Showing posts with label truth. Show all posts
Showing posts with label truth. Show all posts

Sunday, December 28, 2008

Powerful Pictures

I'd like to fancy myself as one who holds his own opinions about things. That's why I get queasy come time for end-of-the-year top-N lists. On one end, I feel like I'm allowing myself to be quasi-brainwashed by other people, but on the other end, some of these toplists are so amazing.

Fortunately, there are the lists of the "top" photographs--beau
tiful and powerful and beyond ranking. Once you view them, whether or not something should be in the top ten seems so inane.

Here are some of my favorite collections and selections:
Boston.com: The Year 2008 in Photographs -- One of the most amazing collections of photographs I have ever seen.
TIME Pictures of the Year 2008 -- With an obvious emphasis on news events.
The Sydney Morning Herald Pictures of the Year -- Music to go with the pictures, too.
Pixcetera People's Choice Awards -- Generally much more lighthearted than the previous three, but there are some repeats.

Saturday, November 22, 2008

Keith Olbermann on Prop 8

Powerful stuff.

Friday, June 6, 2008

The RIAA is Suing My Printer


There is an interesting study out by the University of Washington's Computer Science department revealing flaws in the detection of copyright infringers online. In the study, it was found that enforcement agencies sometimes used merely the fact that an IP address was seen in connection with a certain file online in implicating a user, without verification that the owner of the IP address did actually download anything, or that the owner of the IP is even a person. With IP spoofing tactics, they managed, hilariously, to get three printers in the CS department to receive DMCA takedown complaints from the MPAA. Is this result likely to get noticed in the next wave of countersuits against the RIAA and MPAA? Probably. Are universities going to change their policies because of this? No comment.

See the official page for the paper, an FAQ, and more details.


(Image courtesy of the Washington University CS Department.)

Tuesday, June 3, 2008

Personal References: Toward a Better System

Here's a hypothetical situation that is all too real: A employee has in impeccable record on paper--graduated from Stanford, interned at all the big-name companies, years of experience in key projects at well-known places, both big and small, and a very smooth interviewer, and seems extremely engaged and excited about the job. S/he is hired, and turns out to be a total slacker. He voluntarily leaves 6 months later, and takes a cushy post at another company, with good references from his two friends he made while in the current company. Lather, rinse, repeat.

This is really something to look at in the hiring process. If a company were to spend tens of thousands of dollars on some new servers, the people in IT would probably look into reviews of the servers in forums, written by current and past users, and more than just one or two firsthand accounts of how well the server functions. The IT people probably would not base the decision on the marketing brochures from the selling company and a cursory demo of the platform.

So why is hiring people any different? Hiring a qualified engineer, business development person, or marketer in the Valley will probably cost you at the minimum $70,000 a year, and can often break $120,000 a year, and yet many hiring managers never make a call to even a candidate's listed references, much less random co-workers. This problem is even more apparent in positions that are temporary or have a short hiring process, when there isn't enough time to check references, or when there isn't a good system for contacting past references.

Stanford's Residential Education (ResEd) system is a good example of this. Residential staff, especially freshman residential staff plays a huge role in shaping the lives of all Stanford undergraduates, and upperclassmen who have already been staff for a year usually are preferred over those who have not, yet there isn't a set system in place for references. The staff member's residents over the past year would be a perfect source of information about the candidate, yet there is a stigma against contacting them, resulting in bad candidates being rehired year after year.

When a draft of a scientific paper is sent to a prestigious journal for publication, the authors (and eventual readers or the journal) expect that the article will undergo double-blind peer review by several other scholars in their field. This ensures that the final published article will be legitimate and have at least a backing of several people in the field. While double-blind reference calling may be difficult, why is there such a stigma against calling random people with which the candidate interacts? The standard system of references is inherently flawed because the candidate provides the names. It's as if you were to ask Microsoft if Windows is secure.