Magical thinking and AI

This morning’s Observer column:

”Any sufficiently advanced technology,” wrote the sci-fi eminence grise Arthur C Clarke, “is indistinguishable from magic.” This quotation, endlessly recycled by tech boosters, is possibly the most pernicious utterance Clarke ever made because it encourages hypnotised wonderment and disables our critical faculties. For if something is “magic” then by definition it is inexplicable. There’s no point in asking questions about it; just accept it for what it is, lie back and suspend disbelief.

Currently, the technology that most attracts magical thinking is artificial intelligence (AI). Enthusiasts portray it as the most important thing since the invention of the wheel. Pessimists view it as an existential threat to humanity: the first “superintelligent” machine we build will be the beginning of the end for humankind; the only question thereafter will be whether smart machines will keep us as pets.

In both cases there seems to be an inverse correlation between the intensity of people’s convictions about AI and their actual knowledge of the technology…

Read on

Zuckerberg’s monster

This morning’s Observer column:

Who – or what – is Mark Zuckerberg? Obviously he’s the founder and CEO of Facebook, which is, in theory, a public company but is in fact his fiefdom, as a casual inspection of the company’s SEC filings confirms. They show that his ownership of the controlling shares means that he can do anything he likes, including selling the company against the wishes of all the other shareholders combined.

But the fact that Zuck wields autocratic power over a huge corporation doesn’t quite get the measure of him. A better metaphor is that he is the Dr Frankenstein de nos jours. Readers of Mary Shelley’s great 19th-century novel will know the story: of how an ingenious scientist – Dr Victor Frankenstein – creates a grotesque but sentient creature in an unorthodox scientific experiment. Repulsed by the monster he has made, Frankenstein flees, but finds that he cannot escape his creation. In the end, Frankenstein dies of exposure in the Arctic, pursuing the monster who has murdered his bride. We never learn what happened to the creature.

Facebook is Zuckerberg’s monster. Unlike Frankenstein, he is still enamoured of his creation, which has made him richer than Croesus and the undisputed ruler of an empire of 2.2 billion users. It has also given him a great deal of power, together with the responsibilities that go with it. But it’s becoming increasingly clear that his creature is out of control, that he’s uneasy about the power and has few good ideas about how to discharge his responsibilities…

Read on

Asymmetrical vulnerabilities

This morning’s Observer column:

In their book, The Future of Violence, Benjamin Wittes and Gabriella Blum point out that one of the things that made the Roman empire so powerful was its amazing network of paved roads. This network made it easy to move armies relatively quickly. But it also made it possible to move goods around, too, and so Roman logistics were more efficient and dependable than anything that had gone before. Had Jeff Bezos been around in AD125, he would have been the consummate road hog. But in the end, this feature turned out to be also a bug, for when the tide of history began to turn against the empire, those terrific roads were used by the Goths to attack and destroy it.

In a remarkable new paper, Jack Goldsmith and Stuart Russell point out that there’s a lesson here for us. “The internet and related digital systems that the United States did so much to create,” they write, “have effectuated and symbolised US military, economic and cultural power for decades.” But this raises an uncomfortable question: in the long view of history, will these systems, like the Roman empire’s roads, come to be seen as a platform that accelerated US decline?

I think the answer to their question is yes…

Read on

Hofstadter’s Law and self-driving cars

This morning’s Observer column:

In 1979, Douglas Hofstadter, an American cognitive scientist, formulated a useful general rule that applies to all complex tasks. Hofstadter’s law says that “It always takes longer than you expect, even when you take into account Hofstadter’s law”. It may not have the epistemological status of Newton’s first law, but it is “good enough for government work”, as the celebrated computer scientist Roger Needham used to say.

Faced with this assertion, readers of Wired magazine, visitors to Gizmodo or followers of Rory Cellan-Jones, the BBC’s sainted technology correspondent, will retort that while Hofstadter’s law may apply to mundane activities such as building a third runway at Heathrow, it most definitely does not apply to digital technology, where miracles are routinely delivered at the speed of light…

Read on

Google, Facebook and the power to nudge users

This morning’s Observer column:

Thaler and Sunstein describe their philosophy as “libertarian paternalism”. What it involves is a design approach known as “choice architecture” and in particular controlling the default settings at any point where a person has to make a decision.

Funnily enough, this is something that the tech industry has known for decades. In the mid-1990s, for example, Microsoft – which had belatedly realised the significance of the web – set out to destroy Netscape, the first company to create a proper web browser. Microsoft did this by installing its own browser – Internet Explorer – on every copy of the Windows operating system. Users were free to install Netscape, of course, but Microsoft relied on the fact that very few people ever change default settings. For this abuse of its monopoly power, Microsoft was landed with an antitrust suit that nearly resulted in its breakup. But it did succeed in destroying Netscape.

When the EU introduced its General Data Protection Regulation (GDPR) – which seeks to give internet users significant control over uses of their personal data – many of us wondered how data-vampires like Google and Facebook would deal with the implicit threat to their core businesses. Now that the regulation is in force, we’re beginning to find out: they’re using choice architecture to make it as difficult as possible for users to do what is best for them while making it easy to do what is good for the companies.

We know this courtesy of a very useful 43-page report just out from the Norwegian Consumer Council, an organisation funded by the Norwegian government…

Read on

Zuckerberg’s monster

My Observer review of Siva Vaidhyanathan’s Anti-social Media: How Facebook Disconnects Us and Undermines Democracy:

The best metaphor for Facebook is the monster created by Dr Frankenstein. Mary Shelley’s story shows how, as Fiona Sampson put it in a recent Guardian article, “aspiration and progress are indistinguishable from hubris – until something goes wrong, when suddenly we see all too clearly what was reasonable endeavour and what overreaching”. There are clear echoes of this in the evolution of Facebook. “It’s a story”, writes Siva Vaidhyanathan in this excellent critique, “of the hubris of good intentions, a missionary spirit and an ideology that sees computer code as the universal solvent for all human problems. And it’s an indictment of how social media has fostered the deterioration of democratic and intellectual culture around the world.”

Facebook was founded by an undergraduate with good intentions but little understanding of human nature. He thought that by creating a machine for “connecting” people he might do some good for the world while also making himself some money. He wound up creating a corporate monster that is failing spectacularly at the former but succeeding brilliantly at the latter. Facebook is undermining democracy at the same time as it is making Mark Zuckerberg richer than Croesus. And it is now clear that this monster, like Dr Frankenstein’s, is beyond its creator’s control…

Read on

Why surveillance techology is usually better than we realise

This morning’s Observer column:

The images of the moon’s surface coming down from the orbiters were of astonishingly high resolution, good enough to blow up to 40ftx54ft pictures. When Nasa engineers initially stitched the images together they had to hang them in a church to view them. Eventually, they found a hangar where they could be laid on the ground for astronauts to walk on them in stockinged feet in order to search for suitable landing sites. Sign up for Lab Notes – the Guardian’s weekly science update Read more

For decades, nobody outside of Nasa and the US military knew how good these images were. The few that were released for public consumption were heavily degraded and fuzzy. Why? Because the cameras used in the lunar orbiters were derivatives of the cameras used in high-altitude US aerial reconnaissance planes and satellites and the Pentagon didn’t want the Soviets to know the level of detail that could be derived from them.

In a way, we shouldn’t be surprised by this revelation. It’s an old story: powerful states have often possessed more sophisticated surveillance technology than their adversaries – or their citizens – knew or suspected…

Read on

Censorship 2.0

This morning’s Observer column:

One of the axioms of the early internet was an observation made by John Gilmore, a libertarian geek who was one of the founders of the Electronic Frontier Foundation. “The internet,” said Gilmore, “interprets censorship as damage and routes around it.” To lay people this was probably unintelligible, but it spoke eloquently to geeks, to whom it meant that the architecture of the network would make it impossible to censor it. A forbidden message would always find a route through to its destination.

Gilmore’s adage became a key part of the techno-utopian creed in the 1980s and early 1990s. It suggested that neither the state nor the corporate world would be able to censor cyberspace. The unmistakable inference was that the internet posed an existential threat to authoritarian regimes, for whom control of information is an essential requirement for holding on to power.

In the analogue world, censorship was relatively straightforward…

Read on

Kremlinology 2.0

This morning’s Observer column:

In the bad old days of the cold war, western political and journalistic institutions practised an arcane pseudoscience called Kremlinology. Its goal was to try to infer what was going on in the collective mind of the Soviet Politburo. Its method was obsessively to note everything that could be publicly observed of the activities of this secretive cabal – who was sitting next to whom at the podium; which foreign visitors were granted an audience with which high official; who was in the receiving line for a visiting head of state; what editorials in Pravda (the official Communist party newspaper) might mean; and so on.

The Soviet empire is no more, much to Putin’s chagrin, but the world now has some new superpowers. We call them tech companies. Each periodically stages a major public event at which its leaders emerge from their executive suites to convey messages to their faithful followers and to the wider world. In the past few weeks, two such events have been held by two of the biggest powers – Google and Apple. So let’s do some Kremlinology on them…

Read on

Tech-driven wealth is the new aphrodisiac

This morning’s Observer column:

It’s a quintessential Silicon Valley story. A smart, attractive 19-year-old American woman who has taught herself Mandarin while in high school is studying chemical engineering at Stanford, where she is a president’s scholar. Her name is Elizabeth Holmes. In her first year as an undergraduate she persuades her professor to allow her to attend the seminars he runs with his PhD students. Then one day she drops into his office to tell him that she’s dropping out of college because she has a “big idea” and wants to found a company that will revolutionise a huge part of the healthcare system – the market for blood testing services. Her company will be called Theranos.

Holmes’s big idea was for a way to perform multiple tests at once on a tiny drop of blood, and to deliver the results wirelessly to doctors. So she set about pitching to investors…

Read on