Zenologue

One of the intriguing things about Twitter is its complexity disguised as apparent simplicity. On the face of things Twitter appears to be little more than a very simple communication system involving the exchange of short (often meaningless) messages, very similar in concept to a text message on a mobile phone. In fact, text messaging was the fundamental inspiration for Twitter in the first place. Under the surface, though, things are more complicated, with properties emerging from Twitter that may not have been foreseen in the beginning.

The Power Of A Few Words

Most tweets can be divided into one of several categories: 

• Mundane items of little interest to anyone with a life
• Attempts to sell get-rich-quick schemes
• Promises of thousands of new followers (none of whom are interested in you)
• Other types of spam
• Political and religious messages

But there are some that are completely different and are of a more advanced or intellectually satisfying nature. It’s this group of tweets that could potentially form the basis for a whole new world…

This small group of tweets has been used to compose poetry, write plays, spread news, warn other users of Internet threats, exchange product reviews, sway public opinion, influence political campaigns, reduce the spread of computer viruses and help people get released from foreign jails. The entire world seemed to grieve as a collective on the day Michael Jackson passed away, straining Twitter to its limits in the process; the list goes on.

All of this from answering the simple question of “What are you doing?”

I’m not sure that Jack Dorsey, the creator of Twitter, envisioned these emergent properties when he unleashed his creation on the world.

Ants Tweet Too

Complex emergent phenomena are usually the result of swarms, or crowds of individuals. Mobs and riots are one example, where the size of the “swarm” reaches a critical mass, resulting in the group suddenly acting in unison, usually in a violent or destructive manner.

Insect colonies are a prime example of swarming behavior, and watching Twitter is sometimes reminiscent of observing the social complexities of an ant colony. Individual ants by themselves are very simple creatures that operate by simple rules. There’s really no room in an ant head for a brain capable of understanding the principles of architecture or the strategies of waging war with enemy invaders. No single ant knows how to construct an entire nest, complete with tunnels, chambers, gardens, kitchens, nurseries etc. Yet, the swarm as a whole accomplishes this task with amazing ease and efficiency. The ant nest and its incredible structure is an emergent property of a swarm of individuals. Furthermore, the structure of an ant colony is not predictable in any way by examining individual ants.

People are quite different to ants, and most of them are somewhat more intelligent than an insect, but Twitter strikes me as a good analogy to an ant colony: Twitter is essentially a simple system with a few easy to understand rules. Yet the behavior of the Twitter system as a whole shows signs (albeit small ones at the moment) of emerging properties that are generated by its millions of users and the different ways those users are putting Twitter to work for them. 

What Are You Doing?

What causes Twitter to exhibit complex behavior from an apparently innocuous question where the answer must be in 140 characters or less?

The answer lies in Twitter’s basic concept – the communication of a text message. Text messaging and Twitter are basically the same. However, there is one major difference, and it’s that difference that makes all the difference in the world; it’s the electric spark that breathes life into the creature that Twitter has become. Text messaging doesn’t exhibit any complex emergent properties as far as I know; it’s a one-to-one messaging system. Twitter, on the other hand, expands and multiplies the influence of the text message into a powerful one-to-many system. This amplifies the message so that millions of people can potentially receive it almost simultaneously.

Let’s return for a moment to our comparison of Twitter users with ants. We can consider the ant’s chemical messaging as analogous to a tweet – “food this way”, “follow this trail”, “defend nest” etc. From this, we can imagine how no ant colony could function properly if ant-tweets were received by only a single ant at a time. Communication in ants, the foundation of their complex behavior, is by necessity a one-to-many system that both distributes and amplifies the message and causes complex behavior to occur. 

One Day Twitter Might Be As Smart As An Ant

Thinking about this further, I’m tempted to go out on a limb and suggest that an ant colony, with the benefit of millions of years of evolution, is way more advanced than Twitter is right now. Twitter, in its current state, is like an ant colony gone mad, where most of the messages are disorganized and obscured by meaningless noise. It’s as though all the ants in the colony are all spouting random messages at the same time in a raucous rabble. If that were to actually happen, the ant colony would probably collapse.

Being optimistic, though, I suspect that the overall signal-to-noise ratio of twitter will grow stronger as time goes by. Already, there are pockets of meaningful information out there. It’s a shame that much of that data is masked and drowned out by announcements of the eating status of millions of household pets or the endless repetition of inspirational quotes. Do ants make “noise” too? I’m sure that not all “ant tweets” are perfectly meaningful in the context of the collective need. If that’s true, how are the less useful or incorrect messages filtered out from the real signal to enable the colony to function as it does? If we can find the answer to that question, maybe we can learn how to apply that principle in “tuning” Twitter, to amplify the signal and quench the noise.

You Will Be Assimilated

As any fan of Star Trek: The Next Generation knows, one of the worst fates is to be assimilated by the Borg, a race of evil cybernetic humanoids organized into a swarm (“the collective”) and operating as a single organism. Assimilation meant being absorbed into the collective and becoming a part of the hive mind, sacrificing one’s individuality in the process. Not a pleasant thought for anyone.

While I doubt our own individuality is in too much danger of being assimilated by the Twitter universe, there are some striking similarities between the Borg and the fetal creature currently we know as Twitter.

For example, take the ability of many unrelated individuals in the network to unite as one, in a concerted effort to solve a problem. This happened recently when Demi Moore alerted her Twitter followers to a woman threatening suicide in San Jose. Within moments, the news spread, the local authorities were alerted, and tragedy was avoided.

Or, consider the Twitter Opera; a collaborative work constructed almost entirely of tweets. I’m not sure how “good” the opera turned out to be, but the main point is that it is a work of art made by otherwise unconnected individuals. The question remains of whether or not it is true art, unless it is somehow an expression of the collective consciousness that created it.

Then there’s the Twitter Psychic Experiment conducted by Professor Richard Wiseman. In the experiment, he asked volunteer Twitter users to try to “remote view” the location he was in and select it from a set of five photographs. In each run of the experiment, the collective decision was incorrect suggesting the remote viewing is not a real phenomenon.

Finally, I think one of the really interesting, and possibly useful, possibilities of something like Twitter is as a problem-solving machine…

Welcome To The Machine

We’ve already seen examples of using the Internet as a means for distributed computing. For example, SETI Online, the computational modeling of protein folding and the human genome project, all of which harness the power of many thousands of computers to process small portions of complex tasks that would take a very long time, even on a single supercomputer.

This caused me to wonder if Twitter could eventually be used as a problem-solving computer using principles similar to that found in the wisdom of crowds. Such problems could be the sort that traditional computing might be unable to solve; problems involving concepts that are difficult to express numerically or in a logical form.

Here is a very simple example (simple in the extreme, but it serves the point). “Do I take an umbrella with me to work today?” The answer depends mainly on how likely it is to rain where I’m going, but also on such factors as how much time I intend to spend outside, how windy it will be, how convenient it is to carry, where am I going, what alternatives there are to an umbrella, and so on. Can the wisdom of a crowd decide the answer to this kind of question intelligently? I think maybe it can, as long as the crowd doesn’t fall prey to anti-productive concepts that might lead to a herd mentality or collective irrational fears or prejudices.

This possibility raises some interesting questions, which others might want to try to answer (maybe through Twitter itself as a sort of meta-question).

How many followers might it take for us to effectively solve a qualitative problem that involves rationalization and reasoning rather than mathematical or logical processes? I’m sure the answer probably depends on the complexity of the problem and many other factors, such as when the question is asked.

Is the time taken to solve a problem directly proportional to the number of followers working on it, or might we see a synergistic, exponential or logarithmic effect? That is, would 10 followers be 10 times more effective than 1 at solving the problem, or 100 times more effective?

Conclusion

Twitter is still very much in its infancy at the moment, barely able to feed itself, let alone talk coherently. But it’s learning at a fast rate, and I don’t think it’s going to be too long before we start to see even more startling and amazing things come out of it that will catch us all by surprise.

As a very basic test, I posted a tweet that asked, “what is the square root of 12,345,678 to four decimal places?” Anyone with a calculator can work that out in a few seconds. Twitter took 28 minutes. Maybe the question wasn’t interesting enough to get an answer immediately, but the point is that it did answer the problem. At least it didn’t give me the answer of ‘42’, in which case I would have been a little concerned…

Is it possible that Twitter, or something like it, could be the future answer to artificial intelligence? Such an AI system would be made up of millions of independently smart nodes, each one autonomous in its own right, yet able to function as a small component of a much larger decision-making system. It would be like an ant colony where the ants are each as smart as people…

Maybe the question Twitter should be asking isn’t, “What Are You Doing?” It ought to be “What Do You Want Me To Do?”

General communication, complex system, emergent property, tweet, twitter