IEET > Vision > Bioculture > Contributors > Michael Hrenka > Sociology > Psychology > Technoprogressivism > Artificial Intelligence
Solving Problems With Collective Intelligence – Towards an Internet of Thinkers?
Michael Hrenka   Oct 9, 2015   Ethical Technology  

How can technology that we are able to build with today’s tools help us to solve the big problems of individuals, organizations, and the world at large? More specifically: How can we use the internet in the best way to improve our collective problem-solving capabilities? Questions like these don’t seem to be asked very often, perhaps because people usually focus on specific problems, rather than general problem-solving in its own right.

Today, a vast plethora of different websites, online platforms, and apps exists. Currently, the web is dominated by what can be called web 2.0 platforms, which facilitate social interactions and collaborative co-creation. Can these platforms help us to use some kind of global Collective Intelligence (CI) that is actually good at solving difficult problems?

The answer is probably yes. Nevertheless, there doesn’t seem to be one single absolutely prominent platform that is really dedicated to solving serious real-world problems by using CI. Again, this may come from people not seeing themselves as problem-solvers, or not associating the internet with solving big problems – as opposed to solving “minor” problems like boredom.

An Ordering Framework

In order to classify different online platforms and apps in respect to their fitness for problem-solving, I came up with a simple conceptual framework that is based on different operations that can be done with information. This is not a grand effort to explain problem-solving in general, but rather a mental tool that should help us understand how different apps support different aspects of problem-solving.

When people solve problems they usually either think about the problem at hand, or at possible solutions for that problem. In both cases they need to operate with information that is relevant to the task at hand. This is why the framework I present here is based on operations that are applied to information. These operations are: 

1. Collection: Information is collected at one specific location, for example one website, or eventually in the mind of a person 

2. Generation: New information is generated be processing and recombining already existing information, first in the mind of a person, and later on media that are accessible to other persons 

3. Evaluation: The quality, truthfulness, and utility of information is judged by some entity – then that evaluation is made available in some form for other entities to see 

4. Distribution: Information is transmitted actively to other entities, so that these are likely to make use of it

According to the starting letters of each operation, I tentatively call this conceptual classification system the CGED Framework

Collective Problem Solving Online

To see how this framework can be used to understand how online tools can help us solve problems, I will list some of them which are specially good at one specific operation. After presenting those tools, I will describe four different general use cases for them.

Collection Tools 

Google: Since information is spread out throughout the whole internet, search engines like Google are a real necessity to find any kind of information that is not located within one specific platform. 

Blogs: It’s a bit of a stretch to say that blogs excel at collecting information. Regardless, individuals and organizations occasionally do collect high quality resources about specific topics on their own blogs. WordPress and Medium are among the most prominent blogging platforms. 

Wikis: Wikipedia is still at the forefront of turning the collection, structuring, and curation of information into a collective task for all motivated and knowledgeable users of the World Wide Web. However, wikis have lost popularity over the last years. Still, for some organizations wiki-software like MediaWikiDokuWiki, and wiki-like tools such as Confluence can still be useful for knowledge collection purposes. 

Version Control Systems: When it comes to collaborative software development, using a good version control system like Git is very useful for handling changes to the code in a well structured way. A lot of open-source software is developed on the popularGitHub platform. It not only stores the code in a location that is accessible to everyone with internet access, but it also allows programmers to contribute to such open-source projects in an orderly fashion, or even fork an alternative version of some software. 

Generation Tools 

Etherpad: Although blogs, wikis, and GitHub could also be seen as collaborative information generation tools, they all lack an interface that allows simultaneous real-time editing of texts. Etherpad can do just that! Using a tool like Etherpad, especially in conjunction with voice chat software like Teamspeak or Mumble is one of the most practical and effective ways of collaborative text and code creation. 

Google Docs: Where Etherpad restricts itself to rather simple text files, Google Docs brings simultaneous real-time editing to all kinds of “Office” documents. 

Evaluation Tools 

Reddit: What made Reddit different from previous online discussion boards was its central functionality to rate user contributions up or down. Of course, users could give more detailed evaluations of certain posts, but having them available as binary votes allows Reddit to sort posts according to their up- and down votes, as well as reward contributors who have written many popular post with high “karma” scores. 

Q&A Sites: Getting good support form experts in their fields was never as easy as it is today. Q&A platforms likeStackExchange and Quora also use simple rating systems like that from Reddit. This has many advantages: People can solidify their esteem as experts by accumulating upvotes, and users can see how other people rated the replies from those experts. Overall, these simple mechanisms have improved the quality of answers to all kinds of different questions significantly! 

Yelp: Essentially, Yelp is an app that enables users to rate small businesses, and to see the reviews of other customers. This crowdsourced customer generated business evaluation could be seen as a CI-based service for people to find the best local restaurant or hotel.

It is worth noting that such evaluation tools are generally prone to manipulation, both by malicious users (e.g. Sybil attacks), as well as the organizations that provide these tools (for example fraud or questionable business practices). Countermeasures against these issues exist, but are difficult to implement, and may come with severe trade-offs.  

Distribution Tools

In our current decade, social media platforms have become a quite dominant part of the distribution of news and opinions. 

Facebook: As most popular social network of our time, it is difficult to avoid using Facebook, because it makes it so easy to stay in touch with other people. News and opinions can be sent to other persons directly, posted on a personal wall, or shared with a specific group of users. With its huge number of active users, this makes Facebook a very effective tool for spreading information, especially due to the fact that you only get information from your “friends” and the groups you’ve joined (and the occasional advertiser which usually gets ignored). 

Twitter: The limit to 140 characters per tweet enforces users to compress their messages. This makes Twitter a very information dense medium that is suited for distributing messages that should catch the attention of people quickly. 

A typical use case for these platforms is that blog posts are shared on them. This grants such blog posts a wider reach, but often comes at the price that the discussion of those posts largely happens on Facebook, rather than on the blog in question. 

Use Case 1: You Have a Problem That an Expert Knows How to Solve

In this case, typically the best solution is to research the answer on an expert’s blog, or post your question on a fitting Q&A site. That is, if it’s a problem that is hard enough to warrant more than a simple Google search.

Use Case 2: You Want to Work With a Group of People Dispersed Around the World on the Solution of a Problem

First of all, you need to use suitable communication tools for staying in touch with them. Those can be anything from email over (voice or video) chat apps like Slack to online forums, for example subreddits or a Discourse forum.

Then, you probably need a place to create some content collaboratively. If it’s code you want to create, then GitHub would be a fitting choice. If you want to produce some simple documents, use Etherpad or Google Docs.

Use Case 3: You Want to Share Your Solution to a Problem

You might want to use a blog to publish your solution and then encourage people to share it on social media. If it’s a problem that bothers specific people, find out where they communicate about that problem, and then share your solution on the platform they use.

Use Case 4: You Have a Big Problem That Nobody Knows a Definite Solution For

This is where the current online tools start becoming insufficient. If you have a lot of money at your disposal, you can purchase the services of an organisation which might possibly solve your problem. Alternatively, you can actually try to leverage crowdsourced CI by framing your problem as a challenge on ideation platforms like Innocentive and rewarding the best solution with a prize.

But it’s very likely that you are unable or unwilling to do either of that.

If you have an awful lot of time and energy, you can try solving the problem yourself, or with like-minded people (see use case 2).

What you cannot do easily or effectively though, is delegating the problem to the collective intelligence of all interested internet users. There is no actual “global brain” to ask. Nevertheless, there are some developments which might make the creation of such a CI-based entity more viable.

New Developments

The internet of the future will be a quite different place, which will certainly unlock exciting new possibilities. Some of these possibilities are already materializing, and could be used right now to create better CI-based problem solving tools.

Big Data and the Internet of Things

The increasing collection of all kinds of data is certainly problematic, but big data and the Internet of Things (IoT) could allow us to increase our knowledge of ourselves and the world we live in. A key question is of course: Who will own all that data? “Corporations creating their own centralized and private data silos” is certainly one of the worst answers.

Can we find a satisfying balance between privacy and openness? The more data is freely and openly accessible to all, the more it can be used by (citizen) scientists and analysts to answer interesting and difficult questions.

The Blockchain Revolution

The bitcoin currency has introduced the blockchain as tamper-proof decentralized data storage. Not only has this new data structure enable viable decentralized cryptocurrencies, it also spurred the creativity of many brilliant minds who envision a society based on decentralized networks and services.

Blockchain-based decentralized currencies, marketplaces, identity and reputation systems, websites, apps, social networks, organizations, and even governments are already being created.

One of the most useful-looking apps currently in creation is Augur, a decentralized prediction market. In a prediction market people bet on future events. This is a form of CI that is supposed to produce predictions that are superior to predictions made by individuals.

Reputation and Attention Economies

Reputation systems are getting increasingly common, as they serve the function of facilitating trust between people online. They are even becoming so ingrained in different platforms that there’s a need for inter-platform reputation, and there’s already at least one popular network for that: Traity.

But reputation has the potential to become a cornerstone of a new network economy. This can be seen in the revolutionary decentralized social network Synereo, which not only has reputation scores for its users, but also rewards users for the attention they allocate to shared content! Users will actually get rewarded more, if they have a high reputation score. Synereo’s approach actually combines reputation scores, attention measurement, and cryptocurrencies in a single elegant system.

A similar reputation-based system, which is not restricted to the context of a social network, is my own conceptual Quantified Prestigesystem, which lets users allocate “esteem points” to other users freely. How many esteem points a user gets determines his reputation score in the system. This “Prestige” score can then be used to generate reputation incomes for the users of the system, which are proportional to Prestige.


Using “game mechanics” to improve the interaction of users with a system is typically called “gamification”. Badges rewarded for certain achievements in an app are a typical, yet rather shallow example for gamification. Actual games can be used to achieve gamification of difficult problems like protein folding – see Foldit!

It’s also possible to gamify education, training, work, chores, and even habit creation. Chore Wars and Habitica are games that help you do just that. Accomplishing self-defined tasks is rewarded with virtual gold, experience, and virtual vanity items.

Even the World Bank applied gamification in 2010 when it created an educational Alternate Reality Game called Urgent: Evoke that was aimed at teaching social innovator skills and encouraging players to start their own social projects and enterprises. The success and message of that game was actually examined in a paper called “A Parallel World for the World Bank: A Case Study of Urgent: Evoke”, An Educational Alternate Reality Game written by David I. Waddington from Concordia University, Montreal, Canada.

Finally, there’s a game-like app for creating predictions based on so-called “swarm intelligence”, a special form of CI. This app is called UNUM, and it’s a project that’s currently in the beta stage. Whether this form of swarm intelligence will actually prove better results than other approaches remains to be seen.

Artificial Intelligence

How will CI interact with artificial intelligence (AI)? First of all, AI is already used in the search engine Google and the computational knowledge engine Wolfram|Alpha. In these cases, AI is a basic tool for the collection of information and knowledge.

Furthermore, advanced AI systems based on deep learning such as Watson excel at an ever greater range of tasks, often outperforming human experts. In some cases, Watson can solve certain problems on its own, such as diagnosing diseases. This can enable humans to focus on the virtually infinite number of remaining problems.

This means that humans will work alongside with AI to solve really challenging problems through a CI approach. How this interplay between AI, single human intelligence, and CI will play out is a most interesting question that will encourage very creative and unexpected answers.

Technological Unemployment and Universal Basic Income

If AI actually goes ahead to replace the jobs of a majority of the population, and we as a society fail to create enough sufficient replacement jobs, this will lead to significant technological unemployment. And if this challenge of technological unemployment is met with the implementation of a Universal Basic Income (UBI), paid out unconditionally to everyone, this will create unprecedented opportunities!

With a traditional job market that is unable to provide gratifying occupations for people, they will seek alternative alleys towards meaningful activities. This is because people have psychological needs that are typically met by work, such as developing competence, or relating with other people in a meaningful way (though these needs are not fulfilled by bad jobs). Even if people won’t get rewarded monetarily by alternative activities, they will want to pursue them. Solving problems by interacting with CI networks is probably the best alternative activity that unemployed persons could engage in.

Brain-to-Brain Interfaces

Although brain-to-brain interfaces have already been demonstrated in early experiments, they remain out of reach for any practical use, yet. This will probably change in the coming decades. And then we will have whole new dimensions for experimentation with CI.

Towards an Internet of Thinkers

I’ve been exchanging ideas eagerly with Ken Carroll who is committed to creating a global problem-solving platform that leverages CI. Inspired by the Internet of Things, he envisions an “Internet of Thinkers”. Even though he definitely has some interesting visions about what this Internet of Thinkers could be, he started an open ideation round in order to find out what other people think about such an idea.

Everyone can participate in this first ideation phase by visiting and filling out the feedback form.

Early Results From the Open Ideation Phase

There are already some preliminary results from this ideation phase. I’ve analysed these results according to the frequency of words that were used within the submitted answers. Here are some of the most interesting words that have appeared so far (primarily ordered in descending order of frequency, secondarily ordered alphabetically): 

* people 

* ideas 

* internet 

* another 

* thinkers, thoughts 

* already, problems 

* crowd, system 

* everyone, individuals, thinking, topics 

* Facebook, idea, open, platform, social, solutions, world 

* intelligence, organizations, place

Unsurprisingly, the Internet of Thinkers mainly seems to be seen as people having ideas on the internet. “We don’t need another [enter your favourite online platform]” was a theme that was relatively frequent, too. 

Combining many of the frequent words into a coherent expression, the Internet of Thinkers should be: An internet based open and social crowd intelligence system, used by thinkers and everyone else, thinking about the problems that people, individuals, organizations, and the world already have, as well as related ideas, thoughts, topics, and solutions.

Critical feedback about the idea of an Internet of Thinkers mentioned the thread that it could become just another mere “talking club” without following up ideas with real actions. Also, there was the fear that it could become another governmental or corporate platform that wasn’t really open, and instead inhibited the free flow of ideas by declaring them as intellectual property. Finally, the issue of “quality control” of contributions was raised. If the Internet of Thinkers is a system that everyone can use openly, then the quality of contributions will vary greatly, so it requires mechanisms which promote and reward good contributions.

What Should the Internet of Thinkers be Exactly?

If we accept the definition of the Internet of Thinkers above, at least for a moment, then there’s the open question what a “social crowd intelligence system” should be exactly. After all, there are already platforms that come close to being just that, especially the various different Q&A and ideation sites. What could make an Internet of Thinkers really different from those platforms? Perhaps the answer is connected to the new developments I’ve outlined above, or it lies somewhere else.

In any case, more thinking and discussion is required to figure out the best way to move this idea ahead. You can start right now by commenting on this post, or by engaging in a sustained debate about the Internet of Thinkers on the Fractal Future Forum.

Michael Hrenka is a philosopher from Germany who studied mathematics and physics. He's interested in futurism, transhumanism, and related economic topics like Universal Basic Income and digital abundance, for which he developed a reputation economy system called Quantified Prestige. Besides hosting the Fractal Future Network, which is dedicated to envisioning and creating a better future, he also writes on his personal blog


Great Article… but all the systems you mention are old-school “asynchronous” systems where groups don’t work in REAL TIME.

You might look into “synchronous” methods… like Human Swarming, which allow groups of networked users to actually THINK AS ONE.

You can read the basics of HUMAN SWARMING from this IEEE paper:

Or you can see a real-time HUMAN SWARM here:

The article actually mentions both “asynchronous” systems as well as “synchronous” ones. Etherpad and Google Docs allow for real-time collaboration on documents. Chat rooms are also real-time applications.

The Human Swarming you’ve mentioned is referred to as “swarm intelligence” in the last paragraph of the “Gamification” section. The Human Swarm video you’ve linked displays the UNUM system in action, which is mentioned in the same paragraph. In any case, this is a very new and exciting development. I would love to see this general approach see applied to other areas.

An intelligent combination of “synchronous” and “asynchronous” approaches would be quite interesting. I haven’t seen any app or platform that integrates both of them in a really optimal way. For example, I would love to see some kind of structured forum that included real-time editable wiki-functionality. That would be great for real-time collaborative teamwork online.

This looks like a good way of evaluating software tools for CI-aiding potential, and an excellent analysis of existing platforms (though I notice you left out Google+—I have recently found it to be a good place for some kinds of shared-goal-oriented open discussions).

I recently wrote up some ideas for a new kind of network and software tool for CI, aimed towards “collaborating on complex problems”. You can see it as an attempt to fill exactly the hole you pointed out in existing platforms, i.e. being targeted towards large problems. I’d be very interested to hear your thoughts on it.

“My first blog post is about software for collaborating on complex problems by visualizing the semantic network of a “global brain”—a graph of everyone’s ideas and reasoning involved in the discussion.”

(I’m working on a more technical followup post.)

Comments from Kim Cosmos:

Google docs does have commenting now, but it has no branch and merge functions other than comments. Comments can be fact, consistency and usability checked before merging. However it has no advogato/slashdot style rating or filtering system only the centralised owner, editor, commentator, viewer heirarchy. is going to be my future community platform as I move away from facebook. It is a realtime collaboration tool far more advanced than google docs, built upon the opensourced google wave. The static results can be posted to facebook for the 90% lurkers. It includes an advogato/slashdot low reputation filter and the peer to peer barter system a as used in Japan.

When Synereo makes its glossy advertising real I will jump from facebook to that, It does contain the ability to buy attention whose income is shared with the advertisee but it may need a mod to block this.
If you have been collaborating in facebook and a (probably progressive) moderator decides to block you or delete the group all your posts and comments will be lost. You should backup your facebook posts and comments regularly and permanently in the cloud. Probably dropbox if you are on your own computer or google drive if you have gmail and use a cafe (as poor people do).

Here are the offline collaborative tools that need to come online.

Also it was suprising not to see delegative democracy in the form of not get a mention. It has legal force for all existing proxy voting systems and was developed for the german pirate party. Also has a categorisable delegative ability and very good smart phone support which has proposing, discussing and voting with blocking and is integrated into diaspora. It is used in a number of political parties but mostly collectives.

For industry standards the BPMN language is now infiltrating all business software through Cloud Application Management for Platforms (CAMP) Adapting this to grass roots backoffice control of workflow is not too hard for big enough groups.

YOUR COMMENT Login or Register to post a comment.

Next entry: Gun Control’s MP3 Moment

Previous entry: 3D Printed 9mm Semi-Automatic