Transparency on political advertisement targeting
Annalee Newitz is right that the real problem with fake and lying political advertisements isn't that Facebook and others promulgate them, but that they allow them to be micro-targeted, thereby avoiding detection by the rest of us, and by journalists in particular (16 November, p 24).
If it wanted to, Facebook could easily avoid this problem, while maintaining both its micro-targeting business model and its decision not to censor political advertising. For the current system to work, each advertisement must have micro-targets specified by the advertiser. There is no reason to keep these secret: advertisers being ashamed of them is no justification.
The micro-target specifications could be indexed along with the corresponding advertisements. The index could be searched to see the advertisements that had been targeted at any group of people. Most of the software needed to do this is already part of the system.
We have a model of global climate cooperation now
You ask us to imagine an international research institution dedicated to climate change, bringing together “the best minds from climate science, energy technology, economics, social science and beyond” (Leader, 23 November).
The , based in Reading, UK, is an international institution supported by 34 states. It employs around 360 highly qualified staff from more than 30 countries. More than 60 of its professional staff operate the of the EU’s . Its supercomputer and associated data archive are among the largest in the world. Its data assimilation system and its model of the global atmosphere, oceans, land and ice are the world’s most advanced.
With this foundation, an extension to the international research institution you envisage could surely be managed without too much delay. Can the next UK government provide the required global scientific leadership?
Keep this genie sealed up in its squirt bottle please
Michael Le Page reports genetic engineering of plants using a spray-on mix of carbon dots and DNA coding for a CRISPR system (9 November, p 7). This could be hazardous to humans. Like many nanoparticles, carbon dots easily penetrate the skin, and they are increasingly used for drug delivery.
A human exposed to that spray could be at risk of droplets entering the body through the skin or nasal mucosa, or inhaled into the lungs. If carbon dots enter the systemic circulation, then bound CRISPR DNA could modify the genome of any cells it infects.
While this appears to be a powerful new gene-editing tool in plants, there is a risk that it might edit human genes as well. Until we learn more about it, regulators ought to keep this genie firmly sealed up inside its squirt bottle.
Tree planting is something we can do for the climate
Your article on ways of removing carbon dioxide from the atmosphere was fascinating (10 August, p 18). Options included bioenergy, carbon capture and storage and sequestration of carbon in the soil. The only one on your list that ordinary people can do is to plant trees. But how effective is this?
An average person in a developed country releases about and the average tree absorbs . So to be carbon neutral just by planting trees, the average person would need to plant 680 trees a year.
There are carbon offset schemes, but the link between our money and trees being planted that wouldn't otherwise be planted is tenuous. My family is planting trees in Portugal at €15 per tree: 680 would cost €10,000. Possibly we will never plant all of them, but we count each we plant as 13 hours of carbon offset, even if we do nothing else.
That figure of 680 trees is an underestimate, because as CO2 levels have risen, the oceans have been compensating to some extent by absorbing CO2. The oceans will release some of this when we remove CO2 from the air, however we do it.
But we will do other things as well as planting trees, even as individuals, which may balance things out.
Can a group of AIs actually simulate a chaotic world?
Graham Lawton claims that a type of simulation called multi-agent artificial intelligence is about to upend the world with “highly detailed” simulations of “entire artificial societies” with “extraordinary accuracy” (5 October, p 38). For argument's sake, let's grant that we can construct such simulations – even though we cannot yet simulate even a single artificial intelligence.
Could such an assemblage ever accurately predict the future? What, for example, will the world look like in five years? That might depend on whether Donald Trump gets impeached, whether India and Pakistan get into a nuclear war over Kashmir, whether the UK leaves the EU, whether there is a pandemic of a new zoonotic disease…
What is going to accurately simulate future events like these? Even in much smaller scenarios, such as refugees in a Norwegian city that Lawton mentions, I suspect that social dynamics are chaotic and cannot be predicted.
We are rather sceptical of this 'helical engine' (1)
David Burns’s idea for a “helical engine” with no propellant is ingenious (19 October, p 15). He imagines a movable ring, whose mass is much greater when it slides in one direction than the other. But transferring energy to the ring to increase its velocity and hence, by the rules of relativity, its mass also transfers mass, and therefore creates momentum. It seems to me that the transfer of energy to and from the source exactly cancels out the transfer of mass by the ring.
We are rather sceptical of this 'helical engine' (2)
It looks hard to find any fault with Burns’s design for propulsion without a propellant, but centuries of experience tell us that, in all attempts to get around the conservation of energy, an increase in energy somewhere is always balanced by an exactly equal decrease somewhere else. My guess is that this proposal will turn out to be neglecting some effect in general relativity.
I suspect that the answer to the apparent anomaly might be found in D.W. Sciama’s beautiful paper ““, in which he justifies Mach’s Principle 鈥� the idea that the inertia of local masses is an effect of all the masses in the rest of the universe.
Several layers of the deep mystery of consciousness (1)
Michael Graziano asks how the brain, a material thing, produces what are called qualia 鈥� the redness of a rose, the stabbing of a pain, the aroma of brewing coffee (21 September, p 34).
As basically a materialist, I am forced to be quite puzzled by qualia. There are two implications in the question itself: a brain really exists; and qualia really exist. It seems to me that perhaps the only way we know that the first is true is because the second is true.
If qualia were in some way not real, we couldn’t know that we have a brain to produce them. Perhaps we shouldn’t draw a sharp distinction between qualia and materiality.
Several layers of the deep mystery of consciousness (2)
To me, consciousness means that narrative that runs in my head, like a film with added senses. Graziano appears at first to take the same view, but then conflates it with concepts that probably belong to a “lower” layer.
A key fact about this narrative, as found by Benjamin Libet, is that it runs about half a second behind reality (11 August 2012, p 10). This suggests a model of at least these three layers: independent parallel processes, analysing inputs and competing for attention; a resolution layer that selects actions based on the competing drives and embodies attention and the “global workspace”; and a narrative layer, synthesising inputs and chosen actions into experience. A delay is necessary because different sensory inputs have different processing path lengths. Maladjusted delays may well be responsible for room-spinning and speech-slurring when drunk, and for d茅j脿 vu.
If this is correct, it reopens the question of what consciousness is for. I suggest it comes into its own when we need to stop and think. It provides a feedback to the lower levels, allowing more sustained consideration of a subject.
I think I can think without language, as can my dog (1)
Peter White argues that some people without language can think (Letters, 7 September) and David Werdegar insists that thought does depend on language (Letters, 26 October). That exchange on the subject reminded me, being bilingual, of questions I am often asked: “in which language do you think” and “in which language do you dream”? After much thought, the answer is neither. I think (and dream) in concepts and images.
Language is only used to communicate those concepts and images. Sometimes, people in my dreams do speak a language.
I think I can think without language, as can my dog (2)
People with a severe form of aphasia may not identify a spoon as a means of transporting food, and therefore require help with feeding, writes David Werdegar. He concludes that thought depends on language.
I am always surprised how easily my dog identifies the hatch at the back of any car, even of a make and model she has never seen, as the place she should be let in. In any flat she straight away identifies the kitchen as the place where her water and food bowls should be.
Clearly dogs’ brains work with abstract concepts. Concluding that they therefore possess language would be a bit of a stretch.
Hypnosis and achieving goals in life and letters (1)
I write this letter in a state of hypnosis (9 November, p 34). Now I’m going deeper still, repeating, as you suggest, an “affirmation” that helps me achieve my desired outcome. “This letter will be published. This letter will be…”
Hypnosis and achieving goals in life and letters (2)
Helen Thomson describes the potential of self-hypnosis. What is the difference between this and mindfulness meditation? They seem, to me, to involve exactly the same processes.
The editor writes:
There are certainly many parallels between the two and there is likely to be shared neurobiology. For many people, the states of mind are very similar. The biggest difference seems to be that the end point of mindful meditation is generally focused on letting go 鈥� of worries, say 鈥� and attempting to be fully in the present. Self-hypnosis is much more goal-focused, concentrating on things you want to achieve.
For the record 鈥� 7 December 2019
鈥� Meat and dairy production accounts for 85 per cent of the UK’s total agricultural footprint, in the UK and elsewhere (26 October, p 24).
鈥� When Scottish greyface ewes are injected with male levels of testosterone, it is their daughters that show the symptoms of polycystic ovary syndrome (23 November, p 16).