1 min read29th Jun 20214 comments
This is a special post for quick takes by Nihal M. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.

New to LessWrong?

4 comments, sorted by Click to highlight new comments since: Today at 5:27 PM

Safety is a slippery slope. 

Misplaced excess concern over safety hinders technological progress, and is a slippery slope. 

Experts and regulators who are held accountable when things go wrong with new technology, and are thus incentivized to raise the bar or standards in an effort to make systems "more safe". 

This has detrimental effects in two ways -
- The slippery slope of adding more complex regulations or roadblocks in the name of safety. 
- Disincentivizing technology adoption and slowing progress.

 

Examples from experience :

  • Why we don't have drone deliveries yet. The technology was ready more than half a decade ago, but restrictive regulations didn't help in learning from real word experience and increased adoption. The infrastructure and other problems that would've been discovered and solved at scale have never materialized because everyone is still in the boardroom brainstorming scenarios and ideal workflows
  • Flying Cars suffer the same problem. Just replace "drone deliveries" with "flying cars" in the above paragraph and it mostly stays the same.

External examples with no experience 

  • Roots of progress talking about why Nuclear energy failed. Public perception of safety was a huge driving factor in regulations which made it near impossible to setup new plants in the US. General public perception of the horrors of nuclear technology is still largely negative and cause strong reactions that are pro-safety. This led to a slippery slope in more complex and burdensome regulations resulting in an economic disincentive for industry,

Flying cars are actually a safety problem. In cities they are also likely to be a noise problem with the technology that was available for a long time. Finally energy costs matter a lot for flying cars. The oil shock of the 70's reduced helicoptor usage and that never recovered.  

Flying Cars have larger safety consideration. The point on flying cars I made was a bit hyperbolic which was probably unnecessary in hindsight. 
 

Thinking out loud,
1. While Safety cars expand the risk envelope, the difference is not too much different either. I can expect a drunk driver to ram into a store or building at ground level these days, and I expect that living in the fourth floor of an apartment is safer cause there's no precedence for incidents like that. But someone crashing a vehicle to the ground floor or to the third floor can be handled in a similar legal framework with varying levels of punishment. If there were no cars, we'd be concerned of the idea that someone could crash a heavy machine to our homes at large speeds. The analogy is similar to flying cars, just the precedence is missing. 

2. The current breed of flying cars are heavily electric and shouldn't face too much concern over energy costs.

3. Not sure of this, but could the noise be normalized over time? Like how (some) people living near airports are used to the planes taking off and landing that they kind of tune it out? urban noise is largely from traffic, and this can be another dimension to it? 
 

I'd agree, especially about nuclear energy and perception of safety as opposed to actual safety.  

For most examples, though, "it's complicated".   Many are a mix of safety concerns, NIMBY limitations, and legitimate worries about allocation of resources.   Drones (including passenger drones, not including human-piloted flying cars) are likely pretty safe, with less human harm per delivery than a low-paid contractor driving a truck.  But the airspace rights are hard to value and shouldn't be permanently given to one or a few companies who get there first.  And the ability to be outside without the sky having a bunch of distracting moving things always there is worth something too.