(3 of 3)
On November 24 we took a look at the log files for the last 12 days of polling and discovered a smoldering cybernetic battleground, in which two groups of robots, one favoring answer A and the other answer D, had duked it out on our servers down in the TIME-Life Building basement to see who could distort the results the most.
Using the IP addresses that our server log had recorded during the battle, we set out to track down the origin of this (every machine that connects to the Internet has a unique address, much like a telephone number, called an Internet Protocol Number or IP address) and discovered that many of the addresses were faked, made-up numbers that a savvy hacker had inserted to confuse our tracking devices. All of the IP addresses used by the A-voting, environmentally conscious machines were either fake or untrackable, but several chunks of robot activity favoring the D ("not concerned") response showed up in our server with this interesting return address: "gatekeeper.monsanto.com."
A spokeswoman for agri-giant Monsanto says that their records show no evidence of outgoing robotic activity from their servers during two of the time periods in question. In one case, they saw no traffic at all, and in another, they saw "traffic much too erratic to be the product of robotic activity, and most likely generated by individual employees who feel passionately about the issue."
Our robot tracking expert here at TIME.com, Andrew Arnold, has a different opinion of the vote patterns coming from the address registered to Monsanto: "Robot voting. No doubt about it at all." He then explained: "Anytime you see a pattern like this, with blocks of votes coming in very close together, from the same IP address, with the same user ID number, there is no doubt that they are generated by a machine." Can the IP address that led us back to the "gatekeeper.monsanto.com" be faked? "Yes, it is possible that a very sneaky hacker used a Monsanto IP address [called a "spoof"] and left a fake fingerprint behind on our server log. Given this fact, there is no way to know what really happened for sure."
Polls are all about reader input, and we try to allow as much leeway as we can to allow our users to express themselves. At the same time, we feel we can't become the victims of mischievousness or jokery in the vein of Mick Foley or Dustin the Turkey. We can take the criticism for the appearance of Hitler near the top of our poll — after all, he really did have a huge influence on the current era, and our poll is not a popularity contest — but it's a little disconcerting, as some of our staffers have observed, to have our users imputing wild biases, prejudice and evil intent to us. But that's part of the New Media scene. While it was certainly trying at times keeping our polls together in 1999, we can only hope that the year 2000 will be as interesting.