Cloak And Google

Friday brought a flurry of breathless news stories on the appearance of mysterious billboard along Highway 101 in Silicon Valley.

The billboard states:

{ First 10 digit prime in consecutive digits of e }.com

In almost zero seconds flat, not only was the puzzle solved, but also the owner of the mysterious advertisement and its associated web site revealed to be Google. The second part of the puzzle, a mathematics problem at the aforementioned website, required a little bit of time and lateral thinking to solve, but was similarly discovered and broadcast across the Internet in short order.

The part that interests me about this: here’s Google, renowned for its search technology – why do they have to go to these levels to find good candidates for open positions? While I can understand they are undoubtedly deluged by applications for open positions (my wife, for example, has applied several times for several Google jobs for which she is ideally suited and has not even scored an interview), this is Google we’re talking about – don’t they have a crack team of spare PhDs stowed somewhere to solve this kind of problem? And if not, what hope do they have of solving the problem of enabling search on enterprise desktops (unstructured data) when they appear unable to solve the problem for resumes (an arguably fairly structured form of information).

I’m not exactly sure what problem this advertisement was designed to solve. According to CNet, Google comments on the numerous tactics it has been using to find qualified candidates:

“As you can imagine, we get many, many resumes every day, so we developed this little process to increase the signal-to-noise ratio.”

But does this latest attempt really solve the problem of too many low quality resumes? Anyone looking for a job at Google doesn’t need to figure out the answer for this problem – ironically, they’ll be able to just use Google to find the answer now that everyone on the Internet is gabbing about it. The fatal flaw in Google’s plan was that it failed to incorporate a dynamic element to the problem – the puzzle erects no barriers to prevent a different person from using the answer. If Google had been smart, they would have made the problem different for each person who visited the first puzzle’s web site, enabling them to further cut down on the number of flakes dumping their crappy resume to Google.

That said, it was still a brilliant plan. Google seems to really have the hang of viral marketing and building value by letting its customers/audience do the work. I recently heard that Google’s initial marketing had a similar bent: they handed out golf shirts to VCs. Clever. Gmail also had a very viral marketing plan which allowed people to join by invitation. Even the original search technology is viral to a certain extent – Google leaves the problem of organizing data to web sites, and simply scoops up to the links to extract context and provide matches to search terms.

Maybe that’s what disappoints me about the latest scheme: it clever, but not as clever as I’ve come to expect from Google.

And for those of you who simply arrived here searching for the answer to the math puzzle, here it is. After going to http://7427466391.com you’ll be presented with another anonymous web page stating:

Congratulations. You’ve made it to level 2. Go to www.Linux.org and enter Bobsyouruncle as the login and the answer to this equation as the password.

f(1)= 7182818284
f(2)= 8182845904
f(3)= 8747135266
f(4)= 7427466391
f(5)= __________

So, what’s the next number in the sequence? Once you realize the sequences are from the constant e, there were a number of ways to attack this problem – for one, you could have just brute-forced the login with a simple script to iterate through all the 10-digit sequences in e. But a little observation shows all the sequences have the property that the sum of the number in the sequence is 49. Thus, f(5) is the next 10-digit sequence in e whose individual digits sum to 49. Rather than solve this problem manually, you could just write a quick piece of code to solve the problem. For your reference, the digits of e are:

71828182845904523536028747135266249775724709369995
95749669676277240766303535475945713821785251664274
27466391932003059921817413596629043572900334295260
59563073813232862794349076323382988075319525101901
15738341879307021540891499348841675092447614606680
82264800168477411853742345442437107539077744992069
55170276183860626133138458300075204493382656029760
67371132007093287091274437470472306969772093101416
92836819025515108657463772111252389784425056953696
77078544996996794686445490598793163688923009879312
77361782154249992295763514822082698951936680331825
288693984964651058209392398294887933203625094431…

While I could provide you with the code to perform the final step, I’d hate to kneecap Google’s efforts entirely. Most of the hard work is done – if you’re really that lazy, you might try looking at the screenshots on MarketingVOX – they’ll show you the email address where you end up submitting your resume. I’m not sure if this is cheating or not – sometimes the best way to win the game is to avoid having to play it at all – but I’ll leave it up to you to decide whether that’s right or not.

Non-DRM Constitution

Yesterday, John Beimler pointed Comrade Cory over at Boing Boing to this DRMed version of the US Constitution. Someone must be doing well – it’s apparently Amazon’s 965th highest ranked item. I can only hope that’s only for electronic book sales, otherwise there are a lot of stupid people in this country.

Putting aside the irony of restricting access to a government document with digital rights management software, they’re not the only ones trying to make a buck this way. Hell, even I slapped an Amazon Associates referral ID on the link above – if you’re stupid enough to buy a DRMed copy of a public document from Amazon, I deserve a cut! And arguably, there is some minute amount of value that someone must derive from being able to get a Microsoft Reader copy of the Constitution right now. It’s possible. I guess.

But if you don’t like DRM and you really need the US Constitution (or any other document for that matter) on the go, there’s an easy way to solve this problem: make your own! And make the publisher’s market disappear while you’re at it! Doing this will take you about ten minutes:

  • Install (soul-sucking) Microsoft Word 2002 or later, assuming if you haven’t already succumbed by way of peer pressure.
  • Download and install Microsoft’s Read in Reader conversion plug-in.
  • Open Word and cut’n’paste the contents of you favorite version of the US Constitution into a new document.
  • Click the Microsoft Reader button on the toolbar.

(Or if you’re against The Man, but horribly lazy, you could just download this version I created. And no, I’m not the author of the US Constitution – I’m just really into flagrant self-promotion.)

Voila! Instant un-DRMed version of the US Constitution suitable for reading with Microsoft Reader! Of course, this does beg the question: did I just effectively circumvent the copy protection mechanism and violate the DMCA? Oh well!