Month: December 2009

From Art to Science: what it means

Posted by

0

Most of my research right now is about the evolution of technologies. They go from crafts, requiring skilled experts, to “engineering science,” i.e. mostly automated and very precise. For example, firearms manufacturing took 200 years to undergo this shift. Flying took about 100 years to go from the Wright Brothers, to autonomous aircraft (not just unmanned, but self-directed). How does this happen? Is it a good thing?

Here is a talk I gave on this topic. (Caution: 5 MB PDF file) The subtitle is Why old tasks get easier, but everything gets more complex.

Bohn knowledge evolution 2007

I’m working on a book on this subject, which does side-by-side comparisons of:

  • Flying
  • Medical care – several kinds
  • Firearms manufacturing (from Napoleon to 1980)
  • Semiconductor manufacturing

Each of them has undergone major transformations, with similar patterns.

How 3.6 Zettabytes of Data Get Consumed – hmi – Gizmodo

Posted by

0

You probably already saw that the average American tears through 34GB of data per person per day. Here’s how the media has evolved these last few decades (sorry print), and below a way to compare your consumption with Joe Average.

This chart breaks down each activity by hours, bytes, and words for the total population, average per user, and average per American in 2008. There’s a lot to process here, but my first reaction is: that many people still watch TV in standard def?

Send an email to Brian Barrett, the author of this post, at bbarrett@gizmodo.com.

via How 3.6 Zettabytes of Data Get Consumed – hmi – Gizmodo.

————————————-

Response

What surprised me is not the amount of standard definition TV, but the large amount of live TV overall. It’s hard for me to find people who watch TV live – they either Tivo it, or watch something on the Internet, or rent DVDs with the shows they want. Unfortunately, the Nielsen TV data that we used lumps anything seen within 72 hours of original broadcast into its “live audience” report. Also, our data is for 2008; by 2010 I expect to find a lot less truly live TV. Finally, it is possible that the people I interact with are not typical. But even my computerphobic father watches recorded C-SPAN.

RB

Intel-AMD case

Posted by

1

The lawsuit between Intel and AMD was settled a few weeks ago, five years after it was first filed by AMD. This was a private civil lawsuit; various government investigations are still going. According to reports, Intel agreed to pay $1,250,000,000, and to extend various license agreements with AMD. ($1.25 billion)

I was an expert witness on the case, which was an interesting experience to say the least. As an academic, I wish there were a way to use the data from civil lawsuits  cases for academic research. Of course, the confidential data itself has to be kept secret, but in principle it should often be possible to find ways to provide high-level analysis without revealing anything private.  As far as I know, though, this is never done. Certainly the standard agreements for using the data don’t allow it, and in practice it would probably take an agreement by both sides. The companies don’t seem to have much incentive to allow this. (I’m not speaking of AMD or Intel  or anything about this case in particular.)

There’s one  situation where some of the data does come out, namely if the case actually goes to trial. An example was the US vs Microsoft case in the late 1990s. But this is very rare – most cases settle before trial, and even when they don’t I gather that agreements to seal the data are standard.