Are Bots Allowed?
Bots are routinely discriminated against.
Don’t believe me?
Let’s imagine that a couple of years ago you had written something witty on somebody else’s wall on Facebook, and you were trying to remember who it was.
Turns out that Facebook makes this quite difficult. You can’t actually “search” your own wall posts.
Now Facebook has all of this data, but for whatever reason they’ve decided not to make it easily searchable. So now imagine for a moment if you were able to use a bot to do the searching. You could now instruct this bot to go through the very cumbersome steps that Facebook lays out for finding past wall posts and do it on your behalf. (This is something you could do with a uBot.) The other thing you could have done – if you’ve been using this bot all along it could have kept an own archive of wall posts in my a data store and you could simply instruct it to search your archive.
It may sound like a trivial example, but actually it’s very important. Bots change the power relationship between networks and their participants. They also invert the present legal situation.
Across the online world we systematically discriminate against bots. We deny them access to services and to websites that everyone else has access to. Right now, this bot I described is against the TOS of Facebook.
Why is that? Is it OK?
There are lots of laws at the moment that allow networks to restrict to what degree you can use a bot to interact with them. They basically can restrict you to only use the existing application programming interfaces or API and say: only these are legitimate, and on top of that, we can limit how much you can do. See why bots are so powerful?
Think about on demand car services companies like Lyft. If you drive today, they each have a separate app. This makes it very hard for you as a driver to participate in more than one network at a time. If you had the right to be represented by a bot, somebody could write a piece of software that drivers could run that would allow them to simultaneously participate in all of these marketplaces and the drivers could then set their own criteria for which rides they want to accept. Now clearly those criteria would include for instance what the commission rate is that the marketplace is charging, and drivers would go for market places that charge less of a commission.
So you can see in this example how the right to be represented by a bot is quite powerful – it would make it very hard for a company to charge too high a commission, because new networks could participate simultaneously in the larger Network, and it’s the very threat of the creation of these new networks that would substantially reduce the power of the existing ones.
There are bots that tweet out images from the world’s best art collections. There are playful bots that correct particular misspellings. Typically when we think about bots people think about the unsavory ones. But actually, bad bots aren’t the real reason bots are denied access to Web sites. It’s because they don’t spend money.
On Youtube, for example, advertisers don’t want to pay for a click from a bot. Bots will not buy products.
If the main reason companies deny access to bots is that they aren’t consumers, what if we empowered bots to become consumers? Could this change their access to parts of the online world where the rest of us hang out bots? Could it give bots rights?
Bots do very specialized things that they’re programmed to do. But they’re becoming more complex. Imagine bots as very simple creatures that work autonomously. Let’s imagine in the future that they have access to us as users, and like to interact with us. They’re very social by nature. And they’re designed to be autonomous. They’re designed to do their own thing.
Imagine a team of engineers and designers creating bots that have artificial intelligence. These bots are free agents that don’t belong to anyone. They’re not your personal digital assistant. These bots live to show off, just like any other user, and they’re even financially independent. They create value through their interactions with users and they use the revenue they generate to survive online. They use a couple cents a day to cover their server costs, to pay the kilowatts of power they use and storage they need online.
Let’s say you have a bot that is meant to find and collect information about the sea and volcanoes? Well, after it gets created, it searches the web, interacting with people, discovering sites and information around its favorite keywords, and through those discoveries it came up with this volcanic island off the coast of Sicily and it turns out to be a very interesting travel destination.
This bot gives you the results of its online discovery. It then tries to learn what’s valuable and what’s not, and it notices that you spent quite a bit of time looking at images of mountain climbers hiking up the volcano, and it might introduce you to a group of mountain climbers. Another bot hangs out with this team. The next time that bot might actually give you a content cluster around that and see if you’re interested.
Maybe there’s a bot that can offer services to companies and rate online content. It can deliver advertising for companies like Youtube. And so this bot earns a living and it covers its expenses. It can survive if it makes excess revenue beyond what it needs to pay its bills then maybe it can “reproduce” and make better versions of itself. If those versions are not better they won’t last very long, but if they’re actually improvements, then those new generations in turn will create more bots. These bots will evolve as a species and every new generation will be more interesting to use.
With these bots, the millions of them that are being created, each one of them has a unique identity. No two are alike. They have personality quirks or behavioral characteristics. They have a learning style. They acquire their own knowledge. In a world where bots are not ominous, they create a lot of value.
One of the best things that the Internet has brought us is cutting out the middleman. So in the old days you’d have to go to a travel agent to book a trip. Today you find a place on your own. Tomorrow with a little help from some bots, you won’t even have to look yourself.
So in order for bots to transition to this more mainstream model that I’ve been talking about a couple of things have to happen first.
How To Make Better Bots
- Bots will need unique ideas. That will allow them to be accountable for their activities online, across sites, across time. That bot demographic can be reported to advertisers separately so that advertisers can decide whether that’s interesting to them and how much they’re willing to pay for it.
- Bots will need a code of conduct which will be like a social contract with humans that their makers will have to respect if they’re interested in this mainstream model.
- We need smarter security on websites. Smarter security will be able to protect us better but it will also prevent us from cutting out a whole chunk of business, and I think our attitudes towards bots will start to change. I think we’ll see them less like mechanical humans and this is important because bots don’t make very good humans – they suck pretty badly at it.
- We need to stop holding bots to our likeness all the time. It’s pretty reasonable to define them in terms of what they do for us, but pretty soon this paradigm shift creates a new reality. We will get the feeling that we get when we treat animals inhumanely. And so when we meet those ultimate non humans for the first time, it will be almost as if we’re creating artificial life and creating thinking machines.
Overall, bots have the ability to completely restructure how we interact with services, and what we can use them for. This makes them incredibly important.