Editor's pick: A probable Bayesian court
Allan Reese’s letter discusses an approach to what happens in courtrooms using Bayesian statistics, and asks what prior probabilities jurors bring to the cases they try (11 April). If we accept a Bayesian approach to verdicts, the jury should be given all the evidence, along with information regarding how each part affects the likelihood of guilt.
But in the existing system, some evidence is withheld: for example, previous convictions are concealed in the interests of maintaining the presumption of innocence. In most cases, juries are still likely to set much store by the very fact that the police singled out this individual. If there were prior allegations for similar offences, that fact is likely to have contributed to that choice, so may yet indirectly influence the jury.
Would it not be useful for juries to know how the defendant was singled out: because of previous charges, whether convicted or not; by a trawl of a vast DNA database; or in a far more select group of cases by anonymous tip-off? Juries may, however, be poor at assigning significance to evidence, including eyewitness accounts.
Sydney, Australia
The price of not sequencing babies
Helen Thomson discusses some concerns about sequencing babies’ genes (11 April, p 8) but does not mention the effect on the future cost of healthcare for the babies concerned.
In countries such as the UK where the principle still (just) holds that free healthcare is provided from taxation solely on the basis of need, this should not be a problem. But in countries that rely to varying extents on commercial insurance companies to pay for people’s healthcare, the existence of full genome information would have quite different implications.
It is well established 鈥� at least in folklore 鈥� that employers not only check job applicants’ social media accounts, but also look askance at applicants who do not have such accounts; what have they got to hide? Might not the same attitude develop among health insurers when good healthy genomes attract a discount, with those who make no sequence available being penalised as potential bad risks?
Glasshouses, North Yorkshire, UK
<b>Weekly social</b>
“It’s a shame. Her theorem is the most beautiful one in analytical mechanics”
@JasonVerve on Emmy Noether’s undervalued work on symmetry (25 April, p 33)
Could killer robots be more humane?
In your article on the moral dangers of autonomous, lethally armed robots, Peter Asaro says “most people now feel it is unacceptable for robots to kill people without human intervention” (18 April, p 7). The moral reasoning behind this view is intriguing. How is sending a programmed, armed robot into an area designated as “enemy occupied” any worse than, say, bombing the area from 10 thousand feet?
In fact, the level of precision and the amount of human judgement involved in target selection with the robot would arguably be greater.
There is an even stranger moral angle. Someone who is ordered to go and kill strangers in a war can suffer severe emotional trauma and other mental distress as a result. In the future, there may be societies that decide, on moral grounds, to delegate all killing of the enemy in their wars tofully-autonomous robots so as to protect their citizens from such emotional trauma. In that unnerving scenario, the bots wouldn’t be seen by those citizens as devils, but heroic guardians.
Hampton, Middlesex, UK
Could killer robots be more humane?
As I consider the question of whether we can control killer robots, I also ask: will a killer robot panic and shoot at everything in sight? Will it kill when in doubt? Will it suffer battle fatigue and shoot in uncontrolled rage?
I suggest that in future soldiers should not be authorised to kill without robotic supervision, rather than the other way around.
H枚rja, Sweden
A robot workforce would not work
It seems plausible that within a century the vast majority of work-a-day jobs may be replaced by artificial intelligence (4 April, p 18). As a computer scientist, I have yet to see one that can write code, or a compelling novel or movie: but the creative arts are a rather thin slice of work.
People, from pilots to personal care attendants, who don’t have to invent something new every day are candidates for replacement.
What happens when 95 per cent of jobs are done by machines at 3 per cent of the cost of a human (over their serviceable life)? What happens when the labour, attention or ability to think of the vast majority of people is no longer worth paying for, because a robot can do it better for next to nothing? At some point enough people will be replaced to cause economic hardship and perhaps revolution, demanding some entitlement to the fruits of robot labour in order to survive when all the non-creative jobs are gone.
San Antonio, Texas, US
A robot workforce would not work
Every article on how drones, robots and algorithms are going to replace human workers seems to end with worrying about how people will cope with the increased leisure, instead of moving to the next step, which is worrying how people will cope with not having any money, because they haven’t got a job. How will they afford the goods, services and, indeed, the robots?
This is not just a question of inequality between the rich and the rest. Consumer capitalism is unsustainable if you eliminate paid workers: who will consume the product?
Nuneaton, Warwickshire, UK
Too little noise on ocean commotion
It would have given my husband, , great pleasure to have read Sandrine Ceurstemont’s lucid and comprehensive report on noise in the ocean (11 April, p 38). For many years he taught underwater acoustics to those in defence, the oil industry and environmental disciplines.
It is heartening to see that this work is still progressing and its importance is being recognised. However, political circles and the public are still ignorant of anthropogenic marine noise. The assumption is made that if something is placed “off shore” it will have little effect on us or on the environment.
Trefor, Anglesey, UK
We believe we are rational humans
According to Graham Lawton, “Beliefs, more than anything else, are what make us human” (4 April, p 28). I guess I’m not human, then, since I decided as a geeky astrophysics student many years ago to live in an evidence-based world in which beliefs are replaced by working hypotheses.
At least, I think I did, unless somebody produces strong evidence to the contrary. Once you renounce beliefs, life seems very straightforward, and totally self-consistent. I don’t believe in global warming, but I think there’s strong evidence to support it. I don’t believe in the scientific method, but it seems to work remarkably well, so far. I don’t believe in god, but wouldn’t it be awesome if somebody produced evidence for a god? (I invite gods to try their hand at this).
A world without rigid intolerant belief is probably a gentler, kinder, more fault-tolerant world, with less persecution of those who don’t share our beliefs. Whether you have (irrational) beliefs or (evidence-based) working hypotheses probably makes little difference to daily life, but an enormous difference to the way you view the world. I am always aware of the possibility that anything I say, including this sentence, is wrong.
Racism, sexism and all the other nasty-isms could probably not survive without belief. Unless I believe I’m superior to you, it’s very difficult to justify killing you, or discriminating against you.
Baulkham Hills, New South Wales, Australia
We believe we are rational humans
While I was interested in the detail of how we perceive our personal “reality”, there was little mention of the connection with mental disorders. Studying these might illuminate our understanding of belief and lead to insights into treatments. The list of irrational beliefs held by significant percentages of people in the UK overlapped to a worrying extent with beliefs held by people deemed to have mental disorders. Indeed, often acting in “logical” accord with their beliefs and perception of reality, in a manner that puts themselves or others at risk, is what leads to people being perceived as “ill”.
Many mental disorders involve distortion of belief. People who have anorexia do not believe what they see in the mirror, for example. It seems that a substantial proportion of the UK population might thus be considered to have a subclinical mental disorder. Society would, however, be a more boring place if all eccentrics were obliged to take psychiatric panaceas.
Chichester, West Sussex, UK
Oblivious joggers running on auto
Describing electrodes that can control people’s legs, Evan Peck says that such app-run systems will stop people being chained to their smartphones (11 April, p 18). But instead of attending to a navigation app, they will be on Facebook 鈥� or Periscope (11 April, p 19). Next, will joggers run on auto, watching a feed of a golden beach rolling by?
Imagine all this coordinated in the cloud so people don’t collide. What a risk to those not linked in, with people whizzing about them oblivious to their surrounds.
Canberra, Australia
Virtual tourists from the future
Adrian Ellis suggests using virtual reality at pop concerts (Letters, 28 March). Could the perceived absence of time travellers be due to them using VR, so as not to frighten the locals?
Alicante, Spain
Keys, wallet, passwords…
It is a waste of time trying to come up with clever algorithms for generating easy-to-remember passwords (21 March, p 28). Choose hard-to-remember ones, by all means 鈥� and write them down. You know how to keep your credit cards and house keys safe. Put passwords in the same place.
Hamilton, New Zealand
<b>For the record</b>
鈥� Messenger is a smallish space probe: the crater it made on Mercury is an estimated 16 metres wide (2 May, p 6).