Editor's pick: Artificial arbiters will never work
Ariel Procaccia’s proposal that artificial intelligences should help resolve disputes (22 August, p 27) is one of the scariest things I have ever read. He wants to base this on economics, on the grounds that “a lot of work has been done on how to formalise, in mathematical terms, what fairness means” and to “design algorithms that can provably achieve those notions of fairness”.
It scares me that someone can think this many very silly things at the same time. He thinks it desirable and feasible that computers settle human disputes over issues involving the complex mysteries of the human heart and soul. He thinks that we should do this using 鈥� of all godforsaken ideas 鈥� the theories that underlie economics. He clearly hasn’t looked out of his window recently. Does it look as if the pseudoscience of economics has produced a world that is fair?
Finally, he thinks that a definition of fairness so maimed and truncated that a computer can process it is the best one to choose. This is what happens when education becomes too narrow. Procaccia ought to take a wider view that values history, philosophy and literature. Note what happens when you try to apply the rigorously “fair”, from “cut the baby in half” judgement to ruling to take a pound of flesh but no blood.
Birmingham, UK
Consciousness, illusion and agency
When I pondered mathematical problems in my youth, my brain would sometimes take me on a ride, rushing me through the final stages to a solution. I would then have to backtrack consciously to check it through.
With rose-tinted spectacles, I remember this only leading me to correct solutions, but there probably were misfires, too. I guess my non-conscious mind took over, as Peter Halligan and David Oakley suggest it does (15 August, p 26).
This strikes me as an interactive process between one’s “minds”, rather than the conscious mind being, as they postulated, an evolved social contrivance which doesn’t have much to do with decision-making processes.
This internal interaction seems to me to be highly advantageous, whether for social skills or for individual action.
Sheffield, UK
Consciousness, illusion and agency
I was particularly struck by how Halligan and Oakley’s ideas fit with the concept of responsibility. The conscious self is not in direct control of every little action yet, socially, it is held responsible for those actions.
Compare this with the notion of responsibility in government. Ministers are held responsible for the actions of their ministry. Since ignorance is no defence, they will make sure they know what is going on, as a matter of self-preservation. By arguing that consciousness exists for sociality, Halligan and Oakley are implying that responsibility is an evolutionary invention.
Cygnet, Tasmania, Australia
Consciousness, illusion and agency
I’m reminded of Graham Lawton’s observation that humans tend to seek “agents” with purpose (4 April, p 28). There is a survival advantage in always being on the lookout for the causes of things that happen around you. Could consciousness be a constructed “agent” explaining one’s own unconscious actions as things that occur in one’s local environment? This construct would provide a valuable interface with other local conscious “constructs” (that is, people), and allow the establishment of social groups.
Lymington, Hampshire, UK
Consciousness, illusion and agency
Anthony Castaldo takes issue with the idea that consciousness should be linked to language (Letter, 12 September). He brings up the example of his dog consciously reasoning that it should “ask” to be let out (to avoid making a mess on the carpet and angering its master, I suppose).
I agree that dogs surely possess consciousness. But here the dog is actually communicating, albeit non-verbally. It remains an open question for me whether only social animals like dogs and humans benefit from being conscious. Are naturally lonesome creatures like frogs and turtles conscious too? Maybe consciousness is really just a side effect of communication, even if it appeared long before our species and our verbal language.
Granby, Quebec, Canada
<b>First class post</b>
What is needed is the maximum number of people participating in an economy, driving growth
Michael Knight into Facebook responses to a “living wage” (3 October, p 28)
Invasion of the killer squid
I was intrigued to read Michael Tennesen’s report on Humboldt squid (12 September, p 32). What must be a fear is their potential spread from the Eastern Pacific into the Caribbean Sea and the Atlantic and Indian oceans.
One voracious predator that is already having a powerful influence on coastal marine ecosystems is the lionfish Pterois volitans from the Western Pacific. It has , where reef fish have not evolved to deal with this predator, with disastrous results.
Hurstpierpoint, West Sussex, UK
A sequel to the great land grab
Martin Pratt correctly praises the non-violent process used to divide up the Arctic seabed (29 August, p 24). His nonchalance about the fact that what was once “unowned” will now be split between three already wealthy states is, however, a bit distressing.
Few people recognise that the process of dividing up the seas over the past 40 years has been the largest grab of territory and resources since the (1884-5) divided Africa between European empires. Could Pratt remind us of the benefits of this process for the majority of humanity, especially the poorest?
London, Ontario, Canada
The universe has freedom, if not will
John Clark asked for a definition of free will (Letter, 5 September). Here are three: self-control, self-determination, self-causation.
Freedom, being self-caused, is neither random (uncaused) nor deterministic (externally caused). Since causal loops pervade all living processes, it follows that all life possesses free will, to at least some extent. And if everything in the universe has causes in the universe, then it too possesses freedom, though perhaps not will.
San Francisco, California, US
The universe has freedom, if not will
I agree with Clark that the known laws of physics seem to rule out free will, but I disagree with his implication that they are definitive. The issue in which his letter appears has “10 questions physicists can’t answer” on the cover. I would add that physicists cannot explain dark matter, dark energy and, even with the Higgs, 98 per cent of mass. Nor can they reconcile gravity with quantum mechanics. I could go on.
Free will would require the future to be non-deterministic and non-random, open to our choices. It would mean our observation of that being true is not an illusion. I don’t know what model of physics might support that, but until we have a coherent theory of everything that really does explain everything, I can rest easy because the known laws of physics are clearly not all there is to know, and will undoubtedly be subject to revision.
A hundred-odd years ago, with no knowledge of nuclear fusion, we could not reconcile the apparent ages of Earth and the sun. A hundred-odd years from now, free will might fit quite comfortably within the revised laws of physics. In the meantime, the evidence of my own eyes will take priority over any theory with a dozen holes in it.
San Antonio, Texas, US
Here's one that I invented earlier
Paul Marks’s article on “Eureka machines” (29 August, p 32) that can produce inventions was very interesting, but I was surprised to see no mention of the theory of inventive problem-solving, also known as TRIZ. This was originally based on analysis of Russian patent literature, deriving 40 principles allowing a problem formulated in one technical discipline to be assessed against solutions from others 鈥� often with surprisingly useful results.
Wickham, Hampshire, UK
Here's one that I invented earlier
Working as a patent professional, I likely dealt with algorithmically generated inventions, mainly in the field of pharmaceuticals. But the art of patent examination involves judgement not just on the objective problem but also on the nature of a to whom the claimed innovation would be non-obvious. I assume case law will evolve to accept that this could involve an artificial intelligence.
As Greg Ahahorian says in your article, patent examiners can use the same algorithmic tools as inventors do. So judging the “obviousness” of an invention might turn into a button-pushing contest between the two sides.
There are also inventions whose originality lies in the elucidation of the problem itself. These seem less amenable to automation at present, even if trawling web forums and blogs might be a productive way of harvesting human output.
Chichester, West Sussex, UK
Human origins as a dense network
I look forward to the dating of the Homo naledi fossils discovered in South Africa (12 September, p 8). I hope, too, that surviving DNA fragments may yield data on lineages. I wonder whether the fossils may be of more than one hominin species, with cross-breeding 鈥� as seems to have happened between Neanderthals and our other ancestors. Could hominin evolution have been powered in part by hybrid vigour? If so our genealogy may come to resemble a network more than a tree 鈥� creating yet more headaches for palaeontologists.
Roade, Northamptonshire, UK
Enough with the godswollop
Bryn Glover (Letter, 5 September), like Alan Webb (4 April), trots out that old Aunt Sally: the atheist whose “axiom that there is no god” is based on faith, not evidence. Redefining the word “atheist” to mean nothing but this mythical beast, rather than real godless folk like me and my parents and grandparents, allowed Webb to claim that “being an atheist is just as irrational as being a theist”. Glover similarly, invents “anti-theists… as irrational as… current theists”.
I suspect that Glover and Webb quite unaxiomatically disbelieve in Inari, the Japanese fox-headed god of agriculture, despite his many worshippers, in just the same way as we godless heathen disbelieve in the One God of Christians, Muslims etc. Atheism is a faith in the same way that not-playing-tennis is a sport.
London, UK