Haven't looked at the company in depth, but I think they circumvent the really 'tricky' issues by giving you info only on the 'fun' genes. That is, whether you are likely to turn blue when cold. As opposed to predicting your chances of being diabetic. It's like a half-way solution to getting people more comfortable with DNA testing. (kinda like Sony's robot dogs and RoomBa getting people more comfortable with the idea of having a robot around)

I was reading an article some time ago about Moore's Law as applied to DNA sequencing costs. Not exactly doubling every 18mo, but the rates have certainly come down since early 2000's.

I look at a job pretty plainly as a balanced (or balancing) equation.

Compensation ($) + Learning/Excitement = Work Output + Stress/Dissatisfaction Endured

At any given point in time, each of these four variables will change (some more frequent and volatile than others). If the left-side of the equation is larger, then you've got the better side of the bargain. At which point, the company should either reduce your pay, give you more work, or make you suffer. If the right-side of the equation is larger, than you're getting shafted and should change something or leave.

People quit their jobs (or get fired) when one side of the equation tilts and exceeds a 'threshold' - straining the employer/employee relationship. The value of the threshold is different for everyone and for every company. For instance, a person w/ too big an ego (or even false expecations) may have a higher expectation for the left side and lower expectation for the right side. A company w/ a line of applicants queuing around the block to get in may have the inverse. I suppose we live in a world of mix-and-match - whereby these randomly distributed molecules (applicants and companies) collide.

Side note: I wonder how you would apply prospect theory to employee promotion or hiring strategies. Maybe using horizontal axis as salary increases/decreases and vertical axis as some sort of subjective work variable? How would this encourage or discourage promoting employees early? How would this help predict salary cuts and the impact on work produced?

The goal is to create an integrated solutions offering with multi-product lines. However, I didn't think it smart to compete w/ Intuit's Quicken and Sage on the first go... =/

I have to say that every once in a while, I do take a closer look at the right side ads (may not always click though). However, my behaviour to do so is prompted by something catching the attention of my peripheral vision (ie, my 'exact' search terms appearing in the AdText). Most of the time, my focus goes directly to the left-hand organic search results. So perhaps the eye tracker isn't taking peripheral vision and subconscious absorption into account?

What does everyone think of a website like TopTenReviews.com? It force-categorises various product lines but I find the reviews (especially the side-by-side features comparison) quite helpful.

we should host a competition for next year's IM class... to see who's got the most popular blog!

hmm... still not sold. perhaps to me, microblogging isn't a 'need' just yet...

My knowledge of anything 'semantic' is limited. But I think start-ups like SkyGrid are doing exactly what you described. The hinderance of "Semantic Web" is getting everything meta-tagged... lots of work. Which is why most "semantic" search engines seek to squeeze more info out of existing data (eg, TrueKnowledge, PowerSet) rather than push forward the meta-tagging.

1. Quite right - I have yet to see the Fandango model here. In the UK, there are 4 main cinema chains: CineWorld, VUE, EMPIRE, and ODEON. ScreenRush.co.uk just shows clips and info.

Response to 8: Companies like MaxRoam and UnitedMobile are trying to solve this issue... but dealing w/ MNOs across countries isn't easy...