20 August 2002

UK may ban driving and talking The Government says that research proves we're 4 times more likely to have a car accident if we're on the phone (Redelmeier and Tibshirani, 1997) - so proposes (PDF) to fine drivers up to £2,500 (instead of the current £30/£1,000 fine/conviction for failing to have proper control of a vehicle). Drivers would still be allowed to use hands-free mobile phones, although the paper's definition (Annex A, para 9) of "hands-free" is not mobile-plus-hands-free kit, but instead a phone that (a) is wired into the car and (b) does "not require the driver to significantly alter their position in relation to the steering-wheel to use it" - ie: no-one has a hands-free phone?! Three bodies comment on this: the AA says that drivers should also be stopped from smoking and eating at the wheel if the use of mobile phones is to be banned. The RAC says "The only problem with specific legislation is where do you stop? Retuning the radio probably causes more accidents than using a mobile." And Rospa, one of the sources of the Gov's research (PDF), backs the legislation but sees no difference in the potential danger of hands-free phones and mobiles. The Govt's consultancy report (PDF) notes that the value of preventing a road fatality is £1.1m (original source), but has no data on how many are caused by mobile phone use, nor on how many of those would be prevented by legislation. 10:45:50 PM     comments

Voyager 1 and 2 heading out of the solar system V1 was launched 25 years ago today in 1977, and will reach the heliopause in 7-21 years and leave our solar system. Both spacecraft carry a golden disc containing messages from earth. 10:25:29 PM     comments

Malcolm Gladwell: Can you read people's thoughts just by looking at them? 1. Some people can (scary for the rest of us?) 2. It's a two-way street: facial expressions create as well as display emotion (interesting) 3. how long before security companies use this instead of biometrics for eg airline security?: instead of trying to match an actual face to a watchlist of 100,000s, measure emotion/intent from a "lexicon" of thousands on a face (interesting/scary?) 10:23:45 PM     comments

Schneier and Ozzie: Foolproof doesn't work, smart security and smart failure instead Charles Mann's Homeland Insecurity article for The Atlantic (Google-cached as The Atlantic's own link doesn't work for me) is a great overview of Schneier's view of security: Security that depends on secrecy is doomed to fail Cryptography is not enough; niether is any technology-only solution don't rely on single points of security/failure (eg: the firewall which protects only the boundary of your network) build failure into the design: no single failure should catastrophically break the system decisions must made made by people (not computers) at close range Once humans did the encryption (it was harder to encrypt your messages than to send them plaintext, so many didn't) and everyone hoped that technology would take care of the security (it didn't). Now the reverse is true: today Schneier advocates having humans do the security and technology silently handle the encryption. Very related to this: Ray Ozzie on complacency immune security, which is built into his Groove: It's not the individual's fault! It's up to us - the technology industry - to create systems that are complacency immune - that are designed to be complementary to the way that users and administrators really act. And it's up to IT to realize that it's their responsibility - likely to the point of liability - to broadly deploy technology that is configured to be secure in a complacency-immune fashion. No, it won't be perfect: this is all about risk management. You can't control how people behave - so create an environment in which they do the "right thing" naturally. Also: Will complacency ultimately net out to liability? 12:11:14 AM     comments