I ask because it seems like it would be really bad but I don’t see much talk about it. I do, however, see talk about anonymity being banned from the internet and it being a bad idea. Would a social media site or other such sites be rid of a bot problem by making users ID themselves?
In: 15
It’s a bit of a mixed bag.
Internet traffic is handled by requests. You ask some server for information and it decides to either give it to you or not.
When you’re on a web page, you might make a new request every few seconds/minutes through normal scrolling/viewing/reading.
A simple bot can make thousands of requests per second.
And that’s not always in a nefarious sense.
Automoderators will make a request (maybe multiple) for every single comment made on reddit.
So if you think about it like that, then every comment you make has a bot reading it. That makes half of all requests for comments bot requests.
Then you have marketing agencies making API requests en mass.
And then normal maintenance/debugging which has automated checks.
So while half of all internet traffic is bots, it doesn’t mean that they’re all repost/astroturfing bots.
Automated requests are one of the cornerstones of keeping the internet functioning, and since computers can process many times faster than us, relatively few of them can make the same amount of traffic is hundreds of real users.
There are many different kinds of bots, some nefarious, some helpful. Search engines wouldn’t work without bots regularly combing all the catacombs of the internet. Other bots are relatively harmless, such as automated responses here on Reddit. But bots are also used in information warfare, including here. It’s hard to get an accurate idea of how big or small that is, partly because governments and other agencies try to hide what they’re doing.
The one study that came to mind is that Australian researchers found that in the first month of the Ukraine war, 90% of tweets on the war were done by bots. I try to be aware and it’s so hard to know just how much that affected me.
I would say that harmful/manipulative bots can be a big deal and we should be concerned about them.
So there’s a few facets to this.
The term internet traffic is important here. It is not saying 50% of what you see on the internet is from bots. It’s that 50% of internet consumption is from bots.
Web scrapers, research spiders, crawlers (a la Google or Bing bots), engagement bots, social media astroturfing, and good old fashioned fraud, would all fall under the umbrella of bot behavior.
Some of these usecases are nefarious, most are harmless, and many fall somewhere in-between as a bit sketchy but ultimately “fine”.
One thing to keep in mind is just because it’s a bot, does not mean it’s bad. In line with that, certain people want certain types of bots, and other people don’t want those same bots for different reasons.
Let’s run through a scenario here, an extremely common scenario in the world of online advertising.
——-
Someone has a click bot, it visits websites and clicks on things to see where they go. This particular bot, for one reason or another, does not identify itself (to get a more “legitimate” experience), but also does not discriminate on links, so it will click on ads.
This bot visits YouTube, and clicks on an ad in the sidebar. It redirects out, making it through some basic filtering from the ad unit, and lands on the advertisers page, let’s say it’s Ford.
Now, let’s break down who wants that not to not exist.
Very clearly, Ford doesn’t want that bot. Bots don’t buy trucks. However, does Ford know they were a bot? Google sold it to them, telling them that it wasn’t. Google’s job is supposed to be to catch these things, that’s why you advertise on Google. The other reason you advertise on Google is they have the largest reach, but guess what number is included in that reach.
And hereinlies the problem. We have many people with mixed motivations.
So, we have bots everywhere.
Edit* crap, this was on eli5. Uh, sorry
The answer is yes there’s more bot traffic than people. But a bot can do more than random posts on social media. Bots are used to automate tasks like fetching log files or automatically deleting disk space (over simplification). Bots can also check multiple websites for cheap airline tickets and provide those results to the requester in close to real time. Pretty amazing when you think about it.
What’s crazy is the amount of energy we need for these tasks. We’re basically fueling this…thing…we created to run tasks that requires stupid amounts of power just to find cheap pizza in the area.
It’s probably more than 50%, but that doesn’t necessarily make it a big deal. *Someone*, some real people, set those bots out to do something. Just because they’re having a bot do the web scraping, or posting content across all their social media sites at once or whatever instead of doing it all manually doesn’t mean there’s not a human initiating it and a human wanting it done.
Latest Answers