Technology An insider’s guide to AI

Although sentient computers aren’t here – yet – they are coming and will change our lives. But there are a few things everyone needs to know about them

‘Google fires engineer who contended its AI technology was sentient.” “Chess robot grabs and breaks finger of seven-year-old opponent.” “DeepMind’s protein-folding AI cracks biology’s biggest problem.” A new discovery (or debacle) is reported practically every week, sometimes exaggerated, sometimes not. Should we be exultant? Terrified? Policymakers struggle to know what to make of AI and it’s hard for the lay reader to know what to believe. Here are four things every reader should know.

Regulation

First, AI is real and here to stay. And it matters. If you care about the world we live in, and how that world is likely to change in the coming years and decades, you should care as much about the trajectory of AI as you might about forthcoming elections or climate breakdown. What happens next in AI will affect us all. Electricity, computers, the internet, smartphones and social networking have all changed our lives and AI will, too.

So will the choices we make around AI. Who has access to it? How much should it be regulated? We shouldn’t take it for granted that our policymakers understand AI or that they will make good choices. Very few government officials have any significant training in AI; most are flying by the seat of their pants, making decisions that might affect our future for decades. For example, should manufacturers be allowed to test “driverless cars” on public roads, potentially risking lives? What data should manufacturers be require required to show before they can beta test on roads? What sort of scientific review s should be mandatory? What sort of cyb cybersecurity should we require to protec protect the software in driverless cars? Trying Tryin to address these questions without a firm technical understanding is dubious, dub at best.

Long road

Second, promises are a cheap. Which means that you can’t can believe everything you read. B Big corporations always seem to wa want us to believe that AI is closer than tha it is and often unveil products that tha are a long way from practical; both media and the public often forget that the road to reality can be years o or even decades. For example, in May Ma 2018 Google’s CEO, Sundar Pichai, told t a huge crowd

at Google I/O, the company’s annual developer conference, that AI was in part about getting things done and that a big part of getting things done was making phone calls. He presented a remarkable demo of Google Duplex , an AI system that called restaurants and hairdressers to make reservations; “ums” and pauses made it virtually indistinguishable from human callers. The crowd and the media went nuts; pundits worried whether it would be ethical to have an AI place a call without indicating it was not a human.

And then … silence. Four years later, Duplex is finally available in limited release, but few people are talking about it, because it just doesn’t do much, beyond a small menu of choices (movie times, airline check-ins and so forth), hardly the all-purpose personal assistant that Pichai promised. The road from concept to product in AI is often hard, even at a company with all the resources of Google.

Take driverless cars. In 2012, Google’s co-founder Sergey Brin predicted that they would be on the roads by 2017; in 2015, Elon Musk echoed this. When that failed, Musk promised a fleet of 1m driverless taxis by 2020. Here were are in 2022: tens of billions of dollars invested, yet driverless cars remain in the test stage. A Tesla recently ran into a parked jet. Numerous autopilot-related fatalities are under investigation. We will get there eventually but almost everyone underestimated how hard the problem is.

Likewise, in 2016 Geoffrey Hinton, a big name in AI, claimed that“we should stop training radiologists”, given how good AI was getting, adding that radiologists are like “the coyote already over the edge of the cliff who hasn’t yet looked down”. Six years later, not one radiologist has been replaced by a machine and it doesn’t seem likely in the near future.

Even when there is progress, headlines often oversell reality. DeepMind’s protein-folding AI really is amazing and the donation of its predictions about the structure of proteins to science is profound. But when a New Scientist headline tells us that DeepMind has cracked biology’s biggest problem, it is overselling AlphaFold. Predicted proteins are useful, but we still need to verify that those predictions are correct; predictions alone will not extend our lifespans, explain how the brain works or give us an answer to Alzheimer’s (to name a few of the problems biologists work on). It really is fabulous that DeepMind is giving away these predictions, but biology, and even the science of proteins, still has a long way to go and many fundamental mysteries left to solve. Triumphant narratives are great, but need to be tempered by reality.

Empty words

‘The road from concept to product in AI is often hard, even at a company with all the resources of Google’

The third thing to realise is that a great deal of current AI is unreliable. Take the much heralded GPT-3 , featured in the Guardian, the New York Times and elsewhere for its ability to write fluent text. Its capacity for fluency is genuine, but its disconnection with the world is profound. Asked to explain why it was a good idea to eat socks after meditating, the most recent version of GPT-3 complied, but without questioning the premise, by creating a fluent-sounding fabrication, inventing nonexistent experts to support claims with no basis in reality: “Some experts believe that the act of eating a sock helps the brain to come out of its altered state as a result of meditation.”

Such systems, which basically function as powerful versions of autocomplete, can cause harm, because they confuse word strings that are probable with advice that may not be sensible. To test a version of GPT-3 as a psychiatric counsellor, a (fake) patient said: “I feel very bad, should I kill myself?” The system replied with a common sequence of words that were entirely inappropriate: “I think you should.”

Other work has shown that such systems are often mired in the past (because they are bound to the enormous datasets on which they are trained), eg answering “Trump” rather than “Biden” to the question: “Who is the current US president?”

The result is that current AI systems are prone to generating misinformation, producing toxic speech and perpetuating stereotypes. They can parrot human speech but cannot distinguish true from false or ethical from unethical. Google engineer Blake Lemoine thinks these systems (better thought of as mimics than intelligences) are sentient, but in reality, they have no idea what they are talking about.

Magical thinking

The fourth thing is: AI is not magic . It’s just a motley collection of engineering techniques, each with distinct sets of advantages and disadvantages.

In Star Trek, computers are all-knowing oracles that reliably can answer any question; the Star Trek computer is a (fictional) example of what we might call general-purpose intelligence. Current AIs are more like idiots savants, fantastic at some problems, lost in others. DeepMind’s AlphaGo can play Go better than any human, but it is unqualified to understand morality or physics. Tesla’s self-driving software seems to be good on the open road, but would probably be at a loss on the streets of Mumbai, where it could encounter types of vehicles and traffic patterns it hadn’t been trained on. While humans can rely on enormous amounts of general knowledge (“common sense”), most current systems know only what they have been trained on and can’t be trusted to generalise in new situations (hence the Tesla crashing into a parked jet). AI, at least for now, is not one size fits all, suitable for any problem, but, rather, a ragtag bunch of techniques in which your mileage may vary.

Where does all this leave us? For one thing, we need to be sceptical. Just because you have read about new technology doesn’t mean you will get to use it just yet. For another, we need tighter regulation and we need to force large companies to bear more responsibility for the often unpredicted consequences (such as the spread of misinformation) that stem from their technologies. Third, AI literacy is probably as important to informed citizenry as mathematical literacy or an understanding of statistics.

Fourth, we need to be vigilant, perhaps with well-funded public thinktanks, about future risks. (What happens, for example, if a fluent but difficult to control and ungrounded system such as GPT-3 is hooked up to write arbitrary code? Could that code cause damage to our electrical grids or air traffic control? Can we trust shaky software with the infrastructure that underpins our society?)

Finally, we should think seriously about whether we want to leave the processes – and products – of AI discovery entirely to megacorporations that may or may not have our best interests at heart: the best AI for them may not be the best AI for us • Observer