Wikiopinion - Structured opinions - Wikidata

4 Jan 2017

Hi all, I proposed to create Wikiopinion.org based on our work on AgreeList
that might fit into Wikidata [1]. I paste it here:

Storytelling was the most important way to share knowledge for thousands of
years — before writing was invented — so our brains evolved to be
influenced by stories. As Conor Neil explains, many times we are still
“more easily persuaded by one clear and concrete anecdote than by data and
expert statistical analysis”. He says that, “an anecdote is a one off. It
is not data. It is not science. It is dangerous”.

This made me think about two things:

Firstly, people such as Lydia Pintscher of Wikidata and Dario Taraborelli
of Wikicite are working on projects that improve considerably the quality
of Wikipedia and they could even accelerate world’s research.

Wikidata is a collaboratively edited knowledge base:
...

And Wikicite is building a repository of all Wikimedia citations and
bibliographic metadata. The sum of all citations:
...

Secondly, it also made me think about how this relates to the work we have
been doing with AgreeList.com With AgreeList, we are creating a ‘platform
for informed opinions’ that gathers the opinions of leading experts and
influencers and favors the building of rational opinions on issues of key
importance. Our first issue was ‘Brexit’ where we collected the opinions of
almost 2000 opinion-makers on the impact of Brexit to the UK economy,
immigration, politics, and education, and built a summary of opinions on
both sides to inform the public during the referendum. In other words, we
believe in the value of informed opinions over anecdotes and the data of
who agrees on what and why can help us to build our own opinion. E.g. if
NASA, the Royal Society, Obama, the Pope and a friend of mine who knows
more about climate change than me think that it’s real and we should do
more to tackle it, I believe it.

Similarly, if I read something health-related, I can check the number of
doctors who agree or disagree as fast as I see the number of likes on
Facebook. If it is more than 95%, I believe it straight away. Done. I
learned a new thing today. This way we could fight the fact that false
health content seems to be more popular on social media and we could get
informed of more topics than ever.

When we are interested in a topic and have time, we read about it and
contrast different points of views. But when we don’t have time or are not
interested in something, we believe what our culture, friends and
influencers say. And we are so bombarded with information nowadays that we
can’t get informed about everything all the time.

However, when we want to have an educated opinion about a complex topic
such as Universal Basic Income, we can read the arguments and even go to
the sources where we can find more information. We are still building up
the database on Basic Income and it is currently biased towards opinions in
favour given that it is easier to find them given how early stage the
public debate and the AgreeList tool are, but you can see below what
different opinion-makers say about Universal Basic Income via Agreelist:
...

And when there are many opinions, such as on Brexit, we organise them in a
board or summary that aggregates the arguments per categories.

We can also filter them by profession, university, awards (e.g. Nobel Prize
winners), etc. E.g:
...

How did we get this data? First, the data from occupations comes from
Wikidata. Second, the data of who agrees on topics such as these ones is on
AgreeList. These lists are crowdsourced — people add influencer’s opinions.
Users only need to provide a source, for example an article in the New York
Times or the tweet of the person. Moreover, users of the site can vote and
add their own opinions and, at some point, we could aggregate opinions
automatically by semantic analysis. This way we might organise all the
opinions in the world on key topics or statements. AgreeList or Wikiopinion
could one day become ‘The sum of all opinions’.

We can also play with Google BigQuery to do joins of AgreeList’s tables
with Wikidata’s ones. For example, in order to get all Nobel laureates in
economics that agreed or disagreed on Brexit before the referendum we did a
query and we got:
...

Extent is the degree to which they agree (at least for now it can only be
100=agree or 0=disagree). Therefore we got that from all Nobel laureates in
economics that have ever given their opinion on Brexit (on the BBC, their
twitter account or whatever), all 11 of them disagreed. As every
opinion/vote on AgreeList has a source, we see then that 10 of them signed
a letter published on The Guardian and the other one is Paul Krugman who
gave his opinion in The New York Times.

Then, if for example we go to Paul Krugman’s Wikidata page, we see that he
worked for the MIT in the past. What if we want now to get all the public
figures that supported Hillary Clinton or Donald Trump who work or have
worked for the MIT? Easy, we just change the property to P108 (employer),
set it in the where clause to Q49108 (MIT) and select the statement_id=182
(or we could add a new join and specify the title) and the result of this
new query is:
...

We see that from 7 people who are or have been employed by the MIT, 6 of
them preferred Hillary Clinton to Donald Trump.

This is what we have done until now.

The next natural step of AgreeList is to add social network features where
you can see what people you follow are discussing, opinions from the topics
you follow, etc. Next, if I post/agree that governments should do more to
tackle climate change, other users could add then that I also agree that
climate change is real and that we need to act on climate change —
organising in this way the opinions.

Other important aspect could be that AgreeList questions your opinions.
Besides having the list of people who disagree with you (and why) just
there, it could tell you, do you know that Tim Berners-Lee and your friend
from school disagree with you on this? Would you like to see why?
Therefore, this could be a social network which challenges what you post
and help to tackle the fact that the Internet is allegedly full of myths
and mis-truths — as pointed out by Tim Berners-Lee says. Similarly,
Facebook recently said that it must do more to stop the spread of
misinformation on its platform.

This could be specially important because on mainstream social media there
is a filter bubble — as described by Eli Pariser in one of his books.
Social media networks tend to hide the opinions that differ from what we
think. We only listen to the media that agrees with what we think.

This is so significant that Tom Steinberg said that social media giants
will be remembered, not by their business successes, but by how well they
tackle the problem of the filter bubble.

This filter bubble makes the polarisation of opinions worse. You can see
Barack Obama talking about polarisation of opinions:
...

Taking all of this into account, we are considering whether the best
approach for Agreelist is a non-profit project in a new organisation or
under Wikimedia Foundation if they like the idea — it could be renamed as
Wikiopinion.org — or a for-profit startup.

As a non-profit project it would focus on its social impact and it would
follow the three golden rules of the Internet: nobody owns it, everybody
uses it, and anybody can add services to it — which are what distinguished
the Internet from any previous communications medium according to Vinton
Cerf— so initially that seems to make sense.

On Wikipedia, Wikidata and the other Wikimedia projects, facts precede
opinions. So it is not clear if Agreelist (or Wikiopinion.org) would be a
good fit as a non-profit project under the Wikimedia umbrella.

In fact, my friend Ángel Alberich — CEO @QuibimBiomarker & MIT Innovator
Under 35 — says that facts should not be opinable. However, I argue that
knowing exactly who has a different opinion (and why) might be really
useful. For example, more than 100 Nobel laureates signed a letterendorsing
Genetic Modified Organisms (GMO) and challenging the environmental NGO
Greenpeace to halt its anti-GMO campaigns to prevent the introduction of
potentially life saving options for the world’s poor. Ángel said that there
is not a single scientific paper which says that they present a hazard for
human health. But as Greenpeace and many governments are so reluctant,
isn’t there a clear need for something else that facilitates discussion?
Would it not be useful to know which ones of your friends and
representatives disagree and try to convince them?

Similarly, despite the fact that almost all scientists agree that climate
change is real and that we need to act, there are still many politicians
who deny it or don’t do enough to tackle it.

However, as Greg Mankiw — Harvard professor in economics — said in Leonardo
DiCaprio’s recent film about climate change, if we want to change
politicians view on something, we have to change people’s view first.

Therefore, a social network which organises opinions and challenges what
you think might help to do so. Actually, Barack Obama recently said on
Wired that at some point we might make voting and civic activism as
addictive as scrolling through your Twitter feed.

Let’s do exactly that, a social network where we discover, share and
organise a plurality of opinions where the objective could be to help us
make up our mind. In other words, to accelerate quality decision making.

And this could be really important because of three reasons:

As the MIT professor Alex Pentland said:

The biggest problem in the world is not global warming, is not war, but how
can we organise among ourselves to make good decisions and carry them out.
...

2. According to Terry Jones— disruption occurs when new technology allows
us to deliver new forms of asynchronous communication. And this is what
AgreeList is about. Until now, if you wanted to get a quick opinion on
Basic Income and the arguments on both sides, you had to go to many
different sites or talk to multiple people. Not any more.

3. As the economist Jeffrey Sachs said in his book The End of Poverty, the
single most important reason why prosperity spreads is the transmission of
technologies and the ideas underlying them. But currently we don’t need
more bridges or faster communications (in the privileged part of the
world), but making up our mind accurately in the myriad of the new topics
that arise in our hectic lives.

To sum up, we think that a social network that challenges what you post and
organises who agrees on what and why would complement Wikipedia and the
traditional story telling. What do you think? Would you like to join us?
Should this project be non-profit or for-profit? Would you donate or help
us to fund raise?

Kind regards,

Hector

[1]. Original post: https://medium.com/@HectorPere
z/wikipedias-social-network-578b0257b8ae