Mainstream and even pop sci (New Scientist) coverage of science really is awful, and often the researchers themselves bear some of the blame for bogus hype. Just today Slashdot carried an article where the researcher had discovered that sports participation in kids does not correlate with total energy expenditure -- based on the assumption that an accelerometer around the waist was an accurate measure of relative activity level when comparing "bouncing around" to distance track running. There has also been coverage of an admittedly very promising method in Go game programming, MC/UCT (a variant of Monte Carlo evaluation -- random playout until end of game, score, then repeat -- enhanced to eventually converges towards the correct minimax value), in which one of the developers of the technique says that the programs were approaching pro levels, when the programs are only at the level of advanced intermediate amateurs -- which in turn reminded me how the latest chip fabrication technique is always presented as being poised to blow CMOS out of the water, which makes it hard to understand how CMOS is still around after all these years. OK, I realize the crappiness of pop science and tech coverage is not news, but the rant bears updating every few years, especially considering how the Internet makes it easier than ever for reporters to check their facts before posting nonsense.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
no subject
Date: 2007-03-14 05:36 am (UTC)http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2007/03/11/ngreen211.xml
Mainstream and even pop sci (New Scientist) coverage of science really is awful, and often the researchers themselves bear some of the blame for bogus hype. Just today Slashdot carried an article where the researcher had discovered that sports participation in kids does not correlate with total energy expenditure -- based on the assumption that an accelerometer around the waist was an accurate measure of relative activity level when comparing "bouncing around" to distance track running. There has also been coverage of an admittedly very promising method in Go game programming, MC/UCT (a variant of Monte Carlo evaluation -- random playout until end of game, score, then repeat -- enhanced to eventually converges towards the correct minimax value), in which one of the developers of the technique says that the programs were approaching pro levels, when the programs are only at the level of advanced intermediate amateurs -- which in turn reminded me how the latest chip fabrication technique is always presented as being poised to blow CMOS out of the water, which makes it hard to understand how CMOS is still around after all these years. OK, I realize the crappiness of pop science and tech coverage is not news, but the rant bears updating every few years, especially considering how the Internet makes it easier than ever for reporters to check their facts before posting nonsense.