Bots are tools, designed by people and organizations to automate processes and enable them to do something technically, socially, politically, or economically.
Most of the bots that I have built have been in the pursuit of laziness. I have built bots to sit on my server to check to see if processes have died and to relaunch them, mostly to avoid trying to figure out why the process would die in the first place. I have also built bots under the guise of “art.” For example, I built a bot to crawl online communities to quantitatively assess the interactions.
I’ve also written some shoddy code, and my bots haven’t always worked as intended. While I never designed them to be malicious, a few poorly thought through keystrokes had unintended consequences. One rev of my process-checker bot missed the mark and kept launching new processes every 30 seconds until it brought the server down. And in some cases, it wasn’t the bot that was the problem, but my own stupid interpretation of the information I got back from the bot. For example, I got the great idea to link my social bot designed to assess the “temperature” of online communities up to a piece of hardware designed to produce heat. I didn’t think to cap my assessment of the communities and so when my bot stumbled upon a super vibrant space and offered back a quantitative measure intended to signal that the community was “hot,” another piece of my code interpreted this to mean: jack the temperature up the whole way. I was holding that hardware and burnt myself. Dumb. And totally, 100% my fault.
Most of the bots that I’ve written were slipshod, irrelevant, and little more than a nuisance. But, increasingly, huge systems rely on bots. Bots make search engines possible and, when connected to sensors, are often key to smart cities and other IoT instantiations. Bots shape the financial markets and play a role in helping people get information. Of course, not all bots are designed to be helpful to large institutions. Bots that spread worms, viruses, and spam are often capitalizing on the naivety of users. There are large networks of bots (“botnets”) that can be used to bring down systems (e.g., DDoS attacks). There are also pesky bots that mess with the ecosystem by increasing people’s Twitter follower counts, automating “likes” on Instagram, and create the appearance of natural interest even when there is none.
Identifying the value of these different kinds of bots requires a theory of power. We may want to think that search engines are good, while fake-like bots are bad, but both enable the designer of the bots to profit economically and socially.
Who gets to decide the value of a bot? The technically savvy builder of the bot? The people and organizations that encounter or are affected by the bot? Bots are being designed for all sorts of purposes, and most of them are mundane. But even mundane bots can have consequences.
In the early days of search engines, many website owners were outraged by search engine bots, or web crawlers. They had to pay for traffic, and web crawlers were not seen as legitimate or desired traffic. Plus, they visited every page and could easily bring down a web server through their intensive crawling. As a result, early developers came together and developed a proposal for web crawler politeness, including a mechanism known as the “robots exclusion standard” (or robots.txt), which allowed a website owner to dictate which web crawler could look at which page.
As systems get more complex, it’s hard for developers to come together and develop politeness policies for all bots out there. And it’s often hard for a system to discern between bots that are being helpful and bots that are a burden and not beneficial. After all, before Google was Google, people didn’t think that search engines could have much value.
Standards bodies are no longer groups of geeky friends hashing out protocols over pizza. They’re now structured processes involving all sorts of highly charged interests — they often feel more formal than the meeting of the United Nations. Given high-profile disagreements, it’s hard to imagine such bodies convening to regulate the mundane bots that are creating fake Twitter profiles and liking Instagram photos. As a result, most bots are simply seen as a nuisance. But how many gnats come together to make a wasp?
Bots are first and foremost technical systems, but they are derived from social values and exert power into social systems. How can we create the right social norms to regulate them? What do the norms look like in a highly networked ecosystem where many pieces of the pie are often glued together by digital duct tape?
(This was originally written for Points as part of a series on how to think about bots.)