Seeing through the hype and reality of artificial intelligence

The 1968 movie foresaw threatening general AI in 2001

Last week a group called the “Future of Life Institute” published an open letter urging governments to pause research on artificial intelligence so that a new worldwide regulatory framework can be agreed to prevent the technology taking a highly destructive route. There are quite a few distinguished signatories, of whom the most commented upon is the technology entrepreneur Elon Musk. On the same day the British government published a paper on AI strategy suggesting a minimum of regulation in order for this country to gain a technical edge. The first development bespeaks fear and panic amongst intellectuals, and the second the political reality of countries wanting to win the race. Is the world going to hell in a handcart?

The text of the letter is here. The core of it is this series of fears:

Contemporary AI systems are now becoming human-competitive at general tasks, and we must ask ourselves: Should we let machines flood our information channels with propaganda and untruth? Should we automate away all the jobs, including the fulfilling ones? Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us? Should we risk loss of control of our civilization? 

from “Pause Giant AI Experiments: an Open Letter” – Future of Life Institute

Now let’s take a deep breath. AI is important and will have a profound effect on human life. recent progress has been dramatic, and its most recent capabilities have astonished. But its impact is not necessarily benign, and there are some serious risks attached to its development. But to describe current AI capabilities as a “mind” and to suggest that it is on the road to replacing humanity in general understanding, judgement and control is to misunderstand what it is – it is classic anthropomorphism of the sort that imagines your cat to be a scheming villain. If you want to unpack this a little, a good place to go is this blog piece by Melanie Mitchell an AI academic. The interesting question is why so many intelligent people are being let astray by the hype.

For a long time, people have dreamed of building intelligence that can replicate the autonomy, command and ability to learn new tasks in such a way that any human rôle can be replaced by it. This is referred to as “general AI”. Once you leave behind the understanding that humans, and indeed animals, are animated by some form of supernatural spirit, you are left with the logical possibility of humans building something lifelike in all its capabilities. It’s just atoms and molecules after all. And of course, if we can do that, then we can make robots stronger and better than the original because we can engineer it that way. This has particularly appealed to the military, who can develop robot-powered weapon systems to replace frail soldiers, sailors and airmen, and doubtless spacemen too. And such is the confidence of modern humans, that it is widely assumed that doing so is not so hard. During the Cold War both America and the Soviet Union worried that their opponents were close to developing just such a capability, doubtless promoted by people in search of funding. Such is the grasp of military “intelligence”.

And people continue to believe that such an ability is just around the corner. In their film and novel 2001, a Space Odyssey, published in the 1960s the author Arthur C Clarke and fim-maker Stanley Kubrick speculated that out-of-control general AI would be developed by 2001. In the 2010s it was widely assumed that self-driving cars would be on the road by the early 2020s. But in 2023, instead of mounting excitement about its imminent rollout, there a silence. And now generative AI, which can manufacture very human-sounding bullshit from minimal instructions, is sparking this panic. Each wave of AI generates astonishing advances, and then seems to stop at the bits its promoters assumed were mere details. Researchers and technologists have a strong incentive to hype their latest achievements, and one of the most effective ways of doing this is to claim that you are close to bridging the gap to general AI. The people with the money often have little idea of what is or is not technically possible – not that the developers often have much idea themselves – and investing in the first successful general AI project sounds like a good deal. Just how hard can it be, after all? General AI occurs naturally.

Behind this lies a striking failure to understand what humans are. It is assumed that humans are naturally-occurring robots – machines created by an intelligence to fulfil particular tasks. Back in the days when I followed evangelical Christians there was a popular idea called “intelligent design”. This posited that the world, and animals and people in particular, were far too clever and complicated to have been created by dumb processes like natural selection, so they must have been designed by an intelligence – God. It seems that most people instinctively believe this, even if they disregard the divine revelation of the Abrahamic religions. We create god in our own image. Evolution is often described as if it were an intelligent process; DNA is described as a blueprint; and so on. Man has created God in his own image, and assumed he (well, maybe she) would create things in the way that men would. But the most important thing to understand about humans and the world is that there is no intelligent creator; there is no thread of intent; no design. It just happened. Over a very long time. It follows that trying to reproduce human intelligence through a process of intelligent design is going to be at best slow and frustrating, and at worst not feasible. Indeed a lot of breakthroughs in AI design, such as neural networks, are based on the idea that the thing should build itself, rather than be designed. But that has limits. Development is not going to be an orderly progression of achievements.

And indeed, each phase of AI development comes with inherent limitations. Generative AI, the current craze, requires massive computer processing power, and a huge database of knowledge in digital form. It is easy to see how this might be useful; harder to see how it leads to general AI. The battle over modern AI seems to revolve around training data and computer firepower. One of China’s advantages is that it has access to cheap labour to produce training databases. There is a paradox there. The idea is that in due course that the robots will find and produce their own training data – but problems abound. Indeed for AI to be useful (including self-driving cars) I think it will be necessary to simplify the the world so that it can be represented more accurately in training data. But that makes it of limited use in the wild.

And so the hype cycle goes on. At each turn we will be distracted by the prospect of general AI rather than tackle the more important issues that each iteration throws up. In the case of generative AI it is how to deal with the misinformation and prejudice lurking in the training data, that the technology won’t be able to recognise – when its “reasoning” is so opaque that it can be hard to spot what is causing the problem.

And what of the British government’s AI strategy? I haven’t read it so I can’t comment with authority. Generally I am suspicious of politicians jumping on bandwagons – and any call to invest in this or that technology because otherwise we will be left behind in the global race is suspect. Economic advantage usually accrues either from developing things in areas where people are not already heavily invested, or from copycatting after somebody else has done the difficult bits. But in seeking to develop a light-touch regulatory advantage while the European Union and the US are tangled in moral panic, and China over fears of loss of party control, it might be onto something – though it is hard to see that the US will in the end be weighed down by over-regulation. On the other hand the US and China have advantages in access to finance, computing power, and data – while the multilingualism of the EU perhaps offers an advantage to them. Britain may simply become a base for developing intellectual property owns by others.

Anyway, a it is hard to see what a six-month pause could possibly achieve. Humanity may reach general AI in due course. But not for a while yet.

The BBC’s questionable coverage of the US elections

Photo: USCapitol, Public domain, via Wikimedia Commons

The British political class is obsessed with US politics. The right wish to emulate US culture wars and attack “wokeism”. The left (rather more successfully) pick up on the Black Lives Matter movement, never mind our very different racial history. So perhaps it was inevitable that the BBC would cover the US mid-term elections so heavily. But in doing so they have posed deeper issues about the institution.

I try to resist the temptation to comment on other countries’ politics. I last commented directly on US politics in 2021, when I praised the progress being made by President Joe Biden. That quickly became a very unfashionable view, especially after the Afghanistan debacle. I still think he is underestimated – but then I thought that of Jimmy Carter when he was president – another deeply unfashionable view. But want I want to look at this time is not US politics itself, but how the BBC in particular covered it.

News coverage of the US midterm elections was extensive, and especially on the BBC – my leading source of daily news. A Danish general election came and went without comment – in a country with a close cultural affinity with England, if the not the rest of Britain – while reporting on the 10 O’clock TV News from America was almost daily.

The first striking thing abut this coverage was the BBC’s claim that these elections were hugely consequential, a view widely repeated in Britain. They explained how if the Democrats lost control over Congress, then the presidential agenda would be halted. And yet this usually happens in the US midterms, and life goes on. Mr Biden has fought hard to get as much of his legislative agenda as possible achieved before these elections; he always expected his party would lose them. BBC correspondents eventually seemed to realise that this argument wasn’t strong enough to support the trope that these were the most important midterms ever (how sick I get when this claim is made of elections, as it always is). So they said that democracy itself was at stake. The basis of this claim was the number of Republican supporters of Donald Trump, who claimed that the presidential election of 2020 was stolen by Mr Biden. And some of them sought to take control over state electoral processes, so that they could ensure the right outcome next time. This is indeed an interesting aspect of current US politics. But were the BBC following the Democrats’ line a bit too uncritically? I don’t mind their reporters repeatedly saying that Mr Trump’s claims are false – his supporters have been unable to produce evidence that can be tested in court – perjury is a serious matter, after all. But the endangerment of the electoral process in some states is only one issue that American voters have to weigh up.

But this marks a more general Democrat bias to the BBC coverage – a prejudice which, to be fair, is pretty prevalent in Britain. They are careful to present Republican views in their interviews with politicians and in vox pops. But we are left thinking that most Republican supporters are nutters – repeating those election fraud allegations, for example. If this were all there was to it, then it raises the question of why the Republicans are doing so well. Surely Americans aren’t all that stupid? There is clearly something else going on. And that something else seems to be the perceived extremism of many Democrats, which clearly annoys many Americans, and not just less-educated white ones. We will struggle to get any idea of what lies behind this fear from watching BBC coverage – or I suspect that of most other British news outlets. In the end the story of American conspiracy theorists, and the ramblings of Mr Trump, are the more entertaining story, and that dominates our coverage.

Funnily enough, after all the hype of the run-up, coverage of the election results has been very muted. They were almost exactly in line with the more considered predictions (published by The Economist, for example), but probably a bit worse for the Republicans in the House of Representatives. But they still won there – and there will be significant consequences arising from that. Mr Trump seems to have had a measurable negative impact on his party’s performance, but I haven’t seen that story through on the mainstream reporting. We’re now onto the World Cup and COP27. And there is quite a bit of domestic news to be digested too.

All of this is making me think harder about the role of the BBC in news. It remains a trusted brand, and we are lucky to have it in place of the more partisan free for all that dominates the US. But, as a public service organisation committed to balanced coverage, it faces tricky decisions. Not so long ago it got into trouble by “balancing” scientists warning us about carbon emissions with unqualified s**t-stirrers like Nigel Lawson. They have moved on from that, and it is good to see that they are not giving credibility to Mr Trump’s lies about the 2020 election in the name of balance. But, in general, the BBC feels it must follow the news agenda set by others, in which Britain’s diminishing print media have an outsize role. Since this media is dominated by organisations with a right-wing agenda, this causes much wailing and gnashing of teeth amongst people in the left. But there are bigger problems.

The first is the choice of which stories to cover, or not – what makes an important story in the news media is not necessarily all that important in the great scheme of things. We saw this in the coverage of the death of the Queen. This was a big news story, of course, but the BBC followed other media in excluding all other news reporting in the weeks after, until other media outlets started to let other stories in. The BBC wouldn’t devote so much as a minute to “other news” in extended news broadcasts, and instead went round in circles in with interviews with royal correspondents, there being little actual news to report. The BBC could and should have shown a bit of leadership here – five minutes of other news in a 60-minute programme was surely not have been too brave? The time given to US politics is another example, with European politics being neglected by comparison – in spite of the fact that the latter might have a more immediate impact on the country. That Danish general election was quite an interesting story. And then there is the bias towards covering elections before they happen, and neglecting the hard news of the actual results. Even The Economist does this, though. That is the difference between journalism and “the first draft of history”, I guess. Speculation is more fun than facts.

But what about the accusation of liberal bias that is so often levelled at the BBC – and which its coverage of the US election seems to illustrate? This is more complicated than it looks. Another example is racism and antisemitism. A few on the right might grumble about the apparently uncompromising stand that the BBC takes on these issues. But dig down a bit and you find problems. Racists and antisemites are portrayed as nutters that it is easy to dis-associate from. I have seen a couple of television dramas portraying far-right activism, along with its racist and antisemitic tropes. But these activists are cardboard cut-outs – poorly educated white people with a soft spot for Hitler and Naziism. There is much more to racism and antisemitism than this. Will there ever be a drama covering antisemitism in the far left? To say nothing of the muddle between antisemitism and criticism of Israel and Zionism? Of course not. It is too controversial. But this bias annoys both left and right. The right is annoyed by the persistent portrayal of racism as being confined tot he political right. The left complains that discussion of Israel is heavily constrained.

This is all rather depressing. There seem to be two types of news media: the partisan and the dumbed-down. The partisan media thrives on controversy and isn’t afraid to air conspiracy theories and nonsense tropes – and is consequently useless as a source of information. But the “balanced” media are too scared that controversy will damage their reputation for objectivity – and anyway tend to follow the pack in their coverage of stories. Perhaps there is an inevitability about this – but at least the BBC could try a bit harder to be more informative in its news coverage.

What does the new data protection law say about British politics?

A new fear is quietly stalking the land, haunting administrators of businesses, charities, schools and even political parties. It is called the EU General Data Protection Regulation 2016 (GPDR). It replaces our old friend the Data Protection Act 1998 with a set of regulations that are much more onerous. The usual experts are going back and forth gleefully stoking up fear, no doubt in the hope of consultancy fees and demand for training courses. But even allowing for their exaggerations, the change is significant. The politicians, meanwhile, are nowhere to be seen.

This is information politics in action. Loyal readers of this blog will know that I think that information politics should be, and will be, one of the most critical areas of political engagement, notwithstanding its current neglect. The political issues raised by this legislation illustrate that rather well – and the failure of the British political establishment.

To simplify gloriously, there seem to be three main approaches to data regulation: the American, the European and the Chinese. The Americans are giving general freedom to businesses, leaving privacy largely a matter for civil law to be resolved between businesses and citizens. They do, however, want to place limits on government abuse; many Americans are more worried about Uncle Sam prying on them than Google. The Chinese believe in complete government control, with no right to individual privacy. The Europeans believe in strong privacy rights backed up by criminal sanctions, and severe constraints on government agencies too. The GPDR, as its full title suggests, is very much in the European tradition. The purpose of the the extra regulations is to enhance individual rights to privacy, with rights to rectification, erasure and access. So citizens have a right to know what data is held on them, to correct errors and to be forgotten if they want. That means that the organisations have to know themselves how their data relates to individuals, and to make corrections and deletions accordingly. The new rules are altogether more thorough than their predecessors. They cover all data, not just electronic databases; political parties, which only caught the fringes of previous law, face much more onerous requirements, as do many charities.

What are the implications? It will put a lot stress on small organisations, something the British civil service will shed no tears about; they have always preferred to deal with a small number of big agencies. But some bigger businesses are going to find the going harder too – notably Google and Facebook. Mostly organisations will manage the risk by holding the minimum possible amount of data on their own behalf, digital and otherwise. Ironically this may be no bad thing for efficiency. Efficient people and organisations travel light. It’s an old trick of personal organisation: if you destroy everything, then you don’t waste time looking for things. Clever service businesses should be able to design inexpensive support systems that allow small organisations to comply with the regulations, once they have got into the habit of holding no paper records and regularly purging the digital ones. But until organisations realise that this is the new way, there is going to be a bumpy ride. The law may turn out to be almost impossible to comply with – and feel a bit like one of those Russian laws that are intentionally impossible, to give state agencies more arbitrary power.

But surely Brexit will come to the rescue? If there was an example of onerous European regulation that we can be freed from, then this must surely be it. Why can’t we now move to a more light-touch American regime? Alas no; the British government have made it very plain that this law is built to last after the country leaves the EU. Indeed, I understand, the law has been gold-plated to make it more onerous than the European standard. For what reason I’m not entirely sure; the government just seems to think it is a good idea.

Which it may be. These new rights do empower the citizen. Once explained to the public, they might very well like the new law. They would certainly not think that political parties, for example, should be given a free pass, and the rights to access and rectification look basic. The American way, where big businesses have excessive sway, is not necessarily the best. But it is a political choice, and there has been next to no political debate; if there was any, I missed it. This says a lot about how British politics works. A European regulatory proposal comes along; British officials decide whether they like the idea or not, and negotiate with other EU interests accordingly. They then present it to the British parliament as a fait-accompli, and promptly embellish it. And then it gets dumped on the British public with a shrug. It is no wonder that so many intelligent people became fed up with the EU. The British political establishment is using it as a way to bypass awkward political discussion; no doubt this happens in other European countries too. It is a colossal failure of the political class, but in a long British tradition. British institutions have long thought that secrecy over decision-making ensured its integrity.

Why wasn’t there more political debate? This could, or should, have happened at two distinct stages. The first was when the directive was being put together at EU level. The British government clearly had opportunities to intervene if it wanted to – and probably did, but with the minimum of consultation with its own people. And failing that there was the European Parliament. These institutions failed. Brexiteers will suggest that this is an example of arbitrary Brussels lawmaking; Remainers that it is a failure of the British political class to exercise their responsibilities properly. The second possible intervention was when the directive was translated into British law, when Parliament had a chance to scrutinise the proposals. If the directive was indeed gold-plated, then this would have been the appropriate moment to challenge it. But neither the popularly elected commons, nor the supposedly hard-working and expert Lords seem to have done very much.

Behind this there is a deeper failure. Who are the advocates of a different approach, easier for small organisations, profit or non-profit, to manage? Labour aren’t instinctively for enhancing individual rights, they aren’t very interested in making life easier for businesses either. Some on the left probably hanker after a more Chinese model of data regulation – but that is only hinted at in some dark statements on cracking down on tax evasion. The Lib Dems are not inclined to challenge European integration, which GPDR is part of, and anyway probably quite like the enhanced individual rights, in principle anyway. But you would have expected some resistance coming from the Conservative Party.

Alas no. The radical Brexiteers aren’t interested in detail. To them deregulation is a theoretical idea where somebody else has to do all the hard work subject to their backseat driving. The pragmatists are happy enough to go along with European integration. I have heard a bit of talk that there is fresh new thinking in the party, led by a crop of bright new MPs recruited in David Cameron’s tenure. If so you might expect that somebody would make the running and present a vocal challenge to the new regulations, and an alternative vision on how data regulation should work. But so far there is silence.

The truth seems to be that few British politicians have thought deeply about how me manage privacy and data, and therefore recognise the nature of the choices they are making. That is very disappointing.

Information technology is enslaving us: we must learn to master it

I have a new hobby horse: the politics of information. The development of information technology is transforming our lives, but the politicians are being left behind. This is becoming at least as important as economics and finance to the way we live our lives.

But we amateurs face a problem. The IT industry obfuscates everything in jargon and tech-speak. It is easy to get intimidated. In the FT Gillian Tett draws a parallel with the finance industry before the great financial crisis of 2007-08. It is liable to end just as badly. We really must try to hack back the thicket.

I’ve been here before. Back in the 1990s I was appointed Director of Information Systems by my firm, in spite of having no direct IT background. I was nearly suffocated by the jargon and tech-speak. But gradually I came to realise that IT wasn’t as complicated as people were making it out to be. In fact it wasn’t fundamentally more complex than the average office filing system in pre IT days, and it was dealing with much the same issues. If you stuck to firm logical ground, the techies would retreat. I found a world dominated by bluff and which reasoned in a series of attractive sounding non-sequiturs. People were watching each other and saying whatever they had to to fit in. With just a little clarity of thought you could get a long way.

So I will take a deep breath and start to think about the world of information systems and technology, even though it is a very different one from where I left it in the 1990s, or even in 2005, when I stopped working with information systems professionally as a user.

And so to the basics. How do we, as people, manage information? We do two basic things. The first is to gather data from the world around us. The second is to process that data into information that we can use to achieve goals both passive (looking out for danger) and active (finding food, say). All this requires us to be both aware and focused – two things that tend to be mutually exclusive. How humans (and other animals) do this is a very complex process that is only very loosely understood by scientists. The interesting thing is that at its core is a duality – the outwardly referenced right brain, and the inwardly focused left brain. I am currently reading Iain McGilchrist’s The Master and his Emissary which builds a substantial intellectual critique of the modern world from a right-brain left-brain duality. At this stage, though, all I want to say is that the outward/inward duality is central to the understanding of how we deal with information.

This duality is recognisable in the way modern technology works. Here I think it is useful to distinguish between what I would call “big data” and “useful data”. Big data is the amassing of data from many sources. In the modern age this is often from such things as video footage, photographs and sound recordings. But big data is not directly usable to achieve anything. To do that it has to be reduced to patterns and digits that are useful data. The big modern development is the use of artificial intelligence (AI) to achieve this. Previously, useful data was mainly gathered through human input.

There is one key point that needs to be understood about the useful data sitting in computer databases. It works on the principle of distillation. It is an infinitesimal subset of the real world, and even that is before you deal with problems of reflecting time. To move from the real world to this data requires a series of simplifying judgements.In practice this means that data does not multi-task well. To be efficient the data has to be referenced to a particular need, and it will serve other needs less well. And yet the pressure to make such data multi-task is enormous. And this leads to widespread problems.

Lets take an example. One commonly used bit of data is the British postcode. It is designed to delineate postman’s walks to organise mail delivery. It is not designed to reflect insurance risks, for example – but it is often used for just that purpose. As an example I was told by an eminent geologist about how he was asked to assess landslip risks in a town’s postcodes. He found one code which consisted of a valley with no homes in it, with the edge of the town where people actually lived. The landslip risk in the valley was high, but in the town it was negligible. So how to rate the risk for that postcode? According to the rules he was being asked to abide by, he should rate it as high. And yet that would mean that the homes in that code would be overcharged for their insurance. He refused to do it; but doubtless the insurance company found somebody more compliant. Why should they care about a bit of collateral damage? That kind of problem  predates modern IT, but technology allows it to proliferate in multiple hidden ways.

That perhaps illustrates the scale of the challenge that IT presents to liberal values. We as individuals are being made to conform to a world of arbitrary categories, because that is more convenient for systems builders. Instead of technology giving us more control over our lives, it is forcing us to conform to somebody else’s will.

But humans can be masters of technology, rather than being slaves to it. That is the liberal challenge.