fluxcapacity

Last week I went to the Attic with the guys from Ignighter and Ben from Devver. Ben told us the story of his friend who was talking to a girl he liked at a party in college. Apparently, at some point there was an awkward lull in the conversation. Instead of trying to fill the void by bringing up an interesting subject, Ben's friend pointed his thumbs behind him in a vague direction and asked the girl, "Is this cool?" I'm don't know exactly what happened after this, but it probably didn't involve an exchange of phone numbers.

We started talking about the question "Is this cool?" and we realized that the meaning varies greatly based on the context of what you're doing with your hands and body while asking the question. Here are all the possibilities we came up with (the red arrows indicate hand/body movement):

Figure 1

In Figure 1, modeled after what Ben's friend did, the asker's hands are moving around in a vague circle. The asker is checking whether the situation around the himself and the askee (a party, a funeral, a ballgame,etc) is cool.

Figure 2


The asker in Figure 2 is either checking if the drink he is consuming is considered cool, or he's wondering whether it's cool that he got another drink on your tab even though you told him he's cut off.

Figure 3

If an unfortunate confrontation has occurred and a verbal exchange has seemingly ended, one can use "is this cool?" accompanied by the body and hand movement shown in Figure 3 to certify that the situation has been resolved. It's important to note, however, that this will only work if the asker is reasonably bigger and stronger than the askee.

Figure 4


Asking, "is this cool?" in a normal, relaxed voice accompanied by the back and forth hand gesture shown in Figure 4 means that the asker is romantically interested in the askee and is wondering if the situation between them is cool. Asking, "is this cool?" with the same gesture but with a lowered, more intense voice means the asker is interested in going with the askee to her or his place for the night.

Special thanks to Krista and Austin from TravelFli, and Adam from Ignighter for being "situation models".

My birthday just ended (thanks for the beers Vikas, and the 'Say It Ain't So' Rock Band production), and I really lucked out because it coincided with the TechStars collective "exhale" day, which began last night at St Julien (thanks David and Howard for drinks), and continued with a poker tournament (congrats Austin, Aziz), and a little party with Sidney's and friends (nice handstand, Kayla). I don't know about everyone else, but I think relaxing feels a lot better when it follows several days of no-letting-up intense hyper-productivity (we haz buzzwords, as PSC would say).

Andrew Hyde recently put up a clip commemorating the first 30 days for the 2008 TechStars companies in Boulder. The ten companies have come a long way in a short time. In fact, so has TechStars itself: Check out David's post detailing the inception of TechStars. Here's my summary of how well he had it figured out:

"Hey d00d [David Brown], let's lure entrepreneurs to Boulder with couches and shit."

Just kidding, sort of. We love you David.

Here's Andrew's masterpiece, but check out more on the community site:

When you devote almost all waking energy to our startup, it's not uncommon to find yourself eating and working simultaneously.  Today the dynamic duo is a fish burrito and nonlinear least squares regression (NLLS).

While the burrito is still hot, let's dive into a little compare/contrast.

Many of you may not be familiar with one of these items, so I'll give that item a quick introduction:  A fish burrito is a mexican entree fashioned most typically out of fish, lettuce, a tortilla, and an interstitial sauce of some kind.

Now that we've covered that, you are probably thinking, man, that's almost identical to NLLS.

And you're right.

• Both the burrito and the mathematical technique are customizable.

• Both are subject to local maxima -- if you start looking for a fish burrito in the wrong town, you will end up with the best of the worst.  NLLS will end up with with a solution near the initial guess, but not necessarily the best overall.  Overcoming this problem, in both cases, can be done by randomly jumping around periodically (with an airplane and random number generator, respectively).

• When someone says "fish burrito," there's obviously a lot of ambiguity.  It could mean an authentic, south-of-the-border creation -- a homemade fish burrito fashioned from leftovers -- or, as in today's case, a fish burrito from Illegal Pete's.  Saying "Nonlinear least squares" is equally unspecific.  Everyone's left wondering, do you mean a gradient-descent NLLS?  Do you mean MATLAB's lsqnonlin function?  Or, as in today's case, do you mean the Levenberg-Marquardt algorithm?

Despite these similarities, we can't end without pointing out the key difference:

• Fish burritos do not require an estimate of the Jacobian.

Because the walk to Sydney's coffee was just too far away (and closed on weekends).

Three weeks in and Andrew Hyde is starting to sneak up on TechStars companies and record them in their natural environments.  Last night he caught us at around 9pm and captured this brief podcast.  Listen in to hear about our first three weeks.

Here's an old video of the rear-projection physics whiteboard referenced in the podcast. Maybe one of these days we'll set it up in the Bunker and record the creative madness that ensues.