Time to widen the hunt for solution to dietary crisis
I was very excited to see the interview with Chris van Tulleken about ultra-processed foods. It is a topic that I spend a lot of time thinking about and is very important if we want to take charge in the obesity crisis (29 April, p 46).
I welcome his clear words about what ultra-processed foods mean for us. Unfortunately, in this regard there seems to be no regulation about what the food industry can serve us.
I do think that he wasn’t clear enough (as a scientist, he was probably understandably cautious) about what it takes to change things. I say bring historians and anthropologists into the discussion. Why was it possible for people 50, 60 or 100 years ago to cook a meal every day while working 40 hours a week or more? And no, not because of housewives 鈥� there are plenty of places where both parents worked back then (and often had more children to feed).
We need a shift in our societies about what we value and what we want to do with our time. We need to bring back cooking lessons in schools. There are generations out there that don’t know how to cook, but they can learn (to love) it.
Maybe a dose of rule by logical AI is what we need (1)
I don’t understand why people are worried about artificial intelligences taking over. The dinosaurs lasted 165 million years. The first “upright apes” evolved only 5 million years ago and Homo sapiens didn’t arrive on the scene until about 300,000 years ago (Letters, 20 May).
We have only a few thousand years of recorded history. It isn’t a great record. Our world is run by men with an insatiable lust for power and/or an insatiable greed for money. Despite the warnings of science on climate change, they continue to trash our only planet. The human race will be lucky if it lasts another 1000 years.
But a government run by AI would be like Mr Spock: completely logical. Isn’t that the best hope for preserving humanity?
Maybe a dose of rule by logical AI is what we need (2)
The number of pieces on AI seems to be on the increase, especially those that warn of possible dangers. But all these appear to treat AI as though it were a sentient entity already, busy plotting and planning the downfall of humanity.
The truth is that AI, even in its most advanced form, is merely a collection of circuitry that must be built by people and then have someone take the positive decision to turn its power on.
It would be a step forward if future articles could be directed more towards the 鈥� mostly profit-seeking 鈥� motivations of the humans in the loop.
Fathoming human behaviour is hard
Peter Cundall thinks that there is no puzzle that can’t be solved by human intelligence. However, there is one that has defeated it: human nature (Letters, 13 May).
If intelligence could solve human nature, we would have begun reducing greenhouse gas emissions 30 years ago and might now be at net zero. We wouldn’t just be listening to and applauding Greta Thunberg, we would be doing what she tells us. Our governments would be working seriously to prevent climate change instead of being half-hearted about it.
New wave of chatbots shouldn't be for profit
My hypothesis is that ChatGPT’s output is a form of statistical plagiarism, using the words most probably occurring in its training texts. The only reasonable conclusion is that commercial exploitation of such technologies could be akin to theft of intellectual property, and that such systems should only be made available via not-for-profit firms (13 May, p 13).
As pressure grows, so will demand for online therapy
You correctly point out that the biggest determinant of outcomes in therapy isn’t the type used, but the relationship between the therapist and the individual. This raises an interesting point about the relative effectiveness of face-to-face and online therapy (15 April, p 38).
With growing demand and not enough therapists, the need for online therapy is likely to increase.
Phonics still has plenty going for it
Colin Barras believes that the reason why some children in England fail to learn to read adequately is too much phonics. My article , published in the Oxford Research Encyclopedia of Education in 2022, provides copious evidence for two main conclusions: systematic phonics teaching is effective for teaching children to read and spell in English, and the combination of systematic phonics teaching and whole language teaching is probably more effective than either alone (22 April, p 42).
Dark matter and energy could be red herrings
I enjoyed your recent guide to the quantum realm. However, I continue to doubt the existence of dark energy and dark matter (8 April, p 36).
That we have been unsuccessfully searching for these for many decades and haven’t found something that supposedly accounts for 95 per cent of the stuff in the universe suggests to me that they are red herrings.
There are other explanations for the fact that galaxies don’t fling themselves apart, which is said to point to the existence of dark matter. The simplest of these is that numerous small and large black holes may provide the necessary missing mass.
If physical laws evolved, why are they just right?
Thomas Hertog likens theevolution of physical laws to Darwinian evolution, with “quantum observation” playing the role of selection. He allows such observation to include interactions between particles, though how it causes a collapse of the wave function instead of an entanglement isn’t explained. But the key missing argument is how such selection favours a set of laws conducive to the evolution of life (25 March, p 38).