While 3D printers are becoming smaller and more affordable/accessible (MakerBot, RepRap, Mantis, etc), the software we use to build the virtual models still assumes that the user is a professional trying to design a satellite. Within the Media Lab, we’ve tackled a similar problem of computer programming with a project called Scratch. And so the question became: Is there a way to take CAD software and make it easy and fun to use? Where hobbyists, kids, makers, and anyone else interested in 3D Printing could dive in without having to first learn SolidWorks or AutoCAD.
Simultaneously, I was watching virtual creation after virtual creation being built in Minecraft. Then it occurred to me, Notch (the creator of Minecraft) hadn’t just built a game. He’d tricked 10 million people into learning to use a basic CAD program. Sure, no one would be using Minecraft to design a rocket engine (though, now that I’ve said that, someone will certainly try to prove me wrong). However, the goal was never to replace professional grade CAD software, but to drastically lower the barrier to entry for 3D modeling and printing. I’d say we succeeded.
After recruiting Jason, from the Camera Culture group, it took an afternoon to code up the initial iteration. After a few weeks we felt it was ready to be released into the wild. The response was incredible…
In early January, a number of large QR code advertisements started popping up around MIT/Cambridge. So, of course, I did what any geek would do: pulled out my phone and tapped Google Goggles. The QR code linked to ohaipuzzle.com. Unable to ignore a good puzzle, I spent the afternoon working with several friends (Geoffry, Mike, and Nathan), scattered across the country, to solve the puzzle (we came in 3rd). Now that the puzzle has reached its EOL I’ve put together a little walk through. If you want to take a crack at it first, I’d stop reading about now.
Warning and General Info
This is a puzzle intended for people with a decent amount of computer science knowledge. You’ve been warned! Specifically, this is a recruiting puzzle by a cool company called ITA Software. You may have heard something about Google’s attempts to buy them (usually a pretty good indicator of awesome levels). Speaking of Google, they actually did something similar back in 2004.
Spoilers below!
Phase One
If you look through the source you’ll see there’s nothing abnormal. The only interesting bit is the Google Analytics code at the bottom, but it’s standard. To tackle this section you’ll need to inspect the HTTP response headers. If you’re using Chrome you can do this by opening up your Developer Tools.
Specifically, you’re going to want to look at the X-Encode value. Seems odd, doesn’t it? Well, ctrl v it into your favorite base64 decoder to reveal a link to phase two.
Example X-Encode: aHR0cDovL3d3dy5vaGFpcHV6emxlLmNvbS8/NjB lMTg3ODgtMjgwZC00NmM3LTliMjYtZmNhYTZkMjU0MjMz
Example X-Decode: http://www.ohaipuzzle.com/?60e18788-280d-46c7-9b26-fcaa6d254233
Phase Two
And now the hard part. Let’s tackle the letter block first. 26 by 11 letters. 26. Letters in the alphabet, alright. 11. Why 11? What else is 11? FIND THE WORD. Ahhhh, alright. Let’s take the Fth (6th) letter of the first row, the Ith letter of the second, etc. There’s a couple different ways you can solve this. You can either do it manually (just make sure your count is correct) or you can write a script to solve it if you really want.
In this particular case we get the following: ANCHORCHIPS. What you get will vary as the letter block is different per person.
So, now we have the WORD. Next up, the UUID. By now you may have noticed the formatting of the url, handily providing our UUID. Our’s is 60e18788-280d-46c7-9b26-fcaa6d254233.
822-6. Honestly this one took quite a while to figure out. The key is in realizing that the formatting suggests a section of some document. Specifically RFC822: Standard for ARPA Internet Text Messages. Section 6 just happens to be Address Specification (specifically: email addresses).
With this in hand we now know that we need to give OHaiPuzzle the WORD we solved for, the UUID from our url, and our email address. The key for doing that is in the last part of the clue. 2616-9.5. Running with the way we solved the last clue, we discover that 2616-9.5 is the section for POST.
All we need to do is fire off a POST request. Excellent.
If you formatted everything correctly then you should get a response back congratulating you on completing the puzzle, giving you your place, and then revealing that the puzzle was sponsored by ITA Software. Then, a couple weeks later, a package might appear at your doorstep.
There you go, everything you need to know about the O Hai Puzzle. Twas fun, wouldn’t you say?
Current Stats for Solving Team
30,000+ people tried to complete ohaipuzzle.
152 completed ohaipuzzle
We placed 3rd
Dropbox then ran a dropquest
140k people completed the dropquest
We placed in the top 100
Alright! I’m officially caught back up on a sleep and life. For those of you severely confused, allow me to explain…
Every year MIT holds an event known as the TechFair. It’s pretty much like a Career Fair, except instead of everyone dressing up in suits and passing out resumes; everyone dresses up in jeans, has really awesome conversations about the latest and greatest tech in the field, and occasionally a resume is passed. This year, however, Facebook decided to sponsor an all-night hack-a-thon the night before the TechFair. The idea of a hack-a-thon is simple: bring an idea (hardware or software) that you’ve never worked on before, grab a team (up to 4), and then build it. You have from 10pm to 8am to implement your idea. In the morning, you will be given the chance to demo your idea and Facebook Engineers will judge the product based on your idea and execution.
It could be theorized that Facebook held the event to a) evaluate potential hires and b) ensure that the 100+ Course 6 (CS) people that participated were too exhausted the next day to make sense when attempting to talk with other companies. However, my thinking went something like this: “an excuse to pursue one of the many project ideas that fills the moleskine on my desk? Sold!”
I spent some time going through my ideas, and decided on a more recent one. It had a simple premise, addressed a real annoyance in my life, and seemed completely feasible to finish a set of core deliverables in one night.
Let’s set the stage…
The Pitch: Social Networking sites (Facebook, Twitter, etc) seem to have taken on the stance that your life is now, and now only. This real time approach is very much their core mission, but they’re pursuing it at the expense of our past. And by past I could mean 5 years or 5 hours. Twitter let’s us search the world, but not our own social graph. Facebook does, but only within the immediate past. Echo Chamber is a project to bring true search to your own social graph. Find the cool video you saw in your news feed last week or the tweet to a WIRED article from 2 years ago. Why exactly isn’t this already implemented? Now, take it a step further, visualize your social graph’s interest in certain topics over time. Curious what your friends thought of Inception? Well, now you can easily find out.
I really have no idea why this functionality doesn’t already exist, beyond hypothetical technical issues. Many business applications for monitoring a company’s brand through various social media channels have similar such tools, but none of them target the problem for the normal user.
So, that was the challenge. Build a web app that would allow a user to easily search their own social graph. I decided to go with Twitter for phase 0 as I had a decent amount of familiarity with their API, and focused on the search component, leaving the data visualization piece for later. I also decided to go as a one-man team, to add an additional level of challenge.
The Technical: Python + Django + Elixir/sqlalchemy to interface with SQL + Tweepy to interface with Twitter.
The Results:
Core deliverables: met! It definitely needs a good amount of work before it can be released in the wild, but I have to say that it came together amazingly well, given I hadn’t even installed/setup my server environment before arriving.
Out of the 100+ original hackers, 17 teams completed and submitted their projects. After the judging I managed to walk away with two awards: Runner Up and the Highlander Award (for being the only one-man team to finish).
The Fuel: Two cans of coke, a vitamin water, good music, and 2 twenty-minute naps were all that was needed to make it to morning and on to 10pm the next night. Why so late? Well, as a rule of thumb, never say no when Google offers to take you out to dinner…
Lessons: Next year, I will definitely be going with a good team. It was a blast going solo, but it was also a massive crunch to complete each of the core components necessary for the demo. With a team we could have gone beyond the core, adding features and polish – positioning Echo Chamber in direct competition with the first place team of three.
But hey, it was a lot of fun, and that’s what matters.
Before driving from Missouri to MIT I decided it would be an interesting experiment to take a time lapse of the drive itself – three days and 1400ish miles. The remarkable thing (to me) is that I was able to do it all with my phone (Droid X). The video quality is great, and the X did just fine managing that + calls + podcasts + navigation.
My original plan was to stitch the three days worth of videos together, and edit them a little. However, the time to do so has not presented itself since starting classes, research, and the like. So, for now, I will simply post the raw footage.
On August 18th, Facebook revealed Places – their attempt at a location based service. Similar to Foursquare and Gowalla it would allow you to check-in to locations (restaurants, a concert, or the top of a mountain if you have service). During the press conference Facebook VP Chris Cox talked about the future of Facebook Places, specifically the idea of a shared memory that you, your friends, or even your children could explore later.
I’m not going to delve into the privacy or creepiness concerns right now. I’ll leave that for its own post, someday. However, I wanted to point out that Facebook appears to already be moving in the hypothetical direction Cox spoke about. It would seem that Facebook Memories, are here.
Just under a month ago Verizon released a short 16 second ad on YouTube. It wasn’t long before someone noticed “@DroidLanding” reflected backwards in the video and discovered twitter.com/droidlanding – the start of what looked to become a nation-wide scavenger hunt. Solve the clues, get one of 21 free DroidX’s before it officially launched on July 15th. Now that it’s said and done I thought I would take a moment and examine what we can learn from the DroidLanding “game”.
First, some disclosures and background.
Christena with Designate #7
I love large scale games of this nature – playing and especially planning/executing. Running Humans vs Zombies over the past three years has taught me many lessons in game design, implementation, and flexibility. Basically, you don’t run a 300+ person week long game without learning a little along the way.
With @DroidLanding I quickly discovered that there would not be any locations in my vicinity. Thus, I spent my time helping others on www.androidforums.com and the #DroidXHunt IRC channel. For four of the locations (LA, Dallas, Seattle, and St. Louis) I ran a conference call where we helped four players navigate to and search for the prize. In LA and St. Louis we managed consolation gift certificates and in Seattle we succeeded in helping Christena and Conor become the proud owners of a new Droid X. Not bad for a game with over 20,000 participants.
Alright, so on to the analysis!
It’s alright to be subtle. Did @DroidLanding start with a huge marketing blitz to beat people over the head with the promotion? No. They had a simple, well placed, clue that acted as the rabbit hole for any potential Alice to find. When people stumbled upon it they told their friends and posted about it online. It was engaging because it wasn’t obvious. It spread quickly via word of mouth because Verizon wasn’t hammering everyone over the head with it. Maybe most importantly though, it targeted a very small group – those that actually cared. They were the ones that had been following the Droid X for weeks, if not months, before it had been officially announced. The ones that analyzed blurry, leaked photos and hypothetical spec sheets. To take a term from Malcolm Gladwell’s book, The Tipping Point, they were the Mavens.
Though small in number, they were the perfect audience to target because they will end up pushing many times more sales than the campaign cost to begin with. Speaking of which…
It’s alright to be cheap. While I don’t know how much the total budget for @DroidLanding was I feel safe in saying it was a drop in the bucket compared to the massive marketing role-out currently underway across traditional media platforms. Sure, their version of cheap still involved flying Verizon employees around the nation handing out 21 free phones, but everything is relative. In fact, not having a budget can help you make smarter decisions. The instinct is always to use the budget you have. Often, the best solutions are free or cheap. Why take the time and resources to build a full featured website where players can receive clues when you can use Twitter for free? When you have the money, the temptation becomes to go down the more complicated path because, clearly, complicated solutions = better solutions.
In Humans vs Zombies our budget consisted of whatever the moderators were willing to put into the game. As much as I would love to have a full budget, courtesy of a wealthy sponsor (hint hint), the lack of a real budget resulted in incredibly creative use cases for the resources we did have access to. Long story short, think of simple, cheap solutions before throwing money at it.
Your players will examine everything. Reward them for their engagement. You know that guy who insists that you need to simplify everything so that your audience/players/adoring public “gets it”? Go ahead and kick him out of the room now. Well, maybe just take what he says with a grain of salt. Sure, don’t make things convoluted, but there is no reason to dumb down your story/puzzles/explanations solely because they are complex.
But where is that balance?
Simple on the surface yet full of details for those that take the effort to look. Some of your players will care, others will be happy to just skim along, content that they know the basics (which is all they need to know). The one exception to this, in my mind, is the ruleset. Make it as simple, intuitive, as possible. When engaged, the last thing you want to do is pull up the rulebook.
Verizon half succeeded and half failed in this instance. The rules were incredible simple. Solve the clues, find the Droid X, keep the Droid X. Yet, when players attempted to dive down and analyze the clues there was nothing there – no meat. In fact, as it turned out they did something worse, violating our final lesson(s)…
Don’t lie to your players! And don’t change the rules without a good reason!
For the love of all that is good and sacred in game design, just don’t do it. Remember, the whole “solve the clues, find the Droid X, keep the Droid X” thing? Well, that actually turned out to be false. After a couple weeks of attempted clue solving and location searching it seemed as if the community was royally failing. Individuals had spent weekends driving up to 8 hours one-way to search where they believed the clues clearly designated. They weren’t doing it just for the phone as the cost of gas and hotel easily overwhelmed the cost of the device, but instead were in it for the excitement of the game (exactly what any game should be striving for).
Then, suddenly, the official rules were posted on the Twitter account. What did they reveal? That, to date, the Droids had not been placed in their locations. In fact, the clues that players had spent the previous weeks analyzing were meaningless as @DroidLanding would be posting the exact coordinates of the locations at the time a certificate (representing the Droid X) was placed. Was the fact that the @DroidLanding was a race-style game instead of a clue-based scavenger hunt what angered a large number of core players? No, it was that @DroidLanding had marketed itself one way, leading people to act accordingly, only to change their foundation with no stated reasoning.
It makes sense that Verizon would want to run a race-style event where they could have camera crews on hand to film dozens of people rushing in to search for the certificate. However, this is not the reality that was presented to players that felt like the previous two weeks had been nothing more than a sham to string them along while @DroidLanding built a following.
So, don’t change the rules in the middle of the game or lie to your players. Sometimes things happen and you do have to change the game. When you do, explain why, even if that explanation is within the narrative of the game. You’re players will appreciate it, I promise.
I’ve certainly glossed over many of the aspects that either helped or hurt @DroidLanding succeed, but these were a few that stood out as I looked back. I’d definitely love the chance to chat with the coordinators to discuss their strategy/the past month from their view. Maybe someday. Overall, Verizon’s @DroidLanding provided an engaging month leading up to the launch of the Droid X – far from perfect but successful.
While back on campus for one last Board of Governors meeting I decided to take a walk around campus. We have a ton of projects underway this summer and I was curious to see how they were progressing.
Let’s take a look.
There is now a concrete platform sitting between the Library, Magruder, and the SUB. Expect an awesome sculpture to mysteriously appear in the coming months.
Ryle Hall renovations continue. The first picture shows a few of the new bike racks that now line sit in front of the building. The second shows one of the outdoor balcony area.
The Pershing expansion seems to be moving along nicely.
Passing the DPS building I noticed a sign on the door. Pretty sure this happened last semester, but the move to Grim Smith seems complete. Both the DPS building and a wing of the Grim-Smith Hospital (the one pictured below) are scheduled to be demolished.
Starting this fall all on-campus students will receive a new id card. These are necessary for the implementation of the Perimeter Access System (which you can read about here). Managing to get ahold of one of these cards I was curious how they could practically be used. To open a door you must bring the id up to the small black box above the handle. In practice it would seem like you have to get it within an inch for the door to unlock.
The good news is that this means you should be fine leaving it in your wallet and just tapping your wallet against the box. The bad news is that with the box above the handle I don’t see how you would be able to open the door without taking your id out of your pocket.
And thus the exploration was complete and it was time to drive back to Kansas City.
