still notes that I can connect at a very granular level and it's made
00:15:25
it so much easier for me to you know I write a newsletter I'm doing a PhD I'm doing a thousand different things and it's it's become a lot easier to cross-pollinate ideas in between different projects and
00:15:38
different areas of my life which before we're living in complete isolation so this is something that I'm very excited about the kind of CounterPoint of this and which is why I am excited about the advances that we're going to start
00:15:51
seeing with AI and machine learning is that so much in there that I know they are as Lorenzo very good branding by the way there are some gems in there that are very hard for me to find at the at
00:16:03
the moment and it happens sometimes that I stumble upon a connection or know there's something really interesting maybe a month after I've submitted an article and I'm like oh if only I had
00:16:17
seen that at the time that would have absolutely gone into the manuscript but it didn't because I didn't find it so it's still not perfect but I'm very excited about the potential so I'm going to pick up one word you
00:16:29
said just a moment ago which was manuscript and um I haven't published a book but I've had I've been close and had several books in my in my sort of brain and one
00:16:41
of the things I want to put in the intro to the book is hey thank you for buying this souvenir this this is a slice in time this is a snapshot of a bunch of thinking in time that's been forced into a linear order
00:16:53
because that's what books do the artifact is available openly on the web fully updated and is much more interesting and useful and and by buying the book you can now Mark it up and you can do whatever else but but come join
00:17:06
us because these gems are basically threaded together into an indra's net of context where in fact you can find conversations about the different gems right so so maybe in your in your
00:17:19
dissertation there's a there's a piece about how people process map mapping information in their visual cortex I'm making this up but but you could go talk to people doing the visual cortex research because there's an online
00:17:31
community or a Discord server or whatever else where those people are and if you could get permission to to join the conversation you're suddenly there a long way of explaining that for me like the notion that we still have in our heads that we publish something and then
00:17:44
your book is this dry static thing that you can't go back and update is it makes me sad and I'm really interested in this connected world where connected visual world where we can continue the
00:17:55
conversation and improve you know buff the gems and connect the gems and keep doing this over and over again together as a society and we don't I think we don't have that experience we
00:18:10
don't have that experience for sure or at least most of us don't but we don't even have that dream very often which is interesting to me I definitely have this uh frustration when I I read either
00:18:23
paper books or even digital books or even pages on the internet that I cannot see where they are on the map so I cannot see we don't have I discovered a couple of
00:18:36
years ago which is amazing that Wikipedia actually has backlinks on each page you can see the pages that link back to you to division Wikipedia but it's very interesting that it's not a permanent feature you have to know it's there and you have to dig for it but
00:18:49
that's amazing because it does help you see again it's not a visual map but still kind of see where this page is on the map by seeing all of the other pages that are talking about this page right and I think it would be so interesting
00:19:02
if in books in on web pages everywhere you could see where that page is on the map who's talking about it but conversations are happening and as you said if you could join those conversations learn more maybe contribute yourself
00:19:16
um so yes that's uh that would be the dream that would probably require to change the architecture of the whole way with publishing content at the moment but that would be the dream it's probably if it if you think about it
00:19:28
from the book Publisher's perspective yes it's dramatic changes Etc if you think about it from the perspective of the companies we're busy talking to every day and how this sector of tools for thinking is sort of expanding and growing and what people are doing like
00:19:40
Lorenzo gets a lot closer really quickly um Lorenzo I think you wanted to jump in on the same on the same topic I was just asking I was curious about the dream you mentioned that some people don't have the dream so I was just
00:19:55
curious about if you can elaborate on that uh yeah and and I I realized as I was saying that that I was kind of speaking for everybody and that's a dangerous thing to try to do um but I'm in a lot of conversations
00:20:07
where I'm trying to figure out what does the future of media look like and one of the one of the tropes I I say is in the early days of movie making we put a a new camera which you had to turn my hand
00:20:19
on a tripod in front of a stage where they perform theater because everybody understood theater from way back when and then somebody said Hey what if we put this tripod on a dolly and move it around and suddenly you get the panning shot and then you get jump Cuts because
00:20:32
when they were editing the film they were like we could lose some of this footage and we could move it over here and and then we got you know eisenstein gives us the Montage and a bunch of other and suddenly we have the language of Cinema which with modern tools has
00:20:45
gone kind of you know wacko it's really interesting to see what's possible with Visual Arts awesome the web the our experience of the web and the intertubes and apps and all that seems to be frozen in an early media model where we have radio we're
00:20:58
recording a podcast right now which is both a TV show and a radio show right we're doing both we're sending out the audio in the video uh we have books and magazines and little blurbs which are called blog posts or you know newspapers
00:21:11
or whatever uh and we have a couple other other sort of artifacts we really haven't moved beyond our traditional physically limited ideas of what media is and worse intellectual property over
00:21:24
protection has kind of limited us from the real connectedness of ideas right um and so I think that in in conversations I find that when I start exploring that space of what what what
00:21:36
could be next where could we go there are there aren't enough of us who have said oh yeah here's here's a vision of what that could look like and we're sort of still bouncing off of Ted Nelson's uh vision of you know project
00:21:48
Xanadu um and small funny aside one of the first interviews I ever did as a tech industry Trends analyst was with Roger Gregory the head of the Xanadu company in Palo in Menlo Park I think at the time
00:22:00
um and I wrote let's call it 1991 I wrote an article about Xanadu and I said why sort of listened to them and they were like two years from now we're going to have a platform available imagine my surprise
00:22:13
in 20 oh I'll call it 2010 to read an article about project Xanadu in which one of the lines in there said they're going to have a product in the market in two years so I've been watching this for a long time and even The Visionaries have had a
00:22:27
hard time getting to Market and that's why I'm so excited about what's happening right now so that's that's I don't know if that helps Lorenzo but but it feels it feels like we run into mental barriers somehow um and it also feels like the tech
00:22:39
industry Works in a step function where uh suddenly there's a breakthrough in this oh my God somebody invented HTML and uh the domain name system and a couple other things and then you see the web show up and go crazy and then you
00:22:52
see the web get co-opted and then you see you know layers of things happen and I think we're at this creative cusp right now where new things are happening that that you and others are busy building that's that's really interesting and I think that
00:23:04
um you know I think that all the two most of the two we have in the tools for Thought space really uh have shared this sort of common design principles from a past where really many of the Technologies are available today like AI
00:23:17
was not so available and Powerful right so I really think part of the of the of the thing here is really you know how this interface Are Gonna Change to really enable and unlock the
00:23:31
potential of these Technologies um so it's it's very interesting yeah but how does that Frontier look to you like what are the things you hope and expect to be able to fold in over the next couple years not just sort of you
00:23:43
know the near near term well I think I I see a really spatial canvases in general not because I'm working on it but in general there are some there's a
00:23:54
pretty this is a pretty common um direction that is is explored right now along with probably dialogues with the large language model so these two routes are definitely
00:24:07
um two very interesting path that people are are exploring in terms of how do we actually operate and collaborate more efficiently with the AI um yeah I will say this these two are my
00:24:23
two big bands right now thanks and Laura what um what do you see that you're looking forward to very excited to see more people be creative because I think that
00:24:36
everyone is able to be creative and to produce creative work but that I can understand why for lots of people it's been intimidating to start jumping into this
00:24:49
as an adult if you haven't maybe found that passion when you were a kid if you haven't gotten the training that you wanted to get as a kid if you are not one of those teenagers who downloaded like a cracked
00:25:03
version of Photoshop when they were 13 and started playing with it at that age and and then you're a little bit older and you want to learn but also you have a day job you have kids I don't you know so a lot of people I think have
00:25:15
this feeling that they could do creative work but it's a little bit intimidated because they don't feel like they have the right technical skills to do it and what I'm seeing right now with a lot of
00:25:26
the new tools is that it's removing this artificial barrier between people who have the pure Technical know-how and people who don't so you basically have more space to express pure human
00:25:40
creativity without really having to worry about the fact that oh do I know how to use this tool do I know do I have this technical skill Etc so I'm very curious to see all of the kind of creative work that's going to be
00:25:51
produced by people who currently didn't have access to that space I love that and our call is titled maps and creativity um and and it opens a whole lot of
00:26:03
really lovely questions like how does mapping affect our creativity um I I I've been to and I think you've probably been to and many people listening have probably been to uh events where there's a graphic facilitator or a graphic recorder who's
00:26:16
writing on the wall on large sheets of paper and who's busy channeling what they hear happening in the in the conversation capturing just a few words and phrases to trigger those ideas and then listening for metaphors and oh
00:26:29
there's a bridge here there's a stop sign and a bridge okay good so they'll draw a bridge or there's some biological organism so they'll draw a bacterium or whatever um and that what I love is when on day
00:26:40
two of that event somebody points to the wall to the place where there was a drawing yesterday that was really captivating or that really caught the moment it's like okay so there's space and visual and all of these things are now playing into a glue a group
00:26:52
conversation and part of my my what I'm wondering is how did the these tools that we're busy talking about and building have the same effect on a group conversation in cyberspace or in the metaverse OR in whatever we want to call this thing
00:27:05
I think ultimately like the the real power of these things is the known linear thinking they enabled right which is basically you know
00:27:17
it mirrors perfectly mirrors the actual way your brain works and again one of the problem I think at the design level of the tool we have in the in the text phase is that they come
00:27:29
from a path and where we decided for some reason that the main interface was uh you know a page and so everything happens in there and it's very linear as our lore said before and so I think
00:27:42
that's part of the of the challenge and the problem we have here so maps are really like to me a good way to really unlock the brain powers and and
00:27:55
find you know and mirror the way our neurons and brains work and if you could explain them with AI that would be super great right and uh I would add to this that something that you can do in a
00:28:09
digital space that you cannot do with those kind of physical maps that you have at events which I love also I think they're amazing conversation tools but the conversation still happens between the attendees using the map as kind of
00:28:23
you know like a as a support for the conversation whereas in the digital space you can have a conversation with the map you can change the map as you go you can say you know what actually I don't think this node is in the right
00:28:37
place it should probably be closer to here or what about connecting this and this together what happens if we do that and so in the same way that I've always loved taking notes because it allowed me
00:28:49
to kind of make different authors across different books have a conversation together that they maybe never had because they were living across different centuries I feel like the map is able to capture the conversation at
00:29:02
different points and to morph and adapt as you go as more information connections Etc emerge so it's probably a more Dynamic way to do it and the same way as Lorenzo was saying that we've
00:29:14
been using this page as the metaphor by default so far but by building this in a digital space you can break free from this and you can make the map take any kind of shape and uh and make sure that
00:29:29
it helps you think rather than forcing you to think a certain way it's interesting because as you were talking I was remembering that in graphic facilitation workshops very often people will walk up with Post-its and there'll be some exercise on the
00:29:42
wall and everybody and the way they get to interact visually is by writing something on with a magic marker on a with a Sharpie on a Post-It and putting that someplace and that thing is not an artifact they can then interact with later and one of the things that's made
00:29:54
me sad also is after the workshop you will get snapshots jpegs or PDFs of the images that the graphic recorder created which are you're going to file away someplace because they're not a they're
00:30:07
not a durable artifact in a shared space to prolong or improve the conversation and the tools that I've seen that a lot of people like you know gather dot gather.app or whatever there's a couple different online spaces where you're
00:30:20
actually in a virtual space but not so much in a document and you're not really sharing the closest we've all come to this and this is a pandemic thing as well zoom and Miro or mural right and zoom eventually built a whiteboard
00:30:32
function into themselves during the pen endemic but Miro is the closest that people get to this and it feels like Miro often degenerates into chaos because Miro seems to require really good information Architects like
00:30:44
somebody a few people on deck if it's going to be a multi-user Miro a few people on Deck need to be disciplined about how this what the shape of the space is and a corralling in the Strays
00:30:57
and getting people who are doing dumb stuff to undo the dumb thing and so on and so forth it takes takes a tremendous amount of sort of effort to do that um Lorenzo have you learned a lot for
00:31:08
Jen's notes from interactions in Miro during lockdown and then how is that playing out for you yeah I'm having a I had them I'm still having a lot of interesting interviews about this topic
00:31:20
I will say brainstorming ux research Affinity mapping those kind of activity that normally happens at the Whiteboard and normally you know take time and effort as you were saying right you have to especially the part where you
00:31:33
actually to make sense of this thing because you know brainstorming uh yes we generate idea it's cool it's it's also fun but then you actually have to distill them down and take action on them and sometimes you just leave the
00:31:46
stuff there uh because you know it takes it really takes time and effort um and so one of the really interesting thing that happens once you add Artificial Intelligence on top of these artifacts and and tools is that actually
00:31:59
part of the work can be outsourced to the to this tool and so for example you can say cluster for me the things which doesn't mean that you're going to be replaced that's not the direction I want to go definitely like the thing that I'm trying to do here is really
00:32:12
like a hybrid approach that drives and helps you uh into certain direction or at least to explore some path but definitely uh you know at the end I think we're still the owners we are
00:32:23
still the the drivers because I think that Ai and humans ultimately solve different classes of problems right so AI is much better than us at finding relationship among a lot of
00:32:36
stuff a lot of information again information overload there is a lot of noise how do I find the signal but once you have found this the actual signal then we're much better than Ai and you know find meanings and and get getting
00:32:49
meaningful ideas to move the thing forward so the magic happens when we combine the two together um and definitely brainstorming ux research any type of qualitative
00:33:00
research are definitely great use cases then today happens on mirror but could be improved how how creative do you think AI is Ah that's an interesting question uh it's a fun one
00:33:13
yeah I think we have to think about what do you mean by creativity in the first place good question I mean I'm sort of asking what do you mean by creativity and and when
00:33:25
you're interacting with the AI algorithms that you're busy using do you think of them as creative or when you look at your plans for what to do with gems notes in the future do you think about creativity from the side of the of
00:33:38
the AI That's involved and I feel like as I'm asking the question and refining it I'm making it narrower I don't I don't want to do that I want to keep it as broad as you'd like to take it yeah I don't know like I I hope to not
00:33:51
get into two philosophical stuff here um I don't know I think that creativity in the end is like I wouldn't say that AI is creative is creative itself and that's the reason
00:34:04
why I I'm not sold to the vision of re using AI you know to to actually replace the process that's for the creative for the creative parts right there's there's lots of conversations about either cyborgs or centaurs I don't know which
00:34:17
one's going to win or maybe some other term will win but half human half technology a technology enhanced human and because of my symbiotic relationship with the brain I feel like I'm already well into this Cyborg and so it sounds
00:34:29
like what you're trying to do is be really pragmatic about what AI can actually bring to the party now to free people up to do the Creative work right exactly yeah and I think also one of the
00:34:41
things that I'm observing in with this large language models open Ai and all these things for example generative AI right so is that people are really using it to uh you know creating using it to
00:34:53
generate new ideas uh images the way all these things right text copy copywriting and I think that's great but I think it should be just the first part
00:35:06
of the pipeline like I think the could be like a good trigger for us but we actually have to really take it as an inspiration as a spark and and actually act on it as humans because what I'm
00:35:19
seeing in the market is that sometimes we just take it as the end and thing right the end-to-end process so I have to write this thing and you know I click a button and then create I have it uh and that's let's share it and I don't
00:35:32
really I'm not really encouraged by that kind of vision and future because what's going to happen is that pretty fast we'll have the internet basically made by AI right like the vast majority of
00:35:44
the content out there will actually be made up by AI that to me it doesn't sound like the place where the word very well away right it's not not a world you desire yeah yeah um so I really hope we use this tool use
00:35:58
these tools and models that are actually way more powerful than just creating generating text I think in ways that actually help us to augment our own creative capacities instead of replacing
00:36:10
them um and Laura is AI creative can it be I think AI is creative because I think that creativity is
00:36:23
always combinational so as long as you're combining two things or more and come up with something new then it's creative is it good quality is it interesting I
00:36:36
don't think so at the moment not necessarily but this is a very subjective thing you can work in a museum and see a lot of things that are creative and not everyone is going to agree as to what is worthy of being
00:36:50
called Art and what is not worthy of being whole art what is actually uh what is interesting what is thought provoking what is lazy some people would look at a piece of art and say that's lazy I don't
00:37:01
get it but it's still new and it is still a unique combination that someone has put together to create something so it is creative so I think I think AI is creative but at the moment
00:37:16
and it is the case very often with creativity that by collaborating you do end up having a better output so I think at the moment AI on its own is not the
00:37:28
creative output is is really not at a level where you can deal with that uh a human um kind of like co-pilot to to help it uh so but I do I think in
00:37:42
the future it is possible that we get there and that also worry same as Lorenzo that obviously every time you have a new technology you also have Bad actors that can end up just using it not
00:37:56
to be creative but to spam the web with repetitive lazy Farm content um but I also think that a lot of really good things are going to be created through this process as AI becomes
00:38:09
better and as we become better co-pilots ourselves so there's two examples I point to that tell me that I'm seeing creativity from AI machine learning one of them is Alpha zero which is uh
00:38:23
basically Google's deepmind Project uh where they started with alphago which was a go playing generative system and beat the world's top go player uh then they and and they had trained that
00:38:37
alphago with about bunch of historic games of Go played by experts then they just gave Alpha zero that's why it's called zero no training on human games they just gave it the rules of Go and because it's such a limited domain there
00:38:50
are black and white squares on 19 by 19 grid and here's how you take territory because the rules are so constrained it could play itself and got better than everything else better than it better than alphago better than anything else and there's a really cool chart that
00:39:03
shows you sort of its progression and there's a place at which it sort of Peaks above all the other lines and there's the curve just goes up and that curve to me is creativity and and when because partly because and
00:39:15
maybe this is one of the things I'm realizing from the two examples I'll offer partly because it was doing things that humans thought weren't practical or weren't customary or weren't sort of things you did in that situation it was it was not constrained by by tradition
00:39:27
or custom or habit or whatever and so ghost Scholars are studying those new moves and going oh my gosh this these are ways of winning games and rather than being freaked out and leaving the whole sport similar things happened in
00:39:40
chess where now we have all sorts of freestyle chests that's like combinations of teams of humans and software and the game is Just change in really interesting ways and the second thing is generative algorithms around
00:39:51
texts can create you could say I need to write an article about this theme give me five different ways to talk about it and I know that for me I get into a rut and I tell stories the same way over and
00:40:03
over and over and if something could sit there and sort of boot me out of my rut out of my little track and say Here's three or four other compelling ways to say what you're trying to say I would be thrilled and that would feel combinatorially creative maybe in the
00:40:17
way you were just describing on Laura and those things everything I just described is available today there's no magic needed for that to show up and I'm so that's what's got me really excited about the the space of combining these
00:40:30
different superpowers but you're actually the one who does the synthesis right you get the trigger usually you have to get the triggers I clearly don't want to lose the human in the loop because then we're hosed
00:40:41
like yeah yeah need to keep the need to keep the human in the and there's plenty of dysfunctional you know there's plenty of good dystopian science fiction that pause us that we're no longer necessary once the AIS take over but I also think that that human in the loop and this is
00:40:55
where freestyle chess gets really interesting um and also military things where suddenly there's there's a really quick Sensor Fusion happening with really expanded connectivity with really expanded machine intelligence
00:41:07
and then there's a there's a human trying to to win a battle that that's just that that Frontier is going to feed also what we learn here partly because games you know technology games are sometimes the Leading Edge of
00:41:19
what we what we figure out what would solve a problem that you're facing Lorenzo right now uh if if something were available or somebody were to solve some particular issue yeah yeah great question well again I
00:41:33
I'm really committed to this I think we live in a war in a very noisy world right there is a lot of stuff going on all the time and I think it's getting increasingly hard
00:41:45
to find the signal as I said in the beginning so I'm really committed to do something about that about helping people really get to the meaningful piece of knowledge they need to move their project or
00:41:57
personal growth or whatever forward and one of the possible obstacles as I was saying before is probably generative AI right because the way we're using it right now is actually to create more noise instead of decreasing it right
00:42:11
yeah because you're basically like boosting the process of of putting stuff out there um one of the interesting thing though I think could be explored with the large
00:42:24
language models is actually doing the opposite like using those models to actually try to narrow down and reduce and synthesize knowledge so instead of asking them hey generate for me this piece of content say hey
00:42:37
generate for me this summary or what this thing is talking about or you know work in those kind of Direction with the problems which is totally possible uh it's just you know what I see on the market today
00:42:49
is definitely more towards the creative generative stuff probably because of the name I don't know um thank you and and partly what you're saying makes me think that um the way you'd like to implement AI
00:43:02
would help us overcome enumeracy uh availability bias meaning when we're sitting in front of a large data set or many streams of information we just pay attention to what we saw most recently at sticks in our heads we sort of see
00:43:15
our own neighborhood vividly but we don't see the larger context and and systems like you're describing could in fact say hey you're really excited about this thing over here but in fact in the bigger picture here's kind of what's
00:43:27
going on so I love that and Laura same sort of question for you for the field like what what do you wish would emerge I'm a I'm a teacher and a researcher so selfishly
00:43:41
I would love to see more applications to help us learn better and in particular I'm interested in seeing how we can use even generative AI to create more
00:43:52
adaptive experiences it is kind of connected to what you just said Lorenzo but you know there's this tool for example illicit.org where you can look at research papers and gives you a summary I think taking things a little
00:44:06
bit further it would be amazing if you could take any kind of piece of content of knowledge and have ai adapted presented to you based on your current
00:44:17
level of Knowledge and Skills based on what you know so you can understand it as quickly as possible so for a research paper if you have no scientific background just giving you the very simple explain me like I'm five type of
00:44:31
summary but if you're a scientist in the field just getting into the nitty-gritty Etc and I think this is going to be so helpful in general in the future both for people who just want to teach themselves different topics but also to
00:44:45
better communicate information in a way that's more transparent that people understand better on the internet so I'm very excited about that would both of you like to have a smart assistant that was watching everything you did and making suggestions and going
00:44:58
and running and bringing things or would that creep you out I mean are you is is that a thing you'd like to have or use I mean I already have Google everywhere so technically it's happening it's just
00:45:11
not that useful at the moment yeah but but I mean when I'm busy saving things and putting them in the brain or whatever I would love to have a an AI on the side that says hey Jerry that looks like an author and her book would you
00:45:24
like me to put her other five books in in your mind map and I'm like yeah and I don't use metadata or tags or anything like that in the brain only because it takes time and effort and I know that every every additional step that it's
00:45:36
going to take me to to do the work it decreases the likelihood I'm going to try to do the work so the the I could all I could also say hey here are the tags that I think are are relevant to what you're doing right now can I just
00:45:49
drop this metadata in with uh with the uh the basic data and it'd be like thumbs up let's go so there's a bunch of ways in which work can be completed and made better and then when you have a lot more metadata the possibilities are
00:46:01
endless at the other end as well I think that right now it's the kind of thing that feels a little bit strange to us because we're not really used to it but that maybe in 10 years we'll look
00:46:13
back and think whoa I used to do all of these things manually how weird how much time like I was wasting basically doing all of that logistical stuff when now I
00:46:25
have an assistant doing it so I do feel like I have kind of a natural resistance to it but I don't think it's necessarily rational and I don't think that uh is going to be you know it's it's
00:46:38
probably it's probably going to go the other ways from we're probably all going to have such assistance in the future doing all of these things that are going to show like very high friction if we have to do themselves again uh if at some point our
00:46:51
system doesn't work or something like that we'll be like oh wow I have to create that calendar entry and I have to import that method myself now I remember reading somewhere long ago that that leibniz the co-inventor of
00:47:03
calculus and polymath and other probably corresponded with over 600 people during his sort of working life in that era and I'm like wow I probably contact 600 people through all the different socials
00:47:17
that I'm on you know in in a week actually um no problem and we have instantaneous communication with you know you shrug and there's a message sent so sort of instantaneous sometimes thoughtless
00:47:28
communication and and you know leibniz when he was sitting down was probably thinking for quite a while before sending a message off uh to one of his collaborators now we have machine learning that can be part of that conversation so the the the contrast is
00:47:42
very sharp for me uh in in terms of where we are and what the possibilities are and yet the there's a plenty of Bad actors as you mentioned earlier on Laura in the field and these power tools are
00:47:54
available just as easily to them um so there's a whole other set of questions and probably podcast episodes around hey how do we do good friendly AI good AI whatever else you want to call it there's there's a bunch of efforts
00:48:08
already on the ground to do this just uh we're nearing the end of our time together and I just want to know what what other things matter for you in this General realm of maps and creativity
00:48:18
what uh what Avenues or possibilities um are exciting to you in this space I think at the moment there are lots of amazing tools that are doing a great job at
00:48:31
uh creating a good solo player experience when it comes to thinking in maps you've mentioned how Maro can become very chaotic very quickly so I think when it
00:48:45
comes to maps and creativity this is really the next Frontier how do we make that experience as smooth and enjoyable and productive as possible so you can
00:48:56
actually Unleash Your Collective creativity you did mention actually I just thought about that how you needed some kind of like map moderators when you were using such shared whiteboards
00:49:08
but maybe an AI can do that in the future actually and moderate it for us so we'll see what that looks like but I think once we start having really really good multiplayer map making experiences
00:49:21
then we'll really start seeing the impact that it can have on creativity love that thank you and totally I want that big time Lorenzo I totally agree and I would say
00:49:35
um you know what I'm observing in in the interviews and interacting with the users and any person involved in some way with this I'll totally say that you know that's that's exactly the thing
00:49:47
that has to happen I think that really sounds like the the next logical and useful step for these things because I think part of the reason then you you when you you know when you are by yourself um
00:50:00
Sometimes some stuff just makes sense because you know it's in your brain right um but when it comes to other people and collaborators then that's where the thing starts breaking
00:50:12
um and so you know having an AI or some sort of help that helps you actually Reco like and pre-compose the whole thing and see the big picture that would be very helpful as well exactly and then
00:50:25
I'm not hearing of AI research and maybe I haven't been looking for it that's doing this particular thing we just talked about basically the the stewarding or shepherding or coaching of humans trying to make sense together uh
00:50:38
and figuring out you know how to maintain consistency of the space that they're working in in some way things like that I'm not hearing this uh yet as a as a research topic so I'm I'm now I'm
00:50:50
sort of going to look for it and see what happens any any other concluding thoughts anything else you'd like to add to the conversation yeah just a quick follow-up on this I think we still are at a stage where we have to actually figure out first the design that
00:51:03
actually unlocks our interaction with the AI as I was saying before before actually at the personal level before actually moving to the collected one uh because I think that the really from a design perspective the thing we have the
00:51:16
the metaphors we have today with the interfaces and stuff like that are kind of broken to actually take advantage of these new powerful tools we have so we'll get there eventually of course I
00:51:30
think so too I think so too uh and Laura anything you'd like to add I agree with Lorenzo it's very early but it is exciting it's also really great that a lot of different actors from all sizes are working on this we have some
00:51:43
of the biggest tech companies working on this but also I've seen really good tools built by solo Founders so I think this diversity of thought actually this is a bit meta is really good for the
00:51:55
field and um yeah I'm very excited to see what's coming next uh just the availability of power tools for small groups and individuals is astonishing the the code bases the
00:52:09
availability of large data sets there's a whole bunch of Technology available now that was hard to get to and or or not invented yet just a decade ago and it's on the table right now and it's really exciting with where we're going
00:52:21
thank you for listening to tools for thinking a new podcast that might just help you with your thinking if you're part of a startup in the sector please knock on our door at betaworks.com foreign [Music]
End of transcript