2009 Post-Mortem Chat Transcript (page 2)
18:12:49 facilitator What would you do differently?
18:13:29 BreakingCodes ( Observation: The "perfect game" would have had no competition (other than the game itself). Everyone would submit their findings (sightings, confirmations, etc.) to a well-staffed central site. We'd probably each have earned a few cents, which is better than what we actually got. ;) Everyone wins. [Done.]
18:15:07 Airigami/Kelly I would have attempted to group the smaller teams together sooner, might have been able to get them to "give up" their info if we had reached out to other groups sooner. Also we got started with it about a month ago, but didn't get back into it till the final week before really.
18:15:59 Airigami/Kelly It's interesting because I think the answers would have been found sooner (if that was the goal) if there wasn't an every-man for himself feel to it
18:16:20 Airigami/Kelly but I don't know how you motiviate folks to start looking without some prize... (done)
18:17:52 John/Red40k.com I don't think my site looked as nice as most of the other sites.
18:18:33 John/Red40k.com I probably would have spent more time making it look nice, and less on the back end submission process.
18:19:17 John/Red40k.com I far overestimated the number of responses I got. and it may have been because of my ugly site... not sure.
18:20:16 rchern (DN) This is more on a global level, but I think the challenge could have gotten a lot more exposure.
18:20:30 rchern (DN) I didn't learn about it until Friday, and nobody I talked to had heard about it.
18:21:05 johngreen DeciNena, on what I'd do differently: I would have asked for firmer confirmation from my eyewitness sources.
18:21:18 rchern (DN) Surely there is a point in the middle between that and a local news station putting an article about a balloon in their town on their website. ;)
18:21:53 cwho(DN) Assuming the goal is to build a very big network quickly: 1. monetize the referral process 2. provide a way to provide anonymous list sumbittals for recruting. In other words, I spent a lot of time going through my contacts list to see who to send this too. However, I would have been really happy to DUMP my 4000+ Thuderbird contacts list at a server who would have sent out mailings like "you name has been added to a list of people who might be interested. Then if I sent some ex-es or customers who haven't paid etc. who cares. 3. have some sort of competition for recuritment. 4. provide more "canned" text for emails or "add 3 friends and click to email" tools like some of the policial groups.
18:23:45 lmcelhiney (DN) "Prisoner's Dilemma" must enter into this at some point... A number of teams working together could probably beat MIT handily. Maybe that would be part of the "trading" to "buy" your way into a larger team for critical mass.
18:24:05 nerdfighters We had 9 out of 10 also. We were closer to 10 than we knew, but we were convinced of a false report that kept us from winning.
18:24:30 lmcelhiney (DN) I realize that this sort of steps on "merging teams," but I was impatient! [Done]
18:25:31 Airigami/Cupcak I agree with rchern. I would have liked to see the word get out about the hunt sooner, bigger, more globally. I think what I would have liked to do differently is actually been more on the task of promoting the event with the media/public and let others do more of the organizing of selected data. Of course, it was an interesting problem we faced. If I did more in the public to promote us, I was also promoting to the world that they could join another team that was offering money.
18:26:50 DeciNena (DN) As cwho said, we could have done a better job of forcing the growth of our network.
18:28:24 10balloonies Just curious to know when people became aware of the challenge, esp. MIT. We heard late last Tuesday and scrambled to act.
18:29:00 10balloonies I actually liked the challenge that way, to be honest. If it was a real-world public good sort of thing you wouldn't get much notice, if none at all.
18:29:14 10balloonies a.k.a Amber Alert.
18:29:21 Airigami/Kelly article in Huffington Post about a month ago believe it or not (done)
18:29:28 nerdfighters We heard on Wednesday and announced it to our viewers on Thursday. (done)
18:29:53 10balloonies If an article mentioning geocachers didn't pop up in my alerts, I would have missed it entirely.
18:30:12 John/Red40k.com I heard a month ago on slashdot/huffington post.[done]
18:30:23 hackerfactor Amber Alert? That would have been funny! I could have hacked the highway sighns to say "If you see a red balloon, call ###-####."
18:31:02 10balloonies I was leaning towards the white hat solutions, but kay.
18:31:49 Airigami/Kelly Do you think that the DARPA guys met their goals/created an appropriate experiment?
18:31:59 hackerfactor I'm certain of it.
18:32:28 hackerfactor They expected this to take a week. It took less than a day.
18:32:38 TheMike It was too easy. Ended too fast.
18:32:48 nerdfighters They thought it would take a week? I thought it would take an hour.
18:32:53 lmcelhiney (DN) I believe that one of the critical factors for DARPA was a "short notice" event so that they could see if such an "alternate" system could actually work. Remember the real reason for the creation of the Internet (other than Al Gore) was to have an alternate communication system "just in case..."
18:33:13 hackerfactor The goal was to see if a civilian militia could be organized over a social media network. It succeeded.
18:33:28 hackerfactor (D in DARPA is for Defense)
18:33:36 cwho(DN) NO it was more of a sunt I think. It is not a true experiment. There are no controls and no way to really understand what happened.
18:34:00 rchern (DN) I think it was good for an initial experiment. But, in the case of an Amber Alert or Defense, more exposure in less time would be more useful.
18:34:36 lmcelhiney (DN) Like Engineering 101: You get a pile of tools and have to figure out how they work and what to do with them!
18:34:49 hackerfactor True -- this gives them ideas about the next steps. This was a first-run test case "just to see" and check feasability.
18:34:58 10balloonies I agree. Maybe a contest where the start time isn't announced or the rules. You just sign up for an alert list and they notify you of the contest.
18:35:22 lmcelhiney (DN) "Andomeda Strain"...
18:35:22 DeciNena (DN) I think they will have learned a LOT after this is done. Like, how it SHOULD have been done. And they will interview us about how we responded and what tools and techniques we used.
18:35:28 Airigami/Kelly See now I thought the experiment was more to see how information is shared online/so they can discover things that are going on "first" with the intelligence of the masses... (done)
18:35:31 cwho(DN) But all of this is well known. Look at the recent political organizations. Nothing we did is new to them
18:36:06 hackerfactor Also, there are other networks besides the Internet. This could give big clues about how to monitor and operate.
18:36:31 rchern (DN) I think it was both Kelly. Using networks to get people to do something, and also using networks to see what is being done.
18:36:36 lmcelhiney (DN) ...ah yes, but the political issue had "years" worth of motivation waiting for an outlet.
18:36:38 cwho(DN) Move on.org, OFA etc does this kind of thing and actually gets people to hold events and let people in their doors on nothing more than a shared goal and a website sign up
18:37:15 nerdfighters I think what we did is unique in this respect: We had to sort through more fake information than (say) moveon.org does in order to organize events.
18:37:24 hackerfactor And don't forget counterintelligence. If they were monitoring the teams (and you know they were), then they could see how to better poison the stream.
18:37:37 10balloonies They should do another challenge with no money - only bragging rights.
18:37:38 rchern (DN) An unknown start time is interesting, but I think participation would drop significantly during the work week.
18:37:41 nerdfighters One of the valuable things about this experiment was learning who can differentiate good information from bad information, and how quickly they can do it.
18:38:11 facilitator That leads us to another question. How many teams planted disinformation or fell for it? Say answer if you would like to answer.
18:38:17 10balloonies I think our group was probably the most closed since we already had a well established network.
18:38:19 cwho(DN) information sorting - oK really good point - I stand corrected
18:38:38 BreakingCodes ( I As for why DARPA didn't promote this event more, I think they wanted to see if online communities would do the advertising for them. Whether the cyberworld actually did that or not is irrelevant to the success of the experiment. DARPA now has data.
18:38:43 John/Red40k.com I think it is interesting that DARPA's resource page went away.
18:39:09 10balloonies We had geocachers in several midwest locations sending reports back on faulty data.
18:39:29 10balloonies Many are upset they were largely ignored for the contest.
18:39:31 nerdfighters We planted tens of thousands of pieces is disinformation, because we figured our basic advantage was brute numbers. (Our community has tens of thousands of people, but most of them are young and don't have the ability to drive)....
18:40:35 10balloonies Saturday is optimal for geocachers since they were spotting them while geocaching. It is an older demographic (Mid-thirties) with kids.
18:40:53 facilitator Does anyone want to take credit for the fake #8 in MI that most of us fell for?
18:41:09 rchern (DN) (side note) 10balloonies, Being in Saint Louis, I also noted a severe lack of balloons in the midwest when I looked at the final map.
18:41:13 nerdfighters Seriously, does anyone? Because MIT denied it to me.
18:41:56 DeciNena (DN) It wasn't DeciNena's. ;) Our was in Ohio, but nobody ever saw it as far as we know.
18:42:07 hackerfactor Can someone send me a link to #8? Perhaps I can pull something out of it...
18:42:23 10balloonies We had the correct #8 at 11:26am which was our 5th confirmed find. It was found by two trusted geocachers.
18:42:45 10balloonies 11:26am Pacific
18:43:27 nerdfighters 10balloonies: Wow, which ones did you have trouble with?
18:44:17 10balloonies #5 and #6 were the trouble ones. We submitted the correct info for #10 even though we had two geocachers say it wasn't there. So there was some confusion with that find.
18:44:55 DeciNena (DN) Brian submitted #5 to DeciNena in person.
18:44:59 BreakingCodes ( I was surprised that so few of the small teams seemed to be looking for opportunities to merge. I'm sure each had their own reasons (pride, greed, loyalty to their own joiners, lack of vision, realizing the benefits too late, inability to "do the math," etc.). Does anyone see a way that an individual or a small team could succeed at something like this? I don't.[Done.]
18:46:12 DeciNena (DN) DeciNena actively and strongly campaigned for mergers even before the start. We felt our prize sharing scheme was generally compatible with most other prize-sharing teams who weren't just talking about using it all for one purpose.
18:46:13 10balloonies No. Small teams would lose out, I think. There are well established networks that have some huge advantages.
18:46:21 Airigami/Kelly Who else besides geocachers/us had a pre-existing network of over 1k members, and what was the affiliation/membership numbers
18:46:42 DeciNena (DN) But, none of the smaller teams wanted to merge. I think there's a lot of pride involved, but clearly team size is a major advantage.
18:46:48 10balloonies We have over a million members but we didn't have enough time to disseminate information. Our email went out late Friday.
18:46:49 Iarch(DN) Did MIT cheat? not that it really matters now
18:47:01 nerdfighters Us. 90,000. And is there such a thing as cheating?
18:47:31 facilitator Nerdfighters, what is your organization, exactly?
18:47:31 Airigami/Kelly affiliation? I hadn't hear of you til yesterday- I'm probably out of the demographic
18:47:32 Iarch(DN) Like Kirk in the Kobayashi Maru
18:47:32 hackerfactor Iarch: How can you cheat in an open social experiment?
18:47:41 Airigami/Kelly lol
18:47:51 Iarch(DN) insiders
18:48:22 nerdfighters Um, my brother and I make popular videoblogs on youtube. Those vlogs have spawned a tight-knit community of social activists who call themselves nerdfighters because they fight for nerd rights.
18:48:33 Iarch(DN) I just finised the new version of the day the earth stood still
18:48:40 10balloonies My guess, based on no data, is that some geocachers submitted the coordinates they sent to me to MIT for the less altruistic benefit.
18:48:47 John/Red40k.com I was hoping that some DARPA personel or their close friends might be tempted...
18:48:48 nerdfighters WE did this in an attempt to raise money for a yearly thing we do called the project for awesome, in which charitable causes take over YouTube for one day.
18:48:55 Iarch(DN) and I was thinking we could have identitfied all the spheres in a day or 2
18:49:42 facilitator Ok, question from the queue from hackerfactor: Go.
18:49:54 Iarch(DN) I know the military has advanced spy satelites
18:50:01 hackerfactor Can we archive the set of known photos -- both fake and real? I'd like to do an image analysis on the entire thing later.
18:50:12 Iarch(DN) MIT could have hacked into the DOD mainframe
18:50:34 facilitator hackerfactor: We'll be posting all of the evidence we had on our site when we have time.
18:51:24 Iarch(DN) Does anyone know the total # of participants?
18:51:52 BreakingCodes ( I think "cheating" in a game like this is something that should be encouraged, in the interest of "thinking outside the box" and coming up with imaginative solutions. That would include stuff like false reporting, floating phony balloons, and so on. All part of the fun. [Done.]
18:52:14 Airigami/Kelly So what did y'all think of the giant cupcake?
18:52:40 Iarch(DN) Do you have a twit pic?
18:52:52 John/Red40k.com I still think it's a good idea...
18:53:37 Airigami/Kelly http://www.youtube.com/watch?v=5OAaiKuT97E
18:53:51 Airigami/Cupcak to tag on to Kelly, I'd like to propose that if another challenge like this comes up (DARPA or otherwise) we should work with one of your other teams. We're a force to be reckoned with in the PR world, but could use the techie folks doing the numbers stuff.
18:53:56 DeciNena (DN) Airigami, I think we got recruits by declaring we were 100% cupcake free. You kind of made yourself a target for ridicule. ;)
18:54:19 Airigami/Cupcak that was intentional. :-)
18:54:26 DeciNena (DN) Nicely done, then.
18:54:29 nerdfighters Kelly, thanks for making that a response to our video.
18:54:42 Airigami/Kelly np :-)
18:55:02 DeciNena (DN) Yes, we felt team merging should have been a much bigger force. We got one small group to join us, but no major declared team merged with us. We could have beat MIT if one or possibly two had joined us.
18:55:02 Airigami/Kelly thanks for defending the nerds!
18:55:59 nerdfighters We wanted to merge with a team, but the deal we'd made with our team prevented us from doing so. (100% to charity.)
18:56:34 DeciNena (DN) Yeah, I understand that. We left our terms open, not saying we would be sharing all 40k.
18:56:58 Airigami/Cupcak we didn't think we needed to merge, but the comment came up before about people being busy during the work week. Our work week is Sat-Sun. so my team was a in a bind.
18:57:00 Iarch(DN) I wasn't doing it for the $$$
18:57:02 DeciNena (DN) We wrote our prize plan in terms of "portions of the prize we received" so that if we had to split a prize with another team due to a merge, we still were in the right.
