Tim Urban's Tweet
A while ago, one of my favorite bloggers, Tim Urban, tweeted a tweet in which he expressed sadness about the fact that Nature (the scientific magazine) was posting too much political content.
As it unfortunately often is the case today, this, in my opinion innocent tweet, caused an all out war in the comment section with dozens of people attacking him.
This being sad, but not surprising, I noticed one other thing however. Consider this tweet: (I intentionally deleted the parts that would identify the topic so that I wouldn't get accidentally flamed as well:)
"Come on Tim, you are better than that...remember when you published that multipart hagiography of Elon Musk on yours? What happened to you, dude?? Tim you're better that this. You've had some pretty yikesy takes lately. Hoping it's just a phase and I legitimately don't see the problem here. Very silly tweet. Even if that was a mistake—and I very much hope it was—you’re coming across as what you once called a zealot. ....is this performance art?"
What is interesting about it is, that it is actually not one tweet. These are 10 tweets posted by 10 different people which I pieced together. To me they legitimately seem like written by one person.
Diverting slightly, more and more I've been thinking about the fact that people connected to the internet have turned into this huge neural network where each one of us is acting like a single neuron in that network.
Within this network there are no geographical or physical boundaries and 'superorganisms' emerge very easily and very quickly. When looking at that tweet made up of 10 tweets I was trying to understand this particular superorganism and how and why it formed and I realized it formed around a specific idea - a feeling which these 10 people had and felt the need to express.
The interesting thing here for me is the following: it is impossible to stand against these superorganisms for any one single person. While in real life, you might be able to out-argue 2-3 or, who knows, maybe even 4 people at the same time, it is impossible to argue against 500 or 5000.
While in real life, you might be able to hold your grounds against 4 people, when online trying to stand against a cloud of 500, there always will be someone who maybe might not understand the argument in general better than you but might understand this or that aspect of the argument you are discussing slightly better than you and will attack you from that angle. This may be valid for another 100 people for 100 other angles of the argument you are talking about. Now, even though individually you might be able to out-argue all of them, when they are all behaving like one superintelligent superorganism, it is impossible, because if you piece their 100 excellent niche arguments together, as a whole they form an argument which is approaching perfection.
(notice, I'm using the word 'argue' because I feel like conflict illustrates the behavior of these 'superorganisms' best, not because I like to argue:))
The group is smart
I might not remember the facts of the following story 100% correctly, but the gist of it went something like this: at a rodeo in the US, a bull was shown to an audience of about 100 cowboys once and they were told to try to guess the weight of the bull just by looking at it. Whoever's guess would come closest would win the bull. I don't know who won, but the most interesting thing was, that when the organizers averaged out all the guesses, the average number matched that of the actual weight of the bull. The key takeaway here is, that a group can sometimes be very intelligent. If you allow individual elements of the group to communicate instantaneously (like we do online) it really does become an intelligent superorganism.
As Elon Musk put it, before the discovery of printing press, people essentially communicated through osmosis - just like that children's telephone game. You tell something to someone, that person then says that thing to someone else and that is essentially how information spread. Then before the advent of internet, there was television, radio, books and libraries, but all of these had one thing in common - they didn't allow instant mass communication between people.
With the advent of internet, people were not only able to communicate instantly in huge groups, but given how information is stored online (say a comment section under a news article is a collection of comments that spread sometimes through several weeks or months) they were also able to communicate in time.
Returning to superorganisms, I would argue that as soon as you have two people communicating, whether instantly or through 'time' (for instance a person replying to a comment under a news article which is one month old) the superorganism is already created. So it essentially takes two people to create an organism which is more intelligent than its individual parts. My estimate would be that, on a long enough projection, the more elements there are in the group the more intelligent this organism should be so imagine an organism of 1 000 000 people in it and the intelligence it has.
Notice, I'm not talking about countries, companies, ethnic groups, etc. I'm talking about online entities organically forming around ideas, political views, feelings that have largely flown under my radar as being 'one' organism and I've always looked at them as individuals sharing the same idea, rather than a 'living' group breeding of that idea.
Diverting slightly again, in this sense, I was also thinking about why the app Clubhouse was such a pleasant change compared to other social networks for me. One of the reasons why it didn't bother me was that it is much more difficult for these superorganisms to form on this platform.
Except for the obvious - I simply didn't enter rooms which were discussing topics I didn't like and kept to the ones I found interesting - there is no comment section on Clubhouse and comment sections are a breeding ground for these superorganisms. Also, usually only one person can speak and people listening are unable to communicate with each other and as I wrote before, I feel like 2 people already do create an organism more intelligent than both individuals.
The Future
Doing a bit of a thought experiment about the future and how things might evolve, maybe in the future everyone will have their own AI systems dealing with this. As mentioned, one person cannot out-preform an intelligent superorganism like that but AI might.
This also brings up another question: A lot of people worry about the future of AI and how it might wake up one day (or some say it might already be awake, but intelligently pretending to be silent), how it is dangerous to create something that is more intelligent than we are, but I would argue that the internet has already made it possible for the creation of these superintelligent online entities made up of thousands if not millions of people in some cases, organically grouping around ideas, feelings, interests, political views etc. that are far more intelligent than any one individual or even intentionally formed groups (decision making bodies in companies, governments etc.).
The question is what steers the 'consciousness' of these superorganisms. While if we have a decision making body in a huge company like Google, probably what is happening is that how Google behaves can be traced back to the conscious decisions of a few key decision makers.
But what steers the consciousness of these online super entities is far more complicated to understand. They group around ideas and their behavior reminds me of huge flocks of birds that can be seen flying over Rome:
What a great article, thanks for sharing Vlad!
ReplyDeleteAs we're in a process of creating a superorganism right now, let me expound on one point.
In the last section of the article you wrote: "But what steers the consciousness of these online super entities is far more complicated to understand. They group around ideas"
I like to think of individual human beings as different petri dishes. Each person gravitates towards different ideas, different aspects of certain ideas and ultimately learns more about certain parts of the aspect than about the others. In this sense, each person could be considered a petri dish, different set of predispositions & gravitational pulls...
Now imagine 10 people in the room. You ask them a question about possible AI dangers. And each person will give you a slightly different answer (there will be some overlaps and some dissensions).
Now at this point I need to add that my understanding is formed mainly by my meditation practice. If you seriously consider what happens when one is confronted with a question about AI dangers this is what you get: One listens to the question and suddenly one gets one, two or three possible ideas popping up in one's mind out of nowhere and according to one's petri dish (predispositions/inclinations), one goes with one of them. One cannot influence WHAT pops up, and one cannot even influence what one "chooses". It's very similar as when you come to the kitchen where you have your coffee machine, tea pot and mineral water. The "ideas" will appear by themselves ("Let's get coffee/tea/water"), and according to many different variables one gravitates towards something. I like to call these predispositions/tendencies/gravitational-pulls one's "programs". Each person is basically a set of "programs". Some programs are quite universal like "hungry - let's eat", or "thirsty - let's drink". Some programs are more nuanced, they're opinions, ideas and habits. Person is a bundle of programs, nothing more.
This is also why I place meditation and mysticism before science. Scientists usually aren't even aware that they're a bundle of "programs" too. Who or what is it that is doing science if there is literally no one doing the thinking, deciding etc?? This question is what the sages throughout the ages have been teaching. I mentioned this in one of your previous comments... mysticism, going within, internal introspection, first knowing thyself THEN knowing the world.
Anyway, all that being said, when these "flocks" form, or when two people engage in an argument, it's their "programs" that argue, not them.