In Over Your Head

Cop Bots vs Robber Bots

A simple metaphor for an important phenomenon. I’ll explain.

Cop bots are the enforcers. Google is an organizing algorithm but, more importantly, it’s also an exclusion robot. It says “you’re in,” and “you’re out.” It has to be very good at this, or it makes no money, and the robot gets shut down.

Spammers are infamous for sending millions of emails. These are robber bots. They find new ways around systems and exploit loopholes in the cop bots to give profit to their masters.

Both the cop and robber bots are massively leveraged. Both of them work extremely fast, but there’s one element that’s missing: humans.

Humans are currently sophisticated enough to detect most robber bots. We know when we’re on a splog instead of a real blog, and we know when a spam comment is real or not. But if you have a blog, especially one that gets a fair amount of traffic, you’ll notice it’s taking you longer than before to see what comments are real. Robber bots are getting smarter.

As time goes on, robber bots will get better and better at confusing us, just the same way game bots are getting better than humans at chess and Jeopardy. The cop bots accelerate too, but they need us to triage the grey areas, which is why there are “moderated” comments and CAPTCHAs that require human intervention.

This means humans will have to spend more and more time in the grey area, detecting robber bots. In other words, the robber robots are accelerating. Humans are not.

This is exacerbated by the problem that more and more existing information is going online and becoming spammable, where detection is more difficult due to restriction in trust signals (ie humans can detect each other easier in person).

I hypothesize that the inevitable endgame to this is a 100% non-anonymous internet, which has already begun with Google Accounts, Facebook, and Verified Twitter accounts. I’m not sure I like this idea, but I have a feeling that there is no way to avoid it, because it is the only way to ensure that someone is human, thus giving us our time back (especially since content creators are often moderators, too).

It is highly possible that there are gaps in my logic. If so, please poke holes in them, I’d be more than happy about it.

Posted

in

,

by

Joe Gannon

Comments

9 responses to “Cop Bots vs Robber Bots”

  1. aaron wall Avatar
    aaron wall

    The other big issue … is the CIA dealing drugs (aka: AdSense $ to the content mills).

    In the past Google had less leniency with bulk automated to semi-automated spam efforts. But now that Microsoft has picked up the tab for any anti-trust complaints against Google, it is much harder for them to do things by hand.

    My guess is that more stuff comes online from vetted editorial sources that kills the margins of many spam-based business models.

    When books are online as a part of the web (which can be linked to and wrapped in ads) that changes a lot of the dynamics. And if Google figures out a way to make paid micro-subscriptions work that also changes the dynamics of search.

    Reply
  2. Michael Bigger Avatar
    Michael Bigger

    I have no holes to poke. I am just meditating on the value of algorithms after reading this post. I run a trading algorithm and it is very good. Trading bots are also getting smarter! It is a wonderful thing to own one and it has never been cheaper to build one.

    I am digressing…

    Reply
  3. Ed Avatar
    Ed

    1)Deputize the trustworthy.
    2)More, faster, publicized, vicious prosecutions.

    Reply
  4. Marjorie Avatar
    Marjorie

    Even if web goes completely “non-anonymous,” I think there will be robber people, if not robber bots. There is no way of knowing if this person posting here is really Marjorie. There is no way of knowing for sure whether a Twitter account truly belongs to a celebrity. Maybe we’ll have to register to get an online number along with our Social Security Number sometime soon.

    Thought provoking post. Love it! 🙂

    Reply
  5. Jeffrey Tang Avatar
    Jeffrey Tang

    I question the need for a 100% non-anonymous internet. I agree that connecting online and offline identities to some extent will be necessary for combating robber bots and the like.

    That being said, even in a world where online identity is managed through one (or a few) centralized, verified sources, individual communities and networks may opt to allow anonymous participation (at least at the user-to-user level). For example: I use my verified Twitter account to create a profile on Interesting Forum X under the handle “UserBot”. To other users, I will be known only by that handle, allowing me to keep my “real” offline identity hidden. Should I violate the rules of Interesting Forum X, the administrators have the choice to ban or not ban me, but that will be a decision made by a specific community, not the internet as a whole.

    So instead of heading towards a 100% non-anonymous internet, it might be more accurate to say that we’re creating a non-anonymous back end to the internet, which will be visible and usable to administrators and other enforcers, but minimally disruptive to communities which choose to allow users a measure of anonymity.

    Or maybe it’s just semantics 🙂

    Reply
  6. Duff Avatar
    Duff

    Non-anonymous email looks like a good way to prevent spam, and will probably be proposed at some critical future point or else spam will completely overtake legit emails.

    This will of course increase the payoff for identity theft, which will perhaps then lead to genetic verification of your non-anonymous internet ID, which will lead to faking/stealing genetic information, and so on….

    Reply
  7. Judy Helfand Avatar
    Judy Helfand

    The Robber Bots are really busy today on http://www.chrisbrogan.com! I like to read the comments, so I subscribe to them. But there are days like today where the robbers’ clever ways of writing the comments allows them to slip through Chris’ filters. (where is that personal secretary when you need one!)
    I think I have received in excess of 60 spam comments from Chris’ blog this morning. And I really wish his last name had nothing to do with SHOES!
    Have a good weekend. I think right now I am going decelerate and dust the real furniture.

    Reply
  8. Sophie Davis Avatar
    Sophie Davis

    I think we underestimate humans’ ability to detect a bunch of bs. Robber bots might get some cash in the short run, but can’t establish long lasting relationship (so valuable in social media) because they are not human, therefore not authentic, not transparent.

    I totally agree with your vision that the net is going to be 100% non-anonymous. It already is in someway. It’s getting harder and harder to “hide” behind a user name. But whether it gets to that point doesn’t really matter. There will always be people trying to take advantage of the system. Always. We just have to be confident that we will be able to detect the crap!

    Reply
  9. Ryan G Avatar
    Ryan G

    You bring up a point that ultimately could lead to the eradication of email, which has become so archaic in a way. Sites like LinkedIn and Facebook help give us the filters we need with information to which you alluded recently and will only improve, but like @Sophie suggests people will “always” try to take advantage of the system.

    One could argue that it’s the untrained Internet user that enables spam and robbers to exist and only with training and education can we battle this. These are people that actually are fooled by the robots, which gives them life beyond their time.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *