Human Compatible

Human Compatible
AI and the Problem of Control
by Stuart Russell
Published by Allen Lane

Let’s be clear, Stuart Russell, Computer Science professor at UC Berkeley and Oxford, doesn’t think we have to worry about malevolent robot overlords, sentient Internet or Terminators any time soon. Why waste time worrying about the destination he asks, when the journey is so full of peril?

We’re already experiencing some of those perils in the unintended consequences of filter algorithms contributing to the growth of political polarisation and tribalism. Next up will probably be how our self driving cars make their decisions. Or maybe it will be the rules governing the behaviour of autonomous military robots. Or possibly corporate planning algorithms or medical research or, or, or…

These sort of partially intelligent AI systems will infiltrate more and more of our daily lives in the coming years and decades simply because of their huge potential (an extra $674 trillion per year, or a tenfold increase, in global GDP suggests Russell). AI systems that may not be conscious but are motivated and goal-orientated with an astoundingly greater grasp of situational variables than humans and vastly greater ability to model, and act on, those variables into the future.

Some have suggested putting the AI in a box or just turning off the power if we start having problems. Russell isn’t overly optimistic here, pointing out that a truly intelligent machine will be able to find work rounds for anything our monkey brains can come up with. After all, we can’t build a firewall that keeps other humans out let alone an AI. He’s not even on board with the solution of merging with AI by directly connecting our brains to their silicon. “If humans need brain surgery merely to survive the threat posed by our own technology, perhaps we’ve made a mistake somewhere,” he suggests.

Russell’s solution is to make sure the AI’s goals and ours are aligned before flipping the on switch. You don’t want your self-driving car mowing down pedestrians to get to your destination nor your smart kitchen eyeing up the caloric content of the cat (the shop is 20 minutes away, the cat is right there). However, as Russell’s version of AI becomes more powerful, it starts to sound like a really dumb genie. Huge potential benefits along with lots of opportunities to completely fork things up. Ask an AI to make you happy and you may end up on a permanent heroin drip. How can an AI know what we really (really, really) want when we hardly ever do? What we want at 15 is rarely what we want at 50 and often what we want may not be what we want to want, or indeed, how we want it.

All, however, is not lost. Russell has followed Asimov’s lead by coming up with three, well, not laws as such. They’re principals, suggestions, directions of thought that AI researchers should explore when building AI. They mostly centre around AI being altruistic, humble and learning about human wants and needs from observation. If that last is raising some red flags for you, you’re not alone and Russell explores a bunch of the caveats and provisos implied. “We will need to add ideas from psychology, economics, political theory, and moral philosophy”, he says. Honestly, it seems like a lot to squeeze into an AI. The big problem though, and one Russell only lightly touches on (perhaps because its more political than technological), is making sure everybody sings from the same hymnbook. Asimov’s three laws of robotics were universal because in those stories there was only one company making robots. It’s unlikely that there will be one company or country making superintelligent AI so ensuring there is a universal set of ideals installed in each from the get go will be a challenge.

The Day It Finally Happens


The Day It Finally Happens
by Mike Pearl
Published by Hodder & Stoughton

Mike Pearl – Vice columnist and anxiety disorder sufferer – worries about the future. To help quantify that worry, he’s sought advice from a selection of experts on exactly how much he should be worried about a variety of potential futures.

Pearl sets the scene for each scenario with a little fictional version of the proposed future then gives it a worry and probability rating before delving into the nitty gritty of expert opinion on the hows and whys. With an enviably easy writing style he draws you in and takes you on a tour of optimism and existential dread while pointing out silver linings and unexpected clouds en route.

Humans, Pearl suggests, don’t really believe in the future. We didn’t believe in this one after all which is why we’re in such a mess right now. You probably won’t believe all the futures explored here but you’ll definitely be better prepared to worry about the ones you do.

The Moon


The Moon
A Brief History for the Future
by Oliver Morton
Published by Profile Books

When the Israeli lunar probe ‘Beresheet’ crashed into the Moon in April 2019, it contained a library of 30,000 books engraved on 25 nickel disks. ‘The Moon’ by veteran space correspondent Oliver Morton wasn’t among them. Hopefully the next probe will preserve a copy because the real audience for this book won’t be born for centuries and maybe not on Earth.

The Moon will come to mean something else to them – holiday resort, home, industrial park or maybe just advertising hoarding. However harsh a mistress it may be to those future Selenites, Morton wants to preserve this historical record of what it has meant to us – to all the Earthbound artists and lovers, scientists, explorers, smugglers and dreamers who came before.

In elegant and lyrical prose, collating science along with myths and legends from around the world (mostly the Western parts), Morton entertainingly and informatively teases out the Moon’s place as both symbol and stone in our cultural and scientific zeitgeist.



A Brief History of How We F**cked It All Up
by Tom Philips
Published by Wildfire

“Humans” is a school report card on our first 70,000 years of term as written by Monthy Python. Author Tom Phillips (editor at the fact checking charity has scoured history to illustrate our species’ seemingly unerring ability to just not think things through. One ‘it-seemed-like-a-good-idea-at-the-time’ decision can loose everything. You can loose a battle even if the other side doesn’t show up. You can loose your country’s harvest because you don’t like birds, you can even loose a whole empire (the secret to that last one is to massacre Genghis Khan’s peaceful trade delegation).

Phillips’ humour and engaging style in “Humans” makes it easy to laugh as you learn about those dumb ass decisions made by dumb ass folk in bygone days. Just remember who the last laugh is on though. I suspect that it’s we who will be the butt of the jokes for the 2120 edition.

The End is Always Near

The End is Always Near
by Dan Carlin
Published by William Collins

What if the electricity went off tomorrow and never came back on? It would be the end of civilisation as we know it suggests author Dan Carlin. Of course, civilizations have ended before. The Bronze Age collapse laid waste to the epitome of civilistation at the time. The Assyrian Empire, brutal even by the standards of a brutal age, ruled for 2,000 years yet collapsed so thoroughly it was all but forgotten until its modern rediscovery. What does it even mean for a civilisation to ‘fall’? Did Rome fall or just “transition to a more decentralised era with a distinct Germanic flair” asks Carlin? He looks back at these events as well as some of our own close calls to explore the hows, whys and what-comes-nexts of ‘THE END’.

He starts off with a question; does civilisation make us soft? Is it always rough pioneers building what their effete descendants loose – ‘wooden shoes going upstairs and silk slippers coming down’ as Voltaire suggested? Earlier folk certainly had it tougher than today. Parents could expect half their children to die (diarist Samuel Pepys lost all six of his siblings) and if they did live, the children could expect to experience ‘norms’ of physical and sexual abuse that would consign any modern victim to years on a psychiatrist’s couch. Did all this, wonders Carlin, make for tougher societies or just more barbarous ones?

The chapter on Pandemics makes for pertinent – and chilling – reading back here in early February with the Covid-19 virus taking hold. Hope it works out OK for you future folk. Certainly better than for the 25 million victims of the Plague of Justinian 1,500 years ago or the 75 million lives claimed by the Black Death in Europe or the 50 to 100 million people who died from Spanish Flu in the early 20th century (when they knew all about germ theory, disinfection and isolation).

Wars and their moral mutability get a couple of chapters of their own and, oh, how quickly atrocity becomes routine! It takes less than 30 years for the ‘crime against humanity’ of dropping a handful of small bombs onto Paris from a single, rickety bi-plane to transform into the just-another-day-at-the-office of burning 100,000 souls to death in a 1,000 plane bombing raid on Tokyo. The awful irony is that the road to these and even fresher hells are paved by good intentions. But, asks Carlin, if thwarting an evil empire costs megadeaths, how much evil do you get splashed on you in the process? Carlin reports contemporary accounts from victims and perpetrators are, by turns, heart-wrenching and utterly chilling.

One issue is that Carlin doesn’t really have an argument as such and consequently the book lacks focus. By his own admission, it’s just a loosely connected series of vignettes (albeit engrossing ones). His background is as a radio talk show host turned history blogger (mostly the shooty, stabby kind of history) and he tells us he aims for an unbiased ‘what-would-a-Martian-visitor-think-of-all-this’ viewpoint. I don’t think he quite makes it. For example, the closest we’ve come to extinguishing ourselves by nuclear war (on purpose anyway) is the Cuban Missile crises. While Kennedy doubtless deserves credit for resisting his advisor’s earlier unanimous recommendations for the use of nuclear weapons, some historians believe he prolonged the crises for political capital. There’s no hint of this controversy in Carlin’s ‘Kennedy-the-hero’ take on the subject.

We tend to act as if we are impervious to the sort of historical collapses explored here but that’s just what their citizens thought before they were ground to pulp in the gears of history. If there is a take-away from ‘End…’ it is, to quote Carlin’s dad: “don’t get cocky.”

The Irrational Ape

The Irrational Ape
by David Robert Grimes
Published by Simon & Schuster

Frankly, it’s not looking good for us apes in ‘The Irrational Ape’ by physicist David Robert Grimes. After a strong start courtesy of a couple of Russians you’ve never heard of who saved the world, things go downhill fast. One of the very many low points would be the mathematician Hippasus of Metapontum who, according to legend, was drowned as punishment for discovering irrational numbers. Not by an ignorant mob mind you, but by his fellow Pythagorians (yes, the triangle guy, also religious nut). If mathematicians can’t be logical, what hope do the rest of us have?

Grimes does his best to boost our chances. He starts off with a quick rundown of the steps needed to formulate a reasoned argument and the subtle – and not so subtle – twists of logic and retoric that give the veneer of creditability to hollow or emotive positions. Then he rounds up the usual suspects that scupper our thinking. There’s perennial favorites like echo chambers, confirmation bias and false dichotomies (the with-us-or-against-us ruse popular with demagogues from Lenin to Bush). False memory syndrome makes our memories more Wikipedia than Britannia and manipulated statistics show that dressing as a pirate alleviates Global Warming.

If the book feels familiar it’s not just because it reuses the gorilla hand graphic from Brian Cox’s ‘Human Universe’ for it’s cover. The absences and errors of logic that “hammer nonsense into narrative” have been covered by many previous writers. When I started reading I was unsure that we needed another reminder of the problems but the book provides plenty of motivation. Perhaps we need it because a study of 126,000 internet news stories between 2006 and 2017 showed that ‘falsehood diffused significantly farther, faster and more broadly than truth’. Perhaps it’s because bloggers who claim airplane air is dangerous since it’s not ‘pure oxygen’ can get to No.1 on the New York Times bestseller lists. Most pressingly it might be because the Internet’s promised democratisation of information has instead evolved, as Grimes describes it, into weaponised disinformation. ‘Outrage machines’ pander to the voluntary Balkanization of public discourse by generating pantomime villains & heroes and the comforting illusion of understanding (along with, of course, lots and lots of revenue producing clicks).

And, not least, perhaps we need it because Grimes is a clear and engaging writer with a lovely turn of phrase and personal experience countering anti-vaxers, conspiracists and climate deniers. He also fills the book with plenty of funny / disheartening / heartbreaking examples of tragic error, forehead-smacking-idiocy and outright chicanery to keep the reader entertained.

The only real problem with ‘The Irrational Ape’ then is that it’s destined for an echo chamber of it’s own. Any readers will already consider themselves pretty reasonable. They’re not, obviously. You’re not, I’m not. Even a cursory examination of our own motives reveals exactly the sort of glaring errors in logic and motivation Grimes describes. Books like ‘The Irrational Ape’ can help us get better at countering them but I suspect we’re the wrong market. Some of you reading this are teachers of all stripes from Montessori to tenured Professors. How difficult is it to develop a standard curriculum to teach the basics of how to think? Maybe some simple rules for primary students, funny examples to keep the attention of bored secondary classes, elementary statistics for college courses in business, finance and – especially – politics. The basics of how to reason (or, at least, recognise when others aren’t) should be as ubiquitous at all education levels as our ABC’s and 1,2,3’s. A fourth R to add to the other three. If we are to become even slightly more rational apes, it will have to start (and I apologise for laying this extra burden on your shoulders) with you.


by David A. Sinclair PhD with Matthew D. LaPlante
Published by Thorsons

Ray Kurzweil famously said that he wants to ‘live long enough to live forever’. In ‘Lifespan’, the goals of genetics and ageing researcher David Sinclair are rather more modest. He doesn’t promise us forever, he promises us Harriette Thompson.

Harriette Thompson looked 70 when she broke records at the 2014 San Diego marathon. She was 91. Harriette was a rare exception but she won’t stay that way for much longer Sinclaire says. His position gives him an inside track on the treatments and drugs coming down the pipeline that will extend life and healthspan. The future he sees for us is one where 90 year old marathon runners raise no more comment than 50 year old ones do today. A future where 90 year old teachers instruct 70 year old students and great, great grandparents rough-and-tumble-it with their great, great grandchildren. Its a future, he says, that is sitting on lab benches today.

‘Lifespan’ is full of acronmys; NR and NMN, SIR2 and NAD boosters, AMPK activators, TOR inhibitors and Metformin (a diabetic drug). All molocules, genes and drugs that work to restore vitality to the old and extend health and lifespan. Despite their tsunami of promise and Sinclair’s boundless enthuasism, the caveat is that most of the hard data is from experiments on yeast, fruitflys and mice. Trials with people are ongoing but results and treatments are years away so Sincalir isn’t waiting. He’s confident enough in the research as it stands to take a bunch of the stuff that works so well in animal studies. So does his dad, his brother, his co-workers and their families. Evidence is anacedotal of course but very positive.

On one level, ‘Lifespan’ is basically a funding pitch but Sinclair is up front about it. Ageing isn’t seen as a disease so research grants are hard to come by while doctors won’t prescribe, nor insurance companies cover, treatments. It’s hard to fault his logic. We spend billions on what he describes as ’whack-a-mole’ treatments and research for cancer, heart disease and Alzheimers while ignoring their overwhelming contributing factor; ageing, Being a smoker, he reminds us, increases your cancer risk fivefold but just being over 50 increases it a hundredfold.

Sinclair is aware that all these extra years come with consequences, challenges and risks – who needs an immortal Cyril Smith or Idi Amin? He also expects extending vitality will “fundamentally alter the way our economy works”. Overall though (and unsuprisingly) he’s confident that the personal, social and financial gains will outweigh any downsides.

I have to mention LePlante (the writing half of the duo). He may be over-fond of short, melodramatic sentenses but he makes a decent fist of translating complex research for the layman’s ear – not shying away from the science but not overloading us with it either.

Sinclair’s message is certainly an alluring one, especially for those of us with the cool winds of mortality whistling around our ankles but – for that very reason – it behoves us to make sure our critical faculties are fully engaged while listening to it. Sinclair is involved to varying degrees with commercial companies selling us the stuff he’s talking about. This, by no means, makes him a snake oil salesman but readers should allow for the possibility that it might motivate him – consciously or unconsciously – to ocasionally be more bullish than perhaps prudence might dictate. For instance, he doesn’t discuss the side affects of the Metformin drug I mention above and of which he has such high hopes. In practice it is often discontinued by diabetic patients due to the side affects of explosive flatulence and galloping diarrhoea. An extra ten years of vitality sounds great but less so if you have to spend nine of them in the bathroom.

Human Errors

Human Errors
by Nathan H. Lents
Published by Weidenfeld & Nicolson

Ah, man, that paragon of animals. In form and moving how express and admirable, whose eyes are in backward and whose sinuses are upside down. Nature loves a good joke but all too often we are the punch line. Our eyes are miracles of evolution certainly but by a quirk of early development the light sensing cells face the back of the socket (octopus eyes, with their independent evolution, are the right way round). Our smushed up faces have forced the drainage channel of our main sinus cavities to the top, leaving us with a legacy of blockages and infections our dogs never have to worry about.

So far we’re only on page 11 of ‘Human Errors’, where author Nathan H. Lents reminds us that “we are the long-term products of short-term actions.” In a further half dozen chapters covering anatomy, diet, genetics, reproduction, disease (warning: not suitable for hypochondriacs) and the brain, Lents outlines, in a genial if occasionally gallows-humourish style, some of the many errors and snafus that a myopic evolution has saddled us with.

Take that alphabet of vitamins we need to stay healthy. Becoming omnivorous a few million years ago was a useful survival trick but it took the evolutionary pressure off elsewhere. The GULO gene, for example, allows animals to make their own vitamin C. A few million years ago, a random mutation banjaxed the gene in our primate ancestor. In a forest full of vitamin C-rich fruit nobody noticed but start waking upright and heading out into the big, wide, fruit-free world and suddenly everybody gets Scurvy. Our ability to eat anything has turned into a requirement to eat everything.

Despite our dietary drawbacks we’ve spread ourselves around quite a bit so you’d think that reproducing would be something we’re good at. It turns out the wonder is not how well we reproduce but that we reproduce at all. Besides the obvious problem of our giant heads that made birth one of the leading causes of death for our ancestors, we suffer from a litany of sub-optimal reproductive design choices. That’s the way genetics works – a quick-fix solution for today with no thought for tomorrow.

Our genes, in general, are a bit of a bad news/good news situation. Good news: our cells reproduce with a 99.999% etc accuracy. Bad news: there’s around 100 quintillion chances a day to muck things up. Good news: most of those mistakes won’t hurt you. Bad news: that’s because most of your DNA isn’t actually you but a billion years of accumulated junk. Good news… well, actually, it gets hard to tell the good news from the bad here. All that DNA – dead viruses, duplicated duds and genes that just like to get up and go for a wander – might be bad for you individually but make the species mutable and adaptable.

The most successful of those adaptations is our big brain but Lent’s chapter title ‘A Species of Suckers’ is a clue that even here there are problems. One chapter doesn’t really do justice to the many ways our cerebral shortcuts bamboozle our rational selves but Lents manages to hit a lot of the high – or rather, low – points of our biases, delusions, illusions and irrationality. On the plus side, we’ve discovered monkeys are as bad at economics as we are (though that might be a bit of a low bar).

In the last section, musing on the future of humanity, Lents’ perky prose takes a slightly darker tone. Evolution hasn’t finished messing us about yet. We are on the cusp of being able to re-write our biology while at the same time being on a ‘collision course with our own industrialisation’. We have the tools to save ourselves but do we have the will? If not, Lents wonders, will that prove to be the ultimate human error?


The Intelligence Trap

The Intelligence Trap
by David Robson
Published by Hodder & Stoughton

“No ordinary man could be such a fool”
George Orwell

The thing about really smart people is that they are people first and really smart second. Sometimes a distant second. Starting with a decades long study of genius children and the doomed friendship between Harry Houdini and Arthur Conan Doyle, science journalist David Robson sets out to convince us that smart doesn’t always mean sensible. Honestly, it’s not a hard case to prove. He lines up studies showing nearly half of Mensa members believe in astrology or alien visitation, PhD’s make poor gamblers and people with high IQ’s are more likely to face financial distress. Overall, says Robson, the correlation between high IQ and rationality is very weak. We even have a word for it; dysrationalia – an inability (or disinclination) to think straight.

At the opposite end of the Dunning Kruger scale, the highly intelligent face a bunch of cognitive biases of their own and when clever people are wrong they can be wrong big time. Robson quotes a scary statistic of 80,000 deaths a year in US hospitals being due to misdiagnosis and tells the story of Brandon Mayfield, a US lawyer arrested for the Madrid Train bombings. Two experts mismatched his fingerprints and investigators even mistook his daughter’s Spanish homework for travel documents placing him at the scene. An inability to learn from mistakes is also prevalent. A dozen, nearly identical, close shaves preceded the Deepwater Horizon blowout as well as the Air France Concorde crash in 2000.

All of these cognitive flaws – and many, oh so many, more – are nothing new. Writers and philosophers have been lamenting them for millennia. Robson namechecks Scorates, Descartes and George Orwell and while I was reading on the train, Simon & Garfunkle chimed in with their own take; “a man hears what he wants to hear and disregards the rest” courtesy of my iPod’s playlist! But while the problems are ago-old, what is new is that science, rather than art, is being brought to bear on them. Research is rooting out how and why these errors occur. How often and under what exact circumstance are biases likely to show up? Spoiler: ‘everywhere’ and ‘all the time’. “The people who think they are already immune are probably most at risk”, says Robson. He owns up to his own shortsightedness and even in the midst of reading a book about cognitive biases I managed to galumph straight into an egregious example of my own (buy me a drink sometime and I’ll tell you all about it). All is not lost though. Robson spends the better half of the book recounting a variety of procedures and practices that various organisation have developed to counter bias effects.

In places he sounds like some New Age guru when he talks about ‘mindfulness’ and ‘evidence-based wisdom’ but bare with it, there’s plenty of practical advise too. Thinking a problem through in a second language mutes emotional responses. Doodling as you talk improves your memory. Jotting down your gut reaction and analysing evidence for it afterward improves accuracy. Arguing both sides of a problem – like a solo chess game – gives you a wider appreciation of opposing positions. Even just a little humility can go a long way to preventing errors – you’re probably right but what if, on this one occasion, you weren’t? On a wider scale, Robson explores anti-bias systems put in place by the US navy and Nuclear power industry that have massively reduced human error.

Our intelligence is broadly a matter of biology but, says Robson, how to avoid its drawbacks and use it to its greatest advantage is a wisdom we can, and should, all learn.

Empty Planet

Empty Planet
The Shock of Global Population Decline
by Darrell Bricker and John Ibbitson
Published by Robinson

Overpopulation has been the bugbear of civilisation from Malthus to Attenborough. Every few years a new milestone is reached – 6 billion, 7 billion, 8… UN forecasts have us on track for an 11 to 18 billion population by century’s end. How can the planet survive all those hungry mouths and carbon footprints? Well, say Darrell Bricker and John Ibbitson in ‘Empty Planet’, it won’t have to.

There are a lot of numbers in ‘Empty Planet’ but the most important one to remember is this: 2.1. That’s the number of children each woman must have on average during her lifetime to keep a population level. The authors haven’t just rounded up statistics though, they’ve done the footwork too by talking to demographic experts the world over and face-to-face interviews with focus groups on almost every continent. From immigrant workers in the US, Asian university students and Belgian young professionals to Delhi and Kenyan slum dwellers, – everywhere women have a choice, they choose to have fewer children. The authors are quite clear on the cause of this demographic shift. The economic miracles in the developing world that have lifted millions out of poverty also vacuum the young from the land into city life where children stop being hands to help and become mouths to feed.

It took Europe two centuries to go from 6 children per woman to under 2. It’s taken the developing world just two generations. Today, most of the planet has a birthrate below that magic 2.1 figure.  British women have an average of 1.8 children, slightly higher than the overall European figure of 1.6. Russia’s birthrate is 1.7. America has 1.9 babies per couple (0.95 for Millennials). Chinese women have 1.6 while Thailand, Vietnam and Malaysia are 2 or less. The fertility rate in large cities such as Hong Kong, Singapore, Beijing and Shanghai is below 1. The baby booming fuel that powered the population explosion has run out and, while our numbers are buoyed by all those old folk who stubbornly refuse to die (yet), the year of ‘peak child’ has come and gone. We’ll never see that 18 billion. Not even the 11 billion. The human race, the authors claim, will peak at 9 billion some 20 to 30 years from now and then decline.

Fewer people mean less pollution and cheaper housing sure but it’s not all wine and roses warn the authors. It also means fewer taxpayers to support a graying population plus fewer workers and consumers to drive the economy. They point to Japan as a present day example. “The diminishing working population is one of the biggest causes of long term recession”, says economist Naoyuki Yoshido. Japan today, Bricker and Ibbitson say, is everywhere else tomorrow. One solution, at least for a while, is immigration. ‘Empty Planet’ paints a near future world where countries scramble to attract immigrants rather than turn them away and a US that maintains it’s economic hegemony over an ageing East by dint of it’s appeal to hungry, young workers.

It’s hard to convince people that a book on population demographics is a gripping read (not even a single graph!) but it is. On every page the authors give us hope (and the figures to back it up) that the specter of a teaming, choking, starving world that’s haunted the popular imagination of generations will soon be banished. I could argue with some of their data (which might be editorial choices rather than error) but there’s so much data. Even if some of the information is flawed, the overwhelming direction is down. If Bricker & Ibbitson are correct, people reading this book (and this review) will live to see the human population decrease. Not through war or famine or disaster, but through choice.