Lukas Salecker about deliberAIde and why generative AI makes a difference in deliberative democracy
Ep. 05

Lukas Salecker about deliberAIde and why generative AI makes a difference in deliberative democracy

Episode description

In this episode of Democracy Innovators Podcast, Alessandro Oppo interviews Lukas Salecker about deliberAIde, an innovative platform that uses AI to enhance deliberative democratic processes like citizens’ assemblies. Lukas shares how deliberAIde helps make democratic participation more accessible, efficient, and scalable through AI-powered transcription, analysis, and reporting tools.

Download transcript (.srt)
0:00

why come on another episode of democracy

0:02

innovators podcast and our guest of today

0:05

is lucas celica co-founder and ceo of

0:09

deliberate

0:10

and thank you for your time for

0:12

being here hey alex glad to hear

0:16

i am

0:19

i see that you have cofunded deliberate

0:21

and them how would you describe is

0:24

a platform software or idea i dunno

0:29

the some before i get into the

0:31

nitty gritty of what we're building which

0:33

is our platform and the media get

0:36

give a little bit of background of

0:37

where where does idea comes from said

0:41

wasn't around two years by now and

0:45

when i woke up in the middle

0:47

of the nights and was sunday

0:49

date monday morning at four am and

0:52

i i had this idea to use

0:56

the power of these recently and publicise

1:00

new to technologies of generate of hey

1:02

i am for a lot of people

1:04

known through tattoo between and connect that

1:08

to

1:09

and democratic innovations like citizens assemblies to

1:12

help make them better make them more

1:13

accessible make them are scalable

1:17

and and

1:20

so that's that's exactly what we're doing

1:22

at deliberate to

1:24

i used these generate i had technology

1:27

think about how we can build them

1:29

into applications that are made for processes

1:33

like citizens seventies or other deliberative process

1:37

for different purposes and and the main

1:41

purpose then restarting with 'em with our

1:43

platform for the first

1:45

version of our platform that for launching

1:47

and and mart and and may is

1:51

sensemaking how can we help and organizers

1:56

or commissioners of these kind of participation

1:58

processes to gets to more easily get

2:02

insights and summaries from from the from

2:05

the discussions and and ought to get

2:09

better insights from them

2:11

and so that's exactly what we're doing

2:13

in his first version which and on

2:16

a more technical level starts with recording

2:20

and transcribing discussions and with built in

2:24

anonymization to puppet protect people's privacy then

2:29

we take that transcription data and

2:31

and in the future ultem other kinds

2:33

of data runs around the discussions like

2:36

manual notes or and pictures of of

2:40

whiteboards like whiteboard sticky notes for example

2:44

and we use that data to extracts

2:48

and contributions that participants made

2:51

and to and then make sense of

2:54

these contributions like and what what what

2:57

things that they propose what opinions that

2:59

they that they defend what what's what's

3:03

like wetware similarities across different contributions and

3:07

in the future also where our agreements

3:10

are disagreements

3:11

and so on

3:13

and once the organizers have done the

3:17

sensemaking they can then take their analyses

3:20

and summaries and also use ai assistance

3:23

to turn turn the insights into meaningful

3:26

reports that condemn be passed on to

3:29

the decision makers or the public and

3:32

yep that that's the

3:33

set of functions that we focus on

3:35

in the first platform version

3:40

the in the pasta was thinking about

3:43

the softer for civic participation and the

3:47

i taught a lot about that these

3:49

the automation of about summaries and the

3:53

ai integration and

3:56

and it's very cool that the some

3:59

tools like this one existed and i

4:01

wanted to ask you if you had

4:03

the done the some experiment with some

4:06

group of people and

4:09

he fit is there anything relevant to

4:12

sure

4:14

indeed so and since we're about to

4:16

launch our platform with been going through

4:19

quite a bit of testing in and

4:21

real-life processes over the past weeks and

4:23

months and so we have worked with

4:26

cities and municipalities in germany and or

4:30

also next week we're going to accede

4:32

to a test at a car

4:34

conference buddy european committee of regions and

4:38

and what we can say is that

4:42

that's really a lot of excitement about

4:45

but these tools can offer to to

4:47

people who organize these kind of parts

4:49

patient process to give you an example

4:52

will working with what

4:54

is called social planners and in in

4:57

the municipality of zika wittgenstein atlas a

5:00

relatively small town or municipality in and

5:03

the state of norfolk westphalia in germany

5:07

and day what the source of planners

5:09

do as they organize the series of

5:12

conferences where the involve citizens

5:14

it's and to to decide on future

5:17

social social planning social policy making and

5:22

and as the day they have tested

5:25

our software various occasions now and they

5:28

were already excited simply about the the

5:31

possibility to record and transcribe and the

5:34

discussions and but when they saw for

5:37

example and how they can even going

5:41

beyond transcriptions and can can can use

5:44

a i to extracts and the the

5:47

ideas that participants shared to to cluster

5:50

different ideas together and said to use

5:53

our tagging functionality

5:54

the to extract and even more specific

5:56

insights and all of that and on

5:59

a on a digital whiteboard so we're

6:01

actually we be developed aided to whiteboard

6:03

like mirror or mural and where they

6:06

can work with the insides and some

6:09

reason visually and interactively

6:12

they they did they had a the

6:15

feedback we got from them was that

6:16

does this exactly like what what what

6:19

what they need or it didn't know

6:20

they need it and and they they

6:23

really believed that this could help them

6:25

tremendously in the in the future and

6:27

especially speed up their documentation and reporting

6:30

processes

6:36

i'm thinking about the

6:39

the last thing you he just said

6:41

that the didn't know that to that

6:43

that they didn't know that they need

6:45

the the platform because maybe they were

6:47

not aware of that some platform like

6:51

this or that the technology cool helps

6:53

so much in the seen political processes

6:57

and the

7:01

i was reading the description of the

7:03

libor aid and resulting very interesting our

7:06

mission is to ensure that every voice

7:09

is heard

7:11

and i think this is what is

7:13

very cool about the civic tech that

7:16

could the law literally every voice to

7:18

be heard and the do you think

7:22

that the this platform or like a

7:24

similar platforms

7:26

in the future cool the

7:29

i could be not just related to

7:33

do to some experiment in a town

7:36

but really be something that people use

7:39

the every day and saw all the

7:41

political system eventually with the will change

7:46

i certainly believe so that's that's the

7:48

vision that that drives us and really

7:52

believing that and your leverage this new

7:56

information technology that generate a fe eyes

7:59

that we can create spaces and that

8:02

that extends way beyond the current political

8:05

arenas

8:06

so right now political decision-making political discussions

8:09

are very confined to you know people

8:12

who have been elected or and add

8:15

people who are members of political parties

8:17

or and the media and or social

8:21

media

8:23

but all of these current political arenas

8:25

and spaces have a lot of shortcomings

8:27

am especially when it comes to who

8:29

do they include and who do they

8:31

exclude and for example to be elected

8:34

and it's and it takes quite a

8:38

bit of time first of all but

8:40

then also you need to meet other

8:41

conditions texts to be a

8:42

acted as a representative and thus partake

8:45

in the political conversation and we believe

8:48

that and democratic innovations like citizens assemblies

8:50

and oftentimes of the the selection of

8:52

participants in this and is many public's

8:56

through random selection saw lottery and can

8:59

help overcome some of these shorts or

9:01

cummings of of excuse

9:02

asian and and but they're they're still

9:07

there still even within these new democratic

9:09

innovation processes like citizens' assembly they're still

9:11

challenges to overcome and that's it exactly

9:14

where we believe and technology this new

9:17

information technology of generate for i especially

9:19

can come in add to tackle issues

9:22

ranging from and

9:23

inclusivity to and the depth of conversation

9:26

and quality of dialogue at to scalability

9:29

and costs of these of these processes

9:31

just to give some examples in terms

9:33

of inclusivity oftentimes language is still a

9:36

barrier to participate in and citizens assemblies

9:40

and we already know that generate a

9:42

eyes real

9:43

the good at translating and and they're

9:45

actually start ups popping up that help

9:47

and with real-time translation for example so

9:50

that can help bring down the barriers

9:53

to inclusion in conversations and citizens assemblies

9:58

then depth of dialogue and currently and

10:03

and citizens' assembly set out well facilitated

10:07

and you oftentimes only scratched the surface

10:10

of underlying conflicts and when people have

10:13

disagreements our conflicting opinions you don't really

10:17

to have the time and of the

10:18

air the the insides to actually dig

10:21

deeper and and truly understand

10:23

and why certain conflicts existence society and

10:27

we believe the generate if i and

10:28

and and the and analyses that generate

10:31

a i can enable can help us

10:32

understand conflicts better and and thus mediate

10:36

better between conflicting views and the third

10:38

example related to scale would be if

10:42

if you can use like our girl

10:44

from one platform for example for summarization

10:47

sense making reporting in like very quickly

10:50

and minutes rather than days you save

10:53

a lot of costs for for and

10:56

organize us for an analyst for no

10:59

takers

11:01

and that that allows you to organize

11:03

a lot more of his assemblies and

11:05

or organize such assemblies seventies in the

11:07

first place and if you have a

11:09

very small budget and a and a

11:11

second plane relate to the scared would

11:12

also be and right now

11:16

do you to that texted organize something

11:18

like a citizen assembly like a mean

11:20

of meaningful deliberative dialogue you you need

11:23

most of the time to do it

11:24

well you need professional moderators but there's

11:28

a there's a scarcity of professional the

11:30

moderators so we believe in the future

11:33

and with a i we could empower

11:35

party

11:36

recipients to self moderate for example there

11:39

are already some experiments in that direction

11:40

underway stanford and orlando liberation platform and

11:44

person foremost and but we believe we

11:48

can do better we can help of

11:49

a can use a i to to

11:51

help and creates a i am moderation

11:54

assistance that can truly understand the convoy

11:56

recession and nudge in the right direction

11:58

to make a constructive and that will

12:01

allow for a scaling or multiplying these

12:03

kind of processes by the thousands and

12:06

thus help to include many many more

12:08

people than than current the is the

12:10

case in an average citizen assembly which

12:12

which nominees only around one hundred to

12:15

one hundred one hundred and fifty

12:16

people

12:21

i share we few the excite excitement

12:24

about the the ai agents that moderate

12:28

because i i will agree with feud

12:31

that the good the moderators are not

12:33

the there are not a lot of

12:36

good the moderators and the

12:39

and yeah agents could help a lot

12:41

to and they're thinking about the scalability

12:45

i mean about participation like is there

12:48

anything

12:50

that you were thinking cooler stimulates user

12:53

to participate like i'm thinking about some

12:56

gamification system or like people just participated

13:00

when a you did the experiments in

13:04

the

13:05

i mean with the experiment that you

13:07

have done now like in some towels

13:09

with some institutions people just participated all

13:12

day received some kind of

13:15

the

13:17

now

13:19

southern i'm actually a bit cautious with

13:22

an attempt to incentivise participation that are

13:27

more in the technical sites like gamification

13:29

for example and some not sure if

13:32

they're really dissolution i think that then

13:35

it then need to be and

13:38

other incentives that i'm more in the

13:39

process site for people to want to

13:41

participate and i think a lot of

13:44

times what actually prevents people from participating

13:46

is not that participation is boring and

13:50

therefore you need to gamify it making

13:51

more exciting but there are more structural

13:55

barriers in place like

13:58

do people just next door to job

14:00

or sometimes more than one job just

14:02

not having the time to to to

14:04

take to to go to an event

14:06

like that and in in the evening

14:08

on a work day or there too

14:09

exhausted to to go on participate in

14:12

a and then process and like the

14:13

social the planning process i just talked

14:15

about on the weekend

14:17

and they may also have other responsibilities

14:19

like caretaking that simply cans at they

14:23

simply can't afford to participate and saw

14:25

dead and and in order to tackle

14:27

dose we also need to put structural

14:31

solution that solutions in place for example

14:34

and and big citizens the seventies like

14:36

those that isn't this

14:37

family on a nutrition by the german

14:39

parliament and last year and and twenty

14:43

twenty three twenty twenty four and and

14:46

it's also the case and other assemblies

14:48

people will get paid at a small

14:51

amount but at least they get some

14:52

form of compensation so that they can

14:54

participate and and if

14:57

future of we need to go beyond

14:58

that so that they don't just need

15:00

to get paid like one hundred euros

15:02

per day or something but the actual

15:04

need to need to be able to

15:05

and gagged like at get free from

15:08

their jobs for for the time they

15:10

participate in such processes and or and

15:14

yup it if you if you think

15:16

it even if more law

15:17

long-term white why students someone who participates

15:21

in the in a citizens' assembly be

15:25

almost equally compensated to an elected representative

15:29

so

15:31

they are representatives after all that take

15:34

time to deliberate on the nation to

15:35

inform themselves it's hot work so why

15:38

should they only get like one hundred

15:39

years per day and if they do

15:42

almost the same work as elected representatives

15:45

and so i believe these are structural

15:48

solutions to structural problems that proven participate

15:51

nowadays that are more and the process

15:54

side rather than than than and then

15:56

the technical side

16:01

i agree with from your analysis and

16:04

the

16:06

i would like to ask you

16:08

and if you want to share some

16:10

of your personal but ground eventful also

16:13

starting from when you were a i

16:15

tried

16:18

so starting all the way back from

16:20

from when i was quite i think

16:21

might be a bit boring along but

16:23

i i can start at the like

16:25

quite pivotal moment or or time in

16:28

my life which was almost like half

16:31

my life a gulp around fifteen years

16:34

which is why when i was sixteen

16:37

and before i was sixteen i you

16:40

know i grew up in a small

16:41

town in germany at i've actually never

16:43

left germany until then and and so

16:47

my at you could say i grew

16:50

up in and quite the bubble and

16:53

and but when i was sixteen i

16:54

had the privilege to actually gone and

16:56

exchange here

16:57

abroad

16:58

in mexico for one year and to

17:02

go to high school there for for

17:04

your learned learn language get to know

17:07

the the culture meet meet new people

17:09

extend my heroic horizons and extending my

17:12

horizons it it truly was because it's

17:15

a literally bursts this bottle of of

17:18

privilege

17:19

and that that i grew up in

17:21

because there was the first time when

17:23

i got this outside view of of

17:25

where i grew up in germany and

17:27

europe and the global north and and

17:29

made me realize while how how unequal

17:33

the world actually is and and also

17:35

how unfair the world this and which

17:38

makes

17:38

the world so unequal and so since

17:41

them and you can say in while

17:45

retrospectively i said i call it that

17:48

the moment when i found my purpose

17:49

or the time when i found my

17:50

purpose in life because it really motivated

17:53

me since then to to keep her

17:55

to keep trying to it to to

17:58

learn and stuff

17:59

audi and educate myself and finally find

18:01

a work so that i could do

18:03

something against these inequalities and against is

18:06

ingested injustices that it had first scene

18:09

when i was in mexico for the

18:11

for that year and and this kind

18:15

of purpose only got reinforced over the

18:17

past fifteen years whilst i was studying

18:19

in white is inequalities exists am i

18:22

had other chance to live abroad and

18:24

and other countries of the so called

18:26

godless of and where we're and and

18:30

i saw a more stocker inequalities and

18:32

injustices at work and and for up

18:37

so far long time i thought

18:40

i i might want to help tackle

18:42

these problems as as a development worker

18:45

saw kind of working for the united

18:47

nations working for the world bank and

18:49

your name those institutions german development agency

18:53

and

18:55

but eventually i realized okay i don't

18:58

think while that hasn't been working for

19:01

seventy eighty years now since development corporation

19:04

started and it's probably not gonna work

19:06

because there are other bigger structural issues

19:08

and place power inequality especially

19:12

so actually we need to look at

19:13

ourselves in the global north in europe

19:16

in the west and think about okay

19:18

how are we actually preventing other countries

19:22

around the world especially in his so

19:24

called global south from pursuing their their

19:26

own visions and ideas of of a

19:28

better future

19:30

we we are we have been kind

19:32

of imposing our idea of how how

19:35

societies ought to look like which we

19:38

call it development and we've been trying

19:41

to promote that through development corporation but

19:44

i think that's that's the wrong approach

19:47

and and so let's be what i

19:49

realized eventually

19:50

around the time i was studying their

19:51

masses we need to look at our

19:53

own system that our own economic systems

19:55

and and are on political systems in

19:58

europe in the west in the global

19:59

north and think about how the day

20:01

suppress systematically and how today and takeaway

20:06

space for others to truly flourish and

20:09

pursue their own visions of the

20:11

and so i started to study and

20:14

things like like d grove ideas like

20:17

d grow of posts growth and how

20:19

can be transformed our economic systems to

20:21

make them more a lot of technological

20:23

and social and eventually then i realized

20:26

okay while we probably also need to

20:29

change our political says systems to

20:30

achieved this kind of deep economic transformation

20:34

and so then i became interested in

20:36

deliberative democracy because essentially a deliberative democracy

20:40

is about redistributing political power to ordinary

20:44

people and i i truly believe that

20:47

this is what's needed to achieve the

20:49

that profound transfer

20:51

formation said we need to make our

20:53

our economic political systems our societies more

20:56

ecological more just more more social and

21:00

yeah that's what led me to my

21:03

passion for deliberative democracy and then eventually

21:05

two years ago kind of connecting that

21:07

passion and to

21:11

how how can we use technology to

21:13

enhance these this new forms of democracy

21:21

the will to like to share some

21:23

of your a academic or professional the

21:26

ground

21:28

yep so i studied and

21:33

broadly speaking political science or political economy

21:37

and sided and a bachelor's in liberal

21:40

arts and sciences global challenges that's how

21:43

it was called and in the netherlands

21:46

at leiden university college and that it

21:49

was a very open interdisciplinary him bachelors

21:52

earth where i could really choose my

21:56

my own focus from from a lot

21:59

of options that i was given and

22:01

i decided to focus broadly on an

22:03

economics on an international development and on

22:07

governance and

22:10

and and my masters i then decided

22:13

to do development studies and cause partly

22:17

because i was still thinking about becoming

22:20

a development worker but also because i

22:22

really want to to understand this idea

22:25

of development and now and i wanted

22:27

to critically engage with it and and

22:30

added

22:30

that in in in the uk at

22:32

the university of oxford and what what

22:35

was really special about that masters was

22:37

that it takes a step back and

22:40

and critically questions this whole concept of

22:43

development and i was very glad that

22:47

that ended this mass masters because as

22:50

i just explain

22:50

and it's a really let me to

22:53

question the whole idea of development and

22:55

then vote start to focus more on

22:58

what can be actually do and europe

23:00

in the west and the global north

23:05

yeah i i also had the impression

23:07

that the

23:09

i mean the sometimes corporation

23:13

do it can be no way i

23:16

i mean corporation not as a word

23:17

not the corporation that can be done

23:19

by email you when someone else for

23:21

like a relief to ngos and i

23:25

mean sometimes it cool seems like a

23:27

sort of colonial nation in some way

23:30

saw you sure some of them

23:34

the dubbed about to a if that

23:38

is the solution and the also be

23:41

about older

23:43

you use that structure many times like

23:46

also before and i totally agree like

23:51

there are some structural problems there is

23:53

a some violence that instructor in our

23:57

society and the yeah we can try

24:01

to change it in some way with

24:03

these new tools

24:06

and the

24:09

i see also cooking with the other

24:11

guests do that there are basically two

24:16

ways one is to take a

24:19

i i mean the democracy that we

24:21

have a right now and we just

24:23

put some way i e some automation

24:25

some transcriptions to and we try to

24:30

make them more efficient or the other

24:34

ways to find new solutions completely different

24:38

from the bustle

24:39

and i think that the deliberate the

24:42

is a on this second

24:45

oh wait

24:50

so you're trying to find a new

24:52

solutions for people to be able to

24:56

the side and to talk

25:01

and the i was reading about peer

25:04

parliament's is it the

25:08

it seems connected the no way or

25:11

not

25:12

we're very good yes actually this and

25:15

i just realized that forgot to to

25:17

answer the second part of the previous

25:18

question on my professional background and thought

25:23

i mean it's it's not not a

25:24

very long professional experienced that i have

25:26

and only round and four years now

25:29

and to have that has been deliberate

25:31

and

25:32

but right after my master actually started

25:35

working in an organizing citizen participation at

25:39

a company called he fuck which has

25:41

a german company that's that's 'em now

25:44

a one and among the european market

25:46

leaders who could say and organizing citizen

25:48

seventies and other participation process and i

25:51

worked at for half a year

25:53

and during that it during those six

25:54

months i actually helped organize what's called

25:57

at pier parliament's that he just mentioned

26:00

and i was a process by her

26:03

are organized on behalf of the european

26:05

commission and to involve citizens european citizens

26:09

in you climate policy making

26:13

and what i think was was special

26:16

about this is that this was one

26:18

of the one of the first times

26:20

and not the first time that and

26:23

a high level political institution organized and

26:26

distributors and deliberative participation process and and

26:31

what does that mean

26:33

so basically european citizens were asked to

26:37

organize small scale and discussion groups or

26:43

the article and house parliament's where they

26:46

invite their friends or neighbors or colleagues

26:49

and gathered at home or at a

26:52

bar or

26:53

community center to add as discuss different

26:58

topics related to your climate politics and

27:02

at the end of their discussions they

27:04

were asked to take a decision on

27:06

what what recommendations they they want to

27:10

to make to the european commission what

27:12

the change about

27:13

odds and certain aspects of of a

27:15

you climate politics and they submitted those

27:19

were can commendations back to us and

27:21

more organized this process season total they

27:23

were i think around five hundred or

27:24

six hundred parliament's organized through the whole

27:27

of europe song till around five hundred

27:29

participants and

27:33

i thought that was that was really

27:35

knew because instead of saying like what

27:40

what has happened a lot of times

27:41

over the past ten years all years

27:43

on that participate in platform disco and

27:46

and have a dislike comment on others'

27:48

ideas on and on nine participation platform

27:51

they really locally

27:52

list is processes by us and and

27:54

decentralized the process by art asking people

27:57

to organize discussions at home like real

27:59

time face-to-face discussions and i think that's

28:03

really crucial because what's missing on and

28:06

mass online participation is this very essential

28:10

human element and and a human canal

28:13

action and and ability to look each

28:14

other in and into it and in

28:16

the face see each other's emotions and

28:19

then makes and gestures and and very

28:22

important to hear hear each other's stories

28:24

and all of that is missing an

28:27

online participation platform and it's also missing

28:30

on on social media and i i

28:32

believe lots of

28:33

part of the reason why a social

28:34

media so polarizing because people are not

28:36

looking into each other's eyes monday when

28:38

they discuss but in these peer parliaments

28:42

and they preserved as is this this

28:44

essential element by by focusing on the

28:47

small scale and interpersonal and discussions but

28:51

then they they

28:53

that they scaled stabbed by multiplying these

28:56

small scale deliberate of discussions which eventually

29:01

allowed several thousands of people to participate

29:04

and so where's the connection to what

29:06

i'm doing right now in a way

29:08

that's a part of the vision that

29:10

striving then drove me to start to

29:13

bread and that dry

29:13

having us at at the la bread

29:15

because back in the days when when

29:18

i organized the pier parliaments and i

29:20

had to manually sought through and summarize

29:24

and analyze all the submissions that'd we

29:25

got back from participants and i can

29:28

tell you it was a mess and

29:29

it was not good and was hard

29:31

work a real pain and it

29:33

not do justice to

29:36

at the new ones and motor dimensionality

29:38

that that that their discussions probably contains

29:42

and but i think with generated ai

29:46

we can really use that information technology

29:48

to the first of all and

29:52

do more justice to the complexity of

29:54

discussions that people are having and in

29:56

such distributed processes and second of all

29:59

i think there's also a possibility to

30:01

interact interconnect the different and discussions on

30:05

a pier parliament's the house parliament's they

30:08

were isolated that place isolated in isolation

30:11

from each other there was no

30:12

exchange of information between them but i

30:14

think we can use generate of the

30:15

i to cross-pollinate between them

30:18

and and that's actually what a lindlaw

30:21

more and and one of her articles

30:22

cause and multiple rotating many public's and

30:26

that's really and and and an idea

30:28

that's that's been quite them inspiring for

30:31

me and and kept kept me going

30:34

because i believe like that we can

30:36

really is scaled small scale com

30:39

conversations two thousands or even hundreds of

30:42

thousands of people and thus get back

30:46

the

30:48

the essence of what makes her good

30:50

political conversations by it but at the

30:53

same time allowing masses to participate in

30:56

the conversation

31:01

this will be ozone

31:04

and the i think about the polarization

31:07

the you know we i mean talking

31:11

with someone to someone that the that

31:14

tesla different ideas from as of it's

31:17

what is the needed to not be

31:21

polarized but some time we are used

31:24

to think like in a sort of

31:25

binary way so if if if you

31:27

think in a different way from me

31:29

at the do not talk to you

31:32

and so i see this candle software

31:34

very useful

31:36

and the

31:39

i was wondering because the about the

31:43

platform

31:46

so it can be used by different

31:48

users

31:50

saw institutions political parties i think angles

31:55

i dunno and the is there any

31:58

difference the related to the use the

32:02

i mean how they will they use

32:03

the software or like of would it

32:07

work in the same way for every

32:08

kind of

32:11

entity

32:13

so so right now and as first

32:15

burden it's really mainly targeting and we

32:19

call him organizers and as opposed to

32:23

like every every participant or like an

32:26

ordinary participant in the process and

32:30

and but for organizers like no matter

32:33

if they're part of the political institution

32:35

political party a municipality and or even

32:39

within a company and it's it's all

32:42

the same

32:44

i saw daredevil does one interface like

32:47

sure they are like different roads excess

32:50

different axes like the navy opponents the

32:52

navy facilitators and maybe product managers but

32:55

essentially it's the same and everybody has

32:57

access to and the transcription to and

33:00

the sensemaking to the reporting and

33:04

but yet participants at this point unless

33:07

they are self organizers will not interact

33:10

much with the platform itself that will

33:12

most d c

33:14

interim results like clustered ideas and or

33:18

water final reports that have been created

33:19

with the platform

33:24

i ever thought about the collectivization of

33:27

problems because they something that the

33:31

i mean there was not to think

33:33

about this question when it was preparing

33:34

the interview but the something that i

33:37

thought the about also now

33:41

because sometimes i dunno person as a

33:43

problem do i make these example i

33:46

dunno i fight though i always fight

33:49

with my brother and the i can

33:52

think that that the is an individual

33:54

problem i mean it is my fault

33:57

or my brother fault but then i

33:59

dunno knife in the town

34:01

everyone fight with his or her own

34:04

brother or sister in owais are sort

34:07

of such a problem and the i

34:10

think that nowadays is very hard the

34:12

to be aware of this kind of

34:14

problems

34:17

but using this kind of tools as

34:19

you said the week could the have

34:20

a small discussion groups and then who

34:24

have all the the problems that goes

34:27

in the summer he made by i

34:30

and so then people can be aware

34:31

of certain problem that the the were

34:34

not

34:37

so indeed them i think that has

34:40

real potential to him

34:43

and

34:45

like move move the conversations back from

34:49

kind of private spaces like i fight

34:52

with my my family at home and

34:54

everybody fights with their family at home

34:56

or and i dunno on net neighbors

35:01

fattened fighting with each other and lebanon

35:04

a more private space

35:05

and also from what has become the

35:08

the main public square these days social

35:10

media and and also more traditional media

35:14

and take take those conversations back from

35:16

these two spaces because they have to

35:19

have their issues and really try to

35:21

create a new public's public square and

35:25

are actually not not a new one

35:27

but get back the old once in

35:28

a way to called back to the

35:30

future and cause we used to be

35:32

able to have these conversations and before

35:36

social media the foreign mass media and

35:38

i think we just need to get

35:40

back to depth and and will be

35:42

really believe that and generate of i

35:45

ai and and other emerging technologies can

35:47

help with them

35:51

and the i want to ask you

35:53

what is the actual state of the

35:55

platform like

35:59

i know that these are working in

36:01

some way because you have done a

36:03

tester you know with when his policies

36:05

and other institutions but the i wondering

36:09

do right now i mean what are

36:13

you working on is there are also

36:15

or so if is that is there

36:17

any problem

36:18

mm that your stock on any skill

36:21

that you're searching for maybe someone that

36:23

listen to the podcast can say okay

36:26

it maybe i can help on that

36:29

so we are and still still working

36:33

on developing the first version so there

36:36

there are some features that are already

36:37

quite advanced and that we have tested

36:40

and especially around sensemaking and

36:45

but there are other features that we

36:46

still have to have to fully fleshed

36:48

out like like the final reporting for

36:50

example and and we also need we

36:54

are also working on making our platform

36:56

more more stable and scalable so so

36:59

far sensemaking on the on allows for

37:02

and summarizing individual

37:05

discussion sessions and but one thing that

37:09

we're actually tackling as we speak and

37:11

is to also makes sense of discussions

37:15

and across sessions saw compare different sessions

37:19

with each other and find similarities differences

37:21

in san

37:24

what are we struggling with and

37:30

i think a lot of the

37:33

the the challenges are are under and

37:35

the details and said very easy to

37:39

to produce something that works works eighty

37:41

per cent well and but to actually

37:44

make something reliable and accurate and like

37:48

summaries for example and ninety five percent

37:51

of the time and to make that

37:53

fast and secure and protected data privacy

37:57

and in the process that's the real

37:59

challenge and to anyone who knows how

38:04

to how to make llm systems reliable

38:08

and evaluate them systematically and working on

38:13

know how to do prompt optimization and

38:17

working interested in in and data protect

38:20

one and select anonymization before using a

38:24

lens yet

38:26

at at that would be very helpful

38:27

and then so please please get in

38:31

touch

38:34

was it expensive to be the such

38:37

blocked from or like out there relevant

38:40

cause to i mean i think about

38:42

the transcription they use of ai the

38:44

analysis of the of the transcription

38:50

and the

38:52

yeah was it challenging in a way

38:54

to find the found fundings money to

38:57

build the plateau i dunno also

39:01

if you're working fully full-time on this

39:04

project or if you have aside the

39:06

job and also ah how many people

39:11

are involved in the team

39:14

i like this mani said that that

39:17

the

39:20

makes everything a little bit difficult as

39:23

have heard or saw from other people

39:26

interviewed definitely and so actually the technical

39:32

side of building it was not expensive

39:34

at all and i think our our

39:37

technical costs snack at all

39:39

all llm cogs and hosting costs and

39:43

and you name them their below one

39:46

thousand dollars and saw really not expensive

39:49

that on but that doesn't figure in

39:51

all the like the human labor costs

39:55

and because my my seven co-founders and

39:58

i we have we've been working

39:59

completely for free

40:01

and for two years now

40:04

and some of us fulltime most of

40:07

us next to their phds or a

40:09

studies or a full-time jobs

40:12

i thought of had we had we

40:14

figured and all the costs a labor

40:15

costs than a for probably have been

40:17

much more expensive and and you name

40:21

the key challenge and which is to

40:23

find funding for for this kind of

40:25

project and so

40:29

in a little bit of a tricky

40:32

in between situation where

40:35

i thought so we are a startup

40:37

and may purposely decided to become like

40:40

a business and and not an ngo

40:43

and because we we wanted to be

40:47

independent me wants to be independent from

40:49

public or philanthropic funding reruns of generate

40:52

our own revenues that we could reinvest

40:54

in furthering or

40:55

our mission

40:56

and

40:58

but we are nonprofit startup saw and

41:02

that does not really attracts and big

41:05

funders because the that we with we

41:08

will structurally make it impossible to get

41:11

and

41:13

to privatize profits from what we do

41:17

we wanted to reinvest all our revenues

41:19

all profits into the mission and and

41:22

i puts as and in a very

41:23

difficult place and because

41:26

and makes it hard to apply for

41:28

public funding and my grants and and

41:33

we've been rejected from a lot of

41:35

them i can tell you and at

41:38

the same time it also makes it

41:40

hard to attract more conventional investors for

41:43

example and like vc capital while you

41:45

call it and doesn't speak

41:47

and because they are looking for more

41:48

returns on their investment that we'd just

41:50

can give them and don't want to

41:52

give them however we think my lucky

41:54

to and have half quite of his

41:56

supporters throughout am after all so despite

41:59

all the rejections got and only nice

42:02

needs like a few handful of of

42:04

of people organizations to believe in you

42:07

and as a we like from very

42:09

early on we again to support of

42:11

oxford university innovation and as startup incubator

42:15

and actually gave us a bit of

42:16

funding as bulk and that allowed us

42:19

to i dunno cover cover some costs

42:22

and and recently we got and the

42:25

german gov

42:27

them and startup called a scholarship which

42:29

is called exists and that starting from

42:32

april and that will actually allowed three

42:34

of us to to start working full

42:36

time with a monthly scholarship and also

42:40

quite a bit of additional funding to

42:41

cover other costs like like development for

42:45

example and out that that

42:47

we're really be a game changer for

42:49

us because it's been a long and

42:50

rough two years

42:55

yeah doesn't seem so easy and it

42:58

it is quite crazy i mean because

43:00

i that the this kind of solution

43:04

i mean the

43:07

in a way i dunno in the

43:08

future crew could really avoid the i

43:11

dunno

43:13

something bed

43:16

they could eventually avoid war so civil

43:18

wars i mean we for we forget

43:19

about about all that the

43:22

yeah

43:24

but it was a very common for

43:26

people in the past that the justice

43:28

sometimes kill each other and the i

43:31

think that these is one of the

43:32

most important things like politics and understand

43:35

why people participate and the yeah make

43:40

the process easier for old people to

43:43

to put

43:44

dissipate

43:46

and the

43:48

and yeah i wonder like him

43:51

collaboration in seek deck like

43:56

do you finca of you add the

43:58

collaboration we father

44:01

i dunno platforms or do you think

44:04

that these will be important to

44:10

one hundred percent for sure and very

44:13

important and we have had only good

44:16

experience so far so and

44:21

it's probably like from the perspective of

44:23

established organizations like i dunno go vocal

44:27

make dot org and desert him console

44:31

and not sure how often pro probably

44:34

more often than we think like new

44:36

lay people with new ideas come along

44:38

and and tried to do something in

44:40

the space and

44:44

and while some extent like if your

44:47

he thought about a conventionally like and

44:50

and conventional economic competition terms then

44:57

the the baker established players they probably

44:59

wouldn't welcome you with open arms and

45:02

and the them then they might be

45:04

skeptical like what are what what what

45:06

it is what what are these people

45:07

doing like are they a threat to

45:09

us that that tablet that that's what

45:11

you would would think and eventually i

45:13

guess and but in this space and

45:16

a civic

45:16

tech space i i've i've only experienced

45:18

the opposite and so really all of

45:22

them all the ones that are named

45:23

and more just really welcoming us with

45:25

open arms and being interested and curious

45:28

and would be do and not just

45:31

interests but actually also like pro support

45:34

like realizing or

45:36

okay we are really covering a new

45:38

nice we're not exactly doing the same

45:40

thing so they thought okay does can

45:43

enrich the eco-system saw that's helped them

45:47

and and that's that's transpired and invitations

45:52

and and people and added at the

45:55

has established organization to

45:56

selling their clients their partners about us

45:59

which then led to important partnerships for

46:02

us and just generally multiplication talking about

46:05

us helping helping us gain more visibility

46:09

and

46:11

and so that that has been very

46:13

helpful and for us and generally also

46:17

we couldn't very well combat a civic

46:19

tech association for example even though we

46:22

don't have any resource right now contributes

46:23

to become a member but still there

46:25

they welcomed us and and very recently

46:29

actually and two weeks ago

46:32

at the oh city co-creation boot camp

46:33

in lisbon on using emergent technologies for

46:35

civic participation and we are also cooperated

46:39

with other startups and like ourselves dem

46:43

brain harmony come and to creates a

46:48

a new solution for a consultation analysis

46:52

and and also that was only a

46:55

positive experience and you might think okay

47:00

the startups are doing very similar things

47:02

solid their our competitors and ends to

47:05

admittedly these are also thoughts that had

47:08

previously crossed my mind and because it

47:10

is is really similar but we're doing

47:12

but then just experiencing

47:16

that corporation can actually lead to new

47:20

solution and to him to to kind

47:22

of broadening the

47:25

the the markets so to speak

47:28

i was a really great experience and

47:30

i think permit out there was a

47:32

reinforcement that collaboration always wins over computation

47:38

and because at the end of the

47:39

day everybody benefits if the field growth

47:42

of the industry growth the market growth

47:45

and and ultimately what's most important and

47:48

democracy by

47:48

benefits because of the upper or here

47:50

for the up the same mission yes

47:53

we all need to make our our

47:54

revenues to sustain ourselves that of the

47:57

at the end of the day of

47:58

on to strengthen democracy and and i

48:02

think that that are only goes through

48:04

cooperation and not competing with each other

48:09

i also had the impression that the

48:13

the things that you have said that

48:15

the in some other field the someone

48:17

that does something similar to use our

48:19

competitor while the in this specific field

48:22

the he or she became a collaborator

48:25

this is ozone

48:28

and the

48:29

and

48:32

so a deliberate is a sort of

48:35

standalone plot from and i shingles for

48:37

the same about or monica

48:39

while the city is more modular

48:43

have you had any thought about these

48:45

the

48:50

because i see that the every every

48:52

one is building new platforms

48:55

that is physically different approaches

48:59

do so i wonder a few totals

49:02

about an integration with thus edema or

49:05

like sign of something like this

49:09

definitely i saw the first of all

49:12

and we are thinking mid term long-term

49:15

of deliberate also as more margaret modular

49:18

platform were getting started with one module

49:22

right now and but we want to

49:25

make it's multi modular and the future

49:29

and and that means

49:33

being

49:35

open to other towards integrating with hours

49:39

or our to integrating with other towards

49:43

especially when when they're complementary and i

49:46

think that's really the case for dessert

49:48

in for console for for harmonicas all

49:53

participation processes are very complex and multi

49:59

multi step and i think we're covering

50:04

and different where we're we're focusing on

50:07

a different step in the process then

50:08

for example the decedent and harmony car

50:11

and saw

50:13

we would really love to and have

50:17

conversations with dessert him but also provide

50:19

other providers of online participate in platforms

50:22

because i think they can be

50:24

for example of a great step before

50:27

us citizen the seventy four agenda setting

50:29

for the for instance or afterwards for

50:31

disseminating results and similar to harmonica they

50:36

take a it could serve a function

50:38

like getting a first impression of what

50:41

are the most contentious issues that could

50:43

then be deepened and something

50:45

in the assembly and and so what

50:48

does that mean concretely for us while

50:50

we're we're thinking of the design of

50:52

our platform in a long-term as modular

50:54

and were also thinking about open source

50:58

and and we still have to make

51:00

our minds up about the exact opposite

51:02

strategy that we wanted to pursue and

51:05

that's these our discussion said we're having

51:08

and i think but also often not

51:10

talked about an officer into upper upper

51:12

ability and so kind of our outputs

51:16

are output formats and integrating with other

51:19

other platforms inputs formats and the other

51:22

way round other platforms output formats integrating

51:25

with our and

51:25

put formats and saw that yep it's

51:29

a day they can seamlessly be integrated

51:31

and more different modules from different that

51:34

firms can be connected and tailored to

51:36

the exact sir uses and needs and

51:38

roulette processes

51:42

the boat to interpret ability do you

51:44

know if right now is there any

51:47

kind of protocol for a p i

51:51

related to see with that

51:54

because it's something that the i was

51:55

thinking recently that may be there should

51:58

be some kind of protocol so if

52:00

i want to build a new platform

52:02

i already know how to

52:05

how to be lit club to be

52:07

honest i haven't done a lot of

52:09

research on this yaps and so i

52:12

can't imagine that there's some attempts to

52:14

do it but i haven't come to

52:16

come across them and what i know

52:19

though is that metacafe for example which

52:22

i believe has also funded my money

52:24

cause

52:25

and some i think also a recent

52:28

go vocal project and their recent funding

52:32

call it was also about entire upper

52:34

ability and

52:37

what am i nuts and and not

52:38

not familiar with the outcomes and haven't

52:41

haven't looked into that but yeah i

52:42

i i i see people's to stop

52:45

like start talking about it i've also

52:47

had conversations about it with that civic

52:49

tech association for their attempts to like

52:52

get to that point

52:57

yeah i've also i've seen that through

52:59

some movement inside the mythical about to

53:03

your interoperability because read now there are

53:06

many tools and so will be very

53:08

cool to integrate all of them

53:11

and the you shop do some more

53:15

meals i have a couple of crystals

53:18

of him

53:20

so is if someone as an idea

53:23

related to civic deck to participation how

53:27

will you how cool that person start

53:32

the project

53:35

how it was for you how how

53:36

did you start laker

53:39

so i know it's a bit that

53:42

of a police say but it's true

53:44

i didn't believe it before i did

53:46

myself either i was more of a

53:47

planner and but the the main kind

53:51

of recommendation is to just get had

53:53

started start and start talking about it

53:56

and start doing

54:00

stop endlessly planning and then not implementing

54:04

and just get started somewhere and the

54:08

great news is that nowadays it's

54:12

because of ai it's become hundred or

54:14

a thousand times easier to get started

54:17

there are tools out there were also

54:19

using internally at deliberate that can help

54:22

you prototype in like a minute

54:24

so you have an idea go to

54:27

v zero dot death favor cell and

54:30

within a minute you have a functional

54:32

prototype of your idea

54:35

it might need fewer alterations to actually

54:37

make it nice and but an hour

54:40

later he actually have something that can

54:42

show to people and that will help

54:44

you explain your ideas better

54:47

and and if you want to take

54:48

it further their even tools nowadays for

54:51

for actual software development and web development

54:54

and and thon and i myself i

54:57

knew literally zero about ai before before

55:02

i started this

55:04

the on if the first time i

55:06

ever started engaging with a i was

55:08

touchy between before that i just like

55:11

then completely ignorant of it

55:13

and then tend to be to really

55:15

showed me the potential of it and

55:19

while i i realized the k and

55:21

this is powerful and then i started

55:24

learning more about it i started playing

55:26

around with it added a bootcamp and

55:28

machine learning and data science for three

55:31

months just to get the the basics

55:33

like understand the basic concepts how does

55:35

ai how to a neural and neural

55:38

networks work

55:39

so on

55:40

yeah

55:42

but to be honest if you don't

55:44

have access to boot bootcamp like in

55:45

my case paid by by the employment

55:47

agency in germany he could find a

55:50

wealth of resources on materials online and

55:53

are a website where you can do

55:54

courses and one great example is deep

55:57

learning that i excellent courses very beginner

56:01

friendly

56:02

to just stop familiar familiarizing yourself with

56:05

these more like tech and the tech

56:09

things at and how to say the

56:13

more tech site of of your idea

56:16

but then you can you can get

56:17

started in no time and i am

56:20

not a developer i still don't know

56:22

how to cook

56:22

note but i can use a i

56:24

right now to actually code or develop

56:27

and i think that has to lord

56:29

the entry barrier tremendously for anyone who

56:32

has it has a great idea so

56:34

start doing would be the first part

56:36

of the recombination and start talking about

56:38

it

56:39

the second one might actually be more

56:41

important than the first one because of

56:43

you don't talk about your idea and

56:46

nobody will ever know about it and

56:48

most importantly know at nobody will ever

56:50

come and help you with it so

56:52

no matter how and basic and unrefined

56:56

and flawed your your first initial ideas

56:58

star

56:59

talking about it

57:00

i know it's not easy it requires

57:03

submitted little bit of courage and but

57:07

if you don't believe in your idea

57:10

enough to talk about it then

57:14

while then then maybe you might not

57:15

want to started in the first place

57:17

but if you do believe in your

57:18

idea enough to talk about it then

57:20

go on talking about it

57:22

because then others will get inspired and

57:25

inspired so much the day that it

57:26

will actually want to join you and

57:28

help you create this and i can

57:30

tell you for me over the past

57:32

two years it has been one long

57:35

beautiful and very humbling and are and

57:38

honoring the experience of talking about it

57:40

and then people joining

57:42

and and giving support and making it

57:45

their own idea and yep

57:48

that's at the end of that every

57:50

day but occur recommend to you start

57:52

doing and start talking about it

57:56

yeah as we said the people here

57:58

are really wants to collaborate so talking

58:01

about your idea and network we the

58:04

people if think he is essential social

58:07

the ever more than in other filled

58:11

maybe because of these in specific collaborate

58:14

ocean

58:16

the

58:18

and the do you have any any

58:20

message for the people that are working

58:23

on the

58:25

this kind of tools that are experimenting

58:28

with the

58:30

in the civic tech fields the the

58:32

democracy few the

58:35

something that you would like to share

58:39

lot for for other fellow civic decors

58:43

or a deliberative packers and let's continue

58:46

collaborating and expanding our collaborating and i

58:49

think all of us can only benefit

58:51

from it and and for people are

58:55

working and more more on the and

58:57

the process sides

58:58

of democratic innovation saw people organizing citizens

59:01

assemblies people organizing a dialogue forums town

59:05

halls and so on

59:07

and

59:09

be skeptical of the dangerous of these

59:12

technologies for sure and but also don't

59:15

let the skepticism dominate the conversation cause

59:19

there's so much potential and and using

59:22

the these technologies for goods and especially

59:24

in this context and the that it

59:27

would be a shame if the only

59:29

if

59:29

all we ever kept talking about was

59:31

mustered the dangers of it and that

59:34

let's let's shift the convert let's not

59:36

forget about the dangers but let's shift

59:38

the conversation to what we can really

59:40

use these technologies for because every technology

59:42

can be used for for good and

59:44

bad for bad and good and the

59:46

other way around and that's especially the

59:49

case

59:49

this and democrat in innovation

59:54

thank you thank you are locked the

59:57

for your insight for the conversation and

1:00:00

the issue of any other and anything

1:00:03

that you will like to add the

1:00:04

that may be i didn't task or

1:00:09

otherwise i added all the questions

1:00:12

well thank you very much alessandro i

1:00:16

got our website at www dot to

1:00:19

the red dot com or if you

1:00:21

want to stay at and in the

1:00:24

loop on what's going on follow us

1:00:27

on linkedin

1:00:29

connects also with me on linkedin and

1:00:31

if if you'd like to collaborate and

1:00:34

on the website yields also find will

1:00:35

soon find the next few days a

1:00:37

waitlist so if you are an organizer

1:00:40

was interested in and using our towards

1:00:42

as a pilot partner right now or

1:00:44

as a user from may onwards sign

1:00:46

up to the waitlist and you will

1:00:48

receive other info

1:00:49

what launch and and yet any other

1:00:52

updates firsthand

1:00:57

don't kill getting

1:01:00

thank you