Your writing has a wonderful combination of nerd and wit. Well worth the subscription.
Wasn't sure I'd enjoy a piece on autonomous buses. Then realised I'm also a bit of a public transport geek. And I live in Bristol where the buses are WOEFUL. Good (obvious, but necessary) point about labour costs on self driving buses.
Now to continue reading some other of your posts in a random order...
On the night buses thing. It's true that the big limitation to running at 4am is the labour costs, but the reason they don't run between 11pm and 3am is the limitless potential for the drunk British public to shit (sometimes literally in this case) on a good thing. Many UK cities curtailed their night services not due to lack of demand, but due to large groups of drunk people fighting, vomiting and other whilst onboard. And that was with a driver who could kick the worst offenders off...
One thing to note is the question of deskilling on the job (danah boyd had a good blog post on this: https://zephoria.substack.com/p/deskilling-on-the-job?publication_id=267489&isFreemail=true) which in summary is the way in which as more and more of a given job is automised, the people who do those jobs have less and less practice and therefore get worse and worse at it. Now this would be irrelevant if we get to a point where the buses don’t need any safety drivers at all, but if either a) we need to keep safety drivers during a transition period until we get to that point or b) the tech never gets to that point (or passengers never feel safe enough without them), it could potentially be a problem if the safety drivers get worse and worse and aren’t acc able to help all that much as a safety backup. We’ve seen the same thing with airline pilots (as mentioned in the linked post above) where more and more of a plane’s functions are automised, but pilots are kept in for safety purposes, the pilots end up deskilling cus their just not doing most of what a pilot would need to do very often. Now maybe we’ll get to a point where driverless vehicles are safer than humans, in which case that won’t matter to much, cus it will just be safer generally anywhere (and as I said it will defo be irrelevant if we get to the point where we don’t need drivers at all, which as you argue would be a major benefit of driverless cars). But at the very least it something to keep in mind for at least the transition period
The reporting of that Gove quote really winds me up - It's the modern day equivalent of the Mandelson "intensely relaxed about people getting filthy rich (as long as they pay their taxes)" line. What Gove said was "I think people have had enough of experts (telling them what to do all the time)". Which was arguably quite an astute observation.
Anyway, really interesting article. I think to some extent the clinching argument is the significant reduction in the resource cost of drivers if you can achieve a mostly autonomous system. However, given how little we spend on bus services (relative to other modes, car especially) we can overcome most of this problem by simply upping the state subsidy and/or nationalising services. Not quite as sexy or sellable to the tabloids admittedly.
And that includes the poor who don’t have legal brakes or tyres.
Proving an autonomous vehicle can drive around an american suburb with lowish speeds, wide roads, few pedestrians, cyclists and horses and no ice and little rain doesn't share much with normal driving.
Your writing has a wonderful combination of nerd and wit. Well worth the subscription.
Wasn't sure I'd enjoy a piece on autonomous buses. Then realised I'm also a bit of a public transport geek. And I live in Bristol where the buses are WOEFUL. Good (obvious, but necessary) point about labour costs on self driving buses.
Now to continue reading some other of your posts in a random order...
On the night buses thing. It's true that the big limitation to running at 4am is the labour costs, but the reason they don't run between 11pm and 3am is the limitless potential for the drunk British public to shit (sometimes literally in this case) on a good thing. Many UK cities curtailed their night services not due to lack of demand, but due to large groups of drunk people fighting, vomiting and other whilst onboard. And that was with a driver who could kick the worst offenders off...
One thing to note is the question of deskilling on the job (danah boyd had a good blog post on this: https://zephoria.substack.com/p/deskilling-on-the-job?publication_id=267489&isFreemail=true) which in summary is the way in which as more and more of a given job is automised, the people who do those jobs have less and less practice and therefore get worse and worse at it. Now this would be irrelevant if we get to a point where the buses don’t need any safety drivers at all, but if either a) we need to keep safety drivers during a transition period until we get to that point or b) the tech never gets to that point (or passengers never feel safe enough without them), it could potentially be a problem if the safety drivers get worse and worse and aren’t acc able to help all that much as a safety backup. We’ve seen the same thing with airline pilots (as mentioned in the linked post above) where more and more of a plane’s functions are automised, but pilots are kept in for safety purposes, the pilots end up deskilling cus their just not doing most of what a pilot would need to do very often. Now maybe we’ll get to a point where driverless vehicles are safer than humans, in which case that won’t matter to much, cus it will just be safer generally anywhere (and as I said it will defo be irrelevant if we get to the point where we don’t need drivers at all, which as you argue would be a major benefit of driverless cars). But at the very least it something to keep in mind for at least the transition period
The reporting of that Gove quote really winds me up - It's the modern day equivalent of the Mandelson "intensely relaxed about people getting filthy rich (as long as they pay their taxes)" line. What Gove said was "I think people have had enough of experts (telling them what to do all the time)". Which was arguably quite an astute observation.
Anyway, really interesting article. I think to some extent the clinching argument is the significant reduction in the resource cost of drivers if you can achieve a mostly autonomous system. However, given how little we spend on bus services (relative to other modes, car especially) we can overcome most of this problem by simply upping the state subsidy and/or nationalising services. Not quite as sexy or sellable to the tabloids admittedly.
2 million miles travelled by autonomous car with a test driver isn't that impressive when in america which has a poor safety record there are ~70 million car miles between deaths - https://en.m.wikipedia.org/wiki/Motor_vehicle_fatality_rate_in_U.S._by_year
And that includes the poor who don’t have legal brakes or tyres.
Proving an autonomous vehicle can drive around an american suburb with lowish speeds, wide roads, few pedestrians, cyclists and horses and no ice and little rain doesn't share much with normal driving.