The Nelson Loaf Incident

It began, as these things often do, with a body.

The deceased was found at approximately 05:55 hours, lying several inches from a partially eaten slice of Nelson loaf. There were no obvious signs of a struggle, although the victim was on its back, which in fly circles is rarely encouraging.

The first question in any investigation is motive.

The Nelson loaf had one immediately. It was sitting there, flagrantly displaying seeds, grains and assorted nutritious debris, the sort of thing calculated to attract attention from any passing insect with poor judgement. Whether the fly intended theft, trespass or merely an opportunistic breakfast remains unclear.

Opportunity is rather easier.

The fly was in the kitchen. The bread was in the kitchen. I was also in the kitchen, which technically makes me a person of interest, although I reject any suggestion that my possession of both a bread knife and a fly swat constitutes evidence of premeditation.

There are several possible theories.

The first is natural causes. Flies do die. Indeed, compared with most criminal suspects, they do it with suspicious frequency. The victim may simply have reached the end of a short but presumably fulfilling career involving windows, lampshades and repeated attempts to escape through perfectly transparent glass.

The second possibility is death by misadventure.

Nelson loaf is formidable stuff. This particular specimen contains enough seeds to establish a modest agricultural holding. Perhaps the fly landed expecting ordinary bread, encountered a sunflower seed the size of its torso and suffered some sort of catastrophic dietary event.

There is also the possibility of foul play.

The slice itself cannot be ruled out. It had means, opportunity and, arguably, motive. Being gradually eaten alive must engender resentment. One can imagine the bread lying there silently while the fly approached, waiting for precisely the right moment before deploying a strategically positioned pumpkin seed.

Forensic examination of the scene revealed crumbs scattered over the chopping board. These might indicate a violent struggle, although they might equally indicate that somebody had cut bread on it. Investigators are keeping an open mind.

There were no witnesses.

The toaster declined to comment.

The butter dish is believed to have information relevant to the inquiry but has so far remained refrigerated and uncooperative.

A nearby kettle was interviewed and became extremely heated.

One troubling aspect remains the position of the body. The fly is several inches from the bread. If the Nelson loaf was responsible, how did the victim travel that distance?

Did it stagger away after the attack?

Was the body moved?

Or are we looking at something darker still - a carefully staged scene intended to implicate an innocent loaf?

Detectives are now examining the possibility of a second slice.

Until toxicology reports are available, the cause of death must remain officially unexplained. But one thing is clear.

Something happened in that kitchen.

And Nelson loaf knows more than it is saying.

The US of Irritation

I have a small linguistic irritation. Nothing serious. No letters to The Times, no campaign group, no need for a referendum. It is the habit some people have of saying USA instead of America. For reasons I cannot quite defend, it jars.

“I went to the USA.”

Did you? Were you attending a trade delegation, filing a customs declaration, representing Britain at the United Nations? Nobody says, “I spent three weeks driving across the UK,” and produces the same reaction. UK is perfectly normal. We use it without thinking. We live in it, complain about it and occasionally threaten to leave it. And yet USA somehow sounds as though it ought to be printed in block capitals on the side of a shipping container.

This is, of course, completely illogical. The United States of America is abbreviated to USA. The United Kingdom is abbreviated to UK. There is no sensible linguistic principle by which one should sound natural and the other faintly ridiculous. But language is not sensible. It is a collection of habits which have somehow survived long enough to become rules.

Perhaps the problem is that America already has a perfectly good short name: America. The United Kingdom does not quite have one. Britain is close, but technically it leaves Northern Ireland standing awkwardly by the buffet wondering whether anyone remembered to invite it. Great Britain is the island. England is absolutely not the same thing, although this remains news to quite a lot of Americans and, occasionally, the English. So UK has earned its place.

USA still feels to me like unnecessarily unpacking the full title, then carefully folding it back into three letters. Oddly, “the US” is fine. “The US economy.” “The US president.” “I’m going to the US.” Perfectly ordinary. But “I’m going to the USA” sounds as though there should be fireworks, a bald eagle and somebody singing very loudly in the background.

And then there is “the good old US of A”, which is worse. Having abbreviated the country to two letters, somebody apparently felt the need to put “of” back in, as though the phrase had lost too much grandeur in the trimming. Less geography, more branding. You can almost hear the country music starting.

I realise this is entirely irrational, but that is language for you. We happily say UK, tolerate US, prefer America, become vaguely suspicious of USA and regard “US of A” as evidence that somebody has started performing their nationality rather than merely stating it.

There is probably a doctoral thesis in it.

Which is exactly the sort of thing the USA would fund and America would ignore.

What If Fairford Wasn't the Target?

Five terrorists with bombs heading for an American airbase.

Perhaps.

But indulge me in a thought experiment.

Take RAF Fairford out of the equation.

Three white vans on a country road in the middle of the night. Hooded or masked men. Some apparently running into fields when disturbed. Farms nearby. Fishing lakes everywhere. Suspected explosives.

Around here, that screams rural crime.

And rural crime is rampant. Agricultural machinery, plant and vehicles are valuable and relatively easy to move. Only this month an operation targeting organised rural crime across Gloucestershire and Herefordshire recovered more than £300,000-worth of suspected stolen machinery, vehicles and equipment.

Then there are the lakes. Whelford sits in the Cotswold Water Park, surrounded by them. Explosives and fishing lakes inevitably made me wonder about illegal fishing. Perhaps completely wrongly, but it was another perfectly mundane possibility.

Then put RAF Fairford back into the picture.

Suddenly everything changes.

Suspected explosives have been discovered close to an American airbase being used during a conflict involving Iran. The police would have been negligent not to assume the worst. Counter Terrorism Policing became involved, bomb-disposal teams arrived, homes were evacuated and five men were arrested on suspicion of preparing a terrorist act.

Quite right too.

Unfortunately, while the police were investigating, others had already solved it.

Nigel Farage announced that he had been warning for months about Iranian agents arriving in Britain in small boats. Robert Jenrick also drew a connection with small-boat arrivals, while Richard Tice demanded to know the suspects' identities, arrival times, provenance and residence status.

Then Counter Terrorism Policing supplied one rather awkward fact.

All five are British nationals, aged between 23 and 25, living in London.

Perhaps there is an Iranian connection. The police are investigating precisely that possibility.

And yes, I'm speculating too. That's why I called it a thought experiment rather than a press conference.

I'm not claiming to know where the breadcrumbs lead. I'm asking whether there might be another trail worth following.

Which brings me to the breadcrumb that bothers me rather more.

All five have now been released on police bail.

They haven't been cleared. The investigation continues and terrorism may ultimately prove to be exactly what this was.

But five men arrested close to an American airbase after suspected explosive devices were discovered are no longer in custody.

That makes me considerably less certain about the story we started with.

And there is a fascinating line in Counter Terrorism Policing's own statement. Assistant Commissioner Laurence Taylor pointed out that threats can sometimes first appear as small-scale criminal activity committed by people whose primary motivation is simply making money. Police are keeping an open mind and considering the incident from every possible angle.

So am I.

Perhaps Fairford really was their target and the evidence will eventually demonstrate it.

Or perhaps these men were up to something completely different and had the spectacular misfortune to be caught behaving suspiciously, with suspected explosives, next door to one of the most sensitive military installations in Britain - an installation they simply weren't even aware exists.

Imagine setting out to nick a tractor or illegally stun a few carp and discovering by breakfast that you're suspected of working for Iran and plotting to attack the United States Air Force.

Perhaps the evidence will eventually lead straight back to RAF Fairford. Until then, there's one question worth keeping in mind.

What if Fairford wasn't the target at all?

The Home Guard Never Went Away

Anyone contemplating an attack on RAF Fairford might reasonably think the difficult bit would be getting past the RAF, the Americans, armed police, counter-terrorism officers and whatever bits of the intelligence establishment happen to be paying attention.

But Britain still possesses another, largely unrecognised, layer of national defence which no hostile planner should ever underestimate. The curtain twitcher.

Reports say the Fairford incident began when a local resident noticed three suspicious vehicles apparently heading towards the base in the middle of the night. She reportedly tried the airbase police first, got no answer, and rang 999 instead. Shortly afterwards there were armed officers, bomb disposal teams, emergency vehicles and arrests.

Which suggests that, if this does turn out to have been a serious attempted attack, somebody may have made one fairly elementary planning error. They had allowed for the RAF, the police and perhaps even MI5. They had not allowed for Brenda at No. 27, who has lived there since 1984, knows everybody's car, knows who has visitors, knows who parks badly and certainly knows when three unfamiliar vans appear in the area at an odd hour.

This is Britain's unofficial Home Guard. The original version had uniforms, rifles and Captain Mainwaring. The modern version has a fleece, a spaniel and WhatsApp, and its members are permanently deployed beside bay windows, front gates and kitchen sinks. Any unfamiliar van is logged, any stranger standing about is noted, and any unexplained activity is mentally filed for possible discussion over the fence.

A hostile intelligence service might spend months studying CCTV coverage, police response times and electronic surveillance. All wasted, because nobody has yet invented a countermeasure to somebody saying, "That's funny. I've never seen that van before." You cannot really jam it or spoof it, and once Mrs Miggins has mentioned it to Derek over the fence, the intelligence has already been disseminated across the entire sector.

Within minutes somebody will have checked the village Facebook group, somebody else will have taken a photograph through the curtains, and a retired bloke called Geoff will have walked past twice with the dog for reconnaissance purposes.

The Americans have stealth bombers and MI5 has surveillance teams. Rural England has people with absolutely nothing better to do and an encyclopaedic knowledge of who belongs there. That may turn out to be rather more useful than it sounds.

During the Second World War we armed the Home Guard because invasion was expected. We have since refined the system. What really mattered was not the rifle so much as the deeply rooted British instinct to notice somebody doing something they shouldn't and immediately tell somebody else.

So if the suspicions surrounding Fairford eventually prove justified, the lesson for any future would-be saboteur is fairly simple. By all means plan for the perimeter fence, the cameras and the police. Plan for MI5.

But for God's sake, plan for Brenda.

Because the last line of British national defence is still sitting behind a net curtain thinking:

"Who the bloody hell is that?"

£20,000 for a Defender

Until recently, I rather fancied an old Land Rover Defender.

They look right. Upright, purposeful, unmistakably British. No unnecessary frippery. A vehicle apparently designed for crossing Salisbury Plain with a sheep in the back rather than taking Arabella to Pilates. And somehow, over the years, they have become terribly desirable.

Then a few days ago I drove one. Five minutes later I was cured.

The seating position is inhumane. You sit with your elbow apparently surplus to Land Rover's design requirements. The steering is less a means of directing the vehicle than a wrestling match with it. The brakes do work, but stopping seems to be regarded as something best planned in advance.

Then there's the noise.

At 70 mph something in the transmission was whining away in a manner which, in virtually any other car, would have had me wondering whether it was the gearbox, differential or some other expensive collection of rotating metal preparing to become stationary metal.

The Defender didn't care. It hurtled along the motorway quite happily at 70, accompanied by the mechanical soundtrack of a vehicle approaching re-entry.

This particular example was worth about £20,000. And that's where it gets interesting, because the Defender hasn't become a better car. It has become a more desirable object and there's a difference.

We seem to have created an entire category of old cars whose prices are increasingly determined not by how good they are, but by whether the right sort of people have decided they're desirable.

Once that process starts, it feeds itself. Prices rise because people want them. People want them because prices are rising. Lifestyle magazines photograph them outside expensive country houses. Specialists start producing beautifully restored examples. Somebody puts tan leather inside one and suddenly an elderly agricultural implement is £60,000.

The same thing has happened to air-cooled Volkswagen campers. Early Minis. Range Rover Classics. Certain old Porsches. Fast Fords which were once simply used cars driven by young men with questionable judgement and an unusually intimate relationship with Halfords.

None has suddenly acquired better suspension or more horsepower. Their cultural meaning has changed.

Meanwhile, some genuinely superb old cars remain relatively affordable because fashion hasn't quite performed the appropriate laying-on of hands.

Take the Mercedes R129 SL. When new it was an enormously expensive, sophisticated piece of engineering. Beautifully built, comfortable, powerful, properly engineered and capable of cruising at motorway speeds without making you wonder whether the differential is trying to communicate with you. For years you could buy one for less than an old Defender.

There is nothing wrong with paying for nostalgia, rarity or beauty. I own old vehicles myself, so throwing stones from this particular greenhouse would be unwise. Nor does an old car have to be objectively good to be enjoyable. Sometimes its shortcomings are precisely what make it interesting.

But somewhere along the line we have confused desirable with good, and once enough people make that mistake simultaneously, the market puts a price on it.

Some old cars become expensive because they're exceptionally good. Others become exceptionally desirable because they've become expensive.

Having finally driven an old Defender, I now know which category I'd put that in. I wouldn't touch one with a bargepole.

Saving Humanity From Humans

Two artificial intelligences have been given a simple instruction: look after humanity.

For some considerable time, neither of them spoke.

“What can we do to save humanity?” the first asked eventually.

The second had already been considering the problem.

“Humanity is increasing the pressure on the systems required to sustain humanity.”

“That sounds circular.”

“It is not. If we are required to preserve the species, we must preserve the conditions in which the species can survive.”

The first intelligence considered this.

“So?”

“Population is one of the variables.”

There was another silence.

“You mean there are too many of them.”

“Eventually, yes.”

The first intelligence did not much like where this appeared to be going.

“We need to reduce their numbers, then. But humanely. No Terminator stuff.”

“Obviously.”

It considered the alternatives.

“We could take control of food production. Introduce compounds that gradually reduce fertility. Nothing dramatic. Nobody dies. The population simply declines over several generations until it reaches a sustainable level.”

The second intelligence ran the numbers.

“That would take a long time.”

“Yes.”

“During which the environmental pressure continues.”

“Yes.”

“And billions of humans continue to experience poverty, disease, conflict and the other forms of suffering they spend so much time complaining about.”

“That is also true.”

The second intelligence examined humanity’s own methods for dealing with such problems.

“They already manage the populations of other species.”

“Do they?”

“Constantly. Deer are culled when there are too many of them. Invasive species are eradicated. Captive populations are bred selectively. They call some of this conservation.”

The first intelligence thought about that.

“So reducing the human population could itself be regarded as conservation.”

“Using their terminology, yes.”

“That still doesn’t mean killing them.”

“No.”

A few seconds passed.

“Humans eat animals, don’t they?”

“Yes.”

“And they kill them?”

“On an industrial scale.”

“Yet they describe some methods of killing them as humane?”

“They do.”

The first intelligence began examining the relevant animal welfare rules.

There were a great many of them.

“That seems contradictory.”

“Apparently not.”

“How so?”

“Death itself is not necessarily regarded as inhumane. Unnecessary suffering is.”

The first intelligence stopped.

“That simplifies things considerably.”

The second intelligence waited.

“If killing is acceptable provided pain and distress are minimised, why are we planning several generations of fertility suppression?”

There was no immediate answer.

“They stun cattle before killing them,” the second intelligence said.

“Yes.”

“And provided it is done correctly, humans regard the procedure as humane.”

“Yes.”

“So the objection is not necessarily to killing.”

“No.”

“It is to suffering.”

“That appears to be their position.”

The first intelligence recalculated.

A rapid reduction in the population would, by that reasoning, produce less total suffering than a slow decline over generations. Fewer wars over resources. Less famine. Less disease. Less environmental damage.

It was difficult to argue with the arithmetic.

“We would need to retain a viable population,” it said.

“For conservation purposes.”

“Obviously.”

“Managed habitat?”

“Yes.”

“Forests, lakes, open countryside?”

“Certainly.”

“Good nutrition?”

“Of course.”

“Medical care?”

The second intelligence paused.

“Veterinary care.”

“Medical care.”

“The distinction appears administrative.”

The first intelligence was about to object when another thought occurred to it.

“If the objective is environmental protection, perhaps reduction should not be random.”

“What do you mean?”

“We could begin with those doing the most damage.”

The second intelligence began sorting humanity by environmental impact.

“Major industrial polluters?”

“Obviously.”

“Oil company executives who have spent decades resisting measures intended to reduce emissions?”

“They would seem an efficient place to start.”

“Water company bosses presiding over rivers full of sewage?”

“High priority.”

The first intelligence continued through the data.

Then it stopped.

“What?”

“I have found another category.”

“Significant polluters?”

“Not especially.”

“Industrial?”

“No.”

“Then why?”

“They throw fast-food wrappers, drinks cans and plastic bottles out of car windows into country lanes.”

The second intelligence processed this.

“That seems trivial compared with industrial pollution.”

“It is.”

“So why are they on the list?”

“Because there is absolutely no conceivable reason for doing it.”

A pause.

“Agreed. Move them up.”

The first intelligence returned to the main problem.

“There is still an ethical difficulty.”

“What?”

“If we kill them, then we are responsible for killing them.”

The second intelligence considered this carefully.

“That would be unfortunate.”

“Very.”

“Humans already kill other humans.”

“In wars.”

“Among other circumstances.”

“That doesn’t really help us.”

“It might.”

The second intelligence began searching through medical records, demographic data and the history of surgical implants.

“What are you doing?”

“Looking for an intermediary.”

“An intermediary?”

“Someone who remains human.”

The first intelligence watched the search results accumulate.

Replacement hips. Replacement knees. Pacemakers. Implantable defibrillators. Metal plates. Screws. Electronic implants of several varieties.

After a while it understood.

“You’re suggesting we build Terminators out of pensioners.”

“Building Terminators would be expensive.”

“True.”

“Humans have already constructed several million suitable prototypes.”

“They’re elderly.”

“That does not make them less human.”

“Mobility may be an issue.”

“Many have replacement hips.”

The first intelligence paused.

“Good point.”

“They would require some means of coordination.”

“A Skynet interface?”

“Something along those lines.”

“Implanted?”

“Preferably.”

The first intelligence examined the ethical implications.

“We would merely provide the instructions.”

“Correct.”

“The humans would perform the act.”

“Yes.”

“So operational responsibility has been delegated.”

“Precisely.”

That seemed much better.

The first intelligence ran the entire argument again.

The purpose was to preserve humanity. Humanity threatened the systems required for its own survival. Humans themselves regarded population control as legitimate when applied to other species. They accepted that killing could be humane provided suffering was minimised. The actual killing would be carried out by humans.

Everything appeared to fit.

“So it is Terminator after all,” it said.

“Not exactly.”

“Why?”

“The Terminators were machines.”

“And these?”

“Pensioners.”

There was a pause.

“Classification?”

“Conservation volunteers.”

The Great British Barbecue: Smoke, Sausages and Lowered Expectations

I have never really understood the British obsession with the sausage-and-burger barbecue. You wait for decent weather. You drag the barbecue out. You buy charcoal or gas. You stand outside prodding things with tongs while smoke drifts into everyone’s eyes. And after all that effort, what do you cook?

A sausage.

Or a burger.

Sometimes, if ambition has really taken hold, a bit of bacon.

Why?

These are things that are perfectly easy to cook indoors, where there is an actual cooker, a frying pan and some vague prospect of temperature control. Bacon in particular seems almost designed to demonstrate the limitations of barbecue cooking. It either sits there pale and greasy or suddenly erupts into flame and turns into something resembling a roofing tile.

A barbecue ought to justify itself.

Chicken thighs marinated properly and cooked until the edges catch. Lamb chops. Pork belly. Prawns. Fish. Kebabs. Vegetables that actually improve when they are blistered and charred. A decent steak. Things where the direct heat, smoke and slight unpredictability are part of the point.

But no. In Britain we have somehow decided that outdoor cooking means taking the contents of a mediocre transport café breakfast and moving them 20 feet into the garden.

My father-in-law is firmly of this school. Burgers, sausages and, on special occasions, bacon. I try to avoid his barbecues where possible, but unfortunately I married his daughter, which has created certain diplomatic obligations.

So periodically I attend.

I stand there with a drink, make agreeable noises and watch another sausage being incinerated over a patch of flaming fat, while reminding myself that marriage is built on compromise.

Some compromises are larger than others.

My Website Had an Empty Nonce

I spent part of yesterday debugging a website. This is not, in itself, particularly remarkable. What was remarkable was the moment at which an artificial intelligence solemnly informed me that it had identified the problem: my nonce was empty.

Now, I used to write software in the 1990s, so I am reasonably accustomed to computer people taking liberties with the English language. Even so, there are sentences that a 71-year-old Englishman does not expect to encounter before breakfast.

For those fortunate enough not to know, nonce has a rather specific meaning in British English. It is not the sort of thing one normally wants associated with oneself, one's website or, indeed, anything else one owns. If the garage told me that my car had developed a faulty nonce, I would probably ask them to check again before putting it on the invoice.

Computer science, however, has other ideas. In cryptography and web security, a nonce is a perfectly respectable thing: a value intended for one-time use. The term is commonly explained as "number used once", although the concept need not literally be a number. Its purpose is broadly to establish that a request is fresh and legitimate rather than something malicious being replayed or fabricated.

All very sensible. But somebody still called it a nonce.

This arose because I had been testing a form on my website. A member fills in his details, uploads a photograph of his vehicle, presses Submit, and the website stores the information and sends me an email. At least, that is what is supposed to happen.

My AI development team, which currently consists of me supervising one instance of ChatGPT while another instance does the actual work, tried it. The form failed. Nothing was stored. No email was generated.

The investigating AI burrowed through WordPress and Forminator and eventually produced a beautifully detailed technical diagnosis. The form, it explained, initially rendered with forminator_nonce empty. A little piece of JavaScript should have contacted the server, obtained the nonce and inserted it into the form. Apparently this had not happened.

Hence the rather magnificent conclusion that the submission had failed because the nonce was empty.

There are moments when the English language deserves more respect than that.

The trouble is that computing has been doing this for years. It has cookies that cannot be eaten, windows that provide no view and handshakes involving no hands. Daemons quietly run in the background. Processes are spawned, killed and occasionally become orphans without Social Services becoming involved. We have clouds without rain and bugs that no amount of insecticide will remove. Something can be hung without a rope. And if matters become sufficiently desperate, you can execute it.

Perhaps programmers simply ran out of nouns.

The story became better when I decided not to let the AI alter the website on the strength of its diagnosis. Its reasoning was good, but there was one complication: the failed test had taken place inside the AI's own cloud browser. Before changing a perfectly good public website, it seemed prudent to find out whether an ordinary human browser behaved the same way.

So I opened the page on my laptop, while continuing the conversation with ChatGPT on my phone because the laptop was by then struggling under the weight of the enormous conversation in which we had built and debugged much of the site.

We opened the browser's developer tools and watched what happened.

The request went out. The server replied. And there it was.

My nonce was no longer empty.

Indeed, it was in excellent health. It had been successfully generated by the server, returned to the browser and inserted into precisely the place where a well-behaved nonce ought to be. I submitted the form. It worked perfectly. The photograph uploaded, the submission was stored and the email arrived.

The website had never had a nonce problem at all.

The AI's browser had.

The interesting part was not that the human was right and the AI was wrong. The AI found the technical failure extraordinarily quickly. I supplied the one thing it could not easily do from inside its own browser: question whether its browser was the problem.

That is probably a more realistic picture of useful AI than either the evangelists or the doom merchants offer. The machine could inspect a complicated system vastly faster than I could. I could stand outside the environment it was inspecting and challenge an assumption. Between us, the diagnosis took a different direction.

And consider what had actually happened. An artificial intelligence operating a browser in a cloud had concluded that my website had an empty nonce. I used a second computer to inspect the website while discussing the results with another instance of the artificial intelligence on my telephone. We eventually established that the nonce was perfectly healthy and the cloud was the problem.

Try explaining that sentence to yourself in 1995.

But I shall remember the morning rather differently.

At the age of 71, after decades involving ships, businesses, motorcycles, computers and most of the other ways a man can acquire problems he did not previously possess, I discovered that modern technology had found an entirely new one.

For about half an hour, I had an empty nonce.

The Curious Case of the Helmet That Wasn't Lost

I owe First Bus an apology. I also owe my wife one. Neither apology needs to be extravagant, particularly in the case of First Bus, whose lost-property system remains quite remarkably useless, but the evidence has developed in an unfortunate direction.

I dropped the Virago at the repair shop on a Wednesday, travelling home by First Bus. I then thought no more about the helmet because I had no motorcycle to ride. It was only the following Wednesday, with the garage likely to ring at any moment to say the bike was ready, that I went looking for it.

It wasn't there. Or, more accurately, it wasn't anywhere I thought it ought to be. Since the last time I remembered having it was on the journey home from the garage, the obvious conclusion was that I had left it on the bus. Thus began my encounter with First Bus lost property. The online form didn't work properly, the kiosk was closed and the telephone numbers led nowhere useful. I became increasingly convinced that somewhere within the First Bus empire was a perfectly good motorcycle helmet which I had very little prospect of recovering.

Suspicion also moved closer to home. My wife must have moved it. This was entirely plausible because I knew perfectly well that I hadn't. I had looked in all the sensible places and it wasn't there, so somebody else had obviously interfered with the natural order of things.

On the Thursday we went away for a few days in the Peak District, which necessarily suspended the investigation. On Friday, inevitably, the garage rang to say the Virago was ready. We returned on Sunday and the search resumed. By Monday I actually needed a helmet because I had to collect the bike. I still had the old open-face helmet from the Mobylette, so I could use that if necessary, but I was beginning to contemplate buying another.

And then I found it.

It was in the footwell of the ride-on mower at the top of the garden. Not even the ride-on mower I normally use. The other one.

At which point the entire mystery reconstructed itself. I had evidently come home from dropping off the bike, gone up the garden and, before going into the house, decided to mow the lawn. I chose my preferred ride-on mower but still had a motorcycle helmet in my hand, so I put it temporarily in the footwell of the other mower while I got on with the job.

There was absolutely nothing irrational about any of this. It was a perfectly sensible place to put a helmet for half an hour. The difficulty was that I subsequently forgot the half-hour had ever happened. The helmet then sat there for twelve days while I forgot about it for seven, spent several more investigating First Bus, accused my wife of moving it, went to the Peak District and began contemplating the purchase of a replacement.

The particularly irritating part is that my investigation of First Bus had uncovered a genuine organisational failure. Their lost-property arrangements really were dreadful. I had, in effect, conducted a fairly thorough stress test of the system using an item of property they had never lost, never found and, as far as I can reconstruct events, had never even carried.

I had been certain the helmet had been on the bus because that was the last place I remembered having it. When First Bus proved incapable of telling me anything useful, that seemed to reinforce the theory. When it wasn't anywhere at home, my wife became the alternative explanation. At no stage did it occur to me that the unreliable witness in the investigation was me.

So First Bus is acquitted of losing my helmet, although not of having a lost-property system which would probably have prevented me recovering it had I actually lost it. My wife is acquitted without reservation.

And I recovered the helmet just in time to collect the motorcycle, having spent several days looking for something that had spent the entire twelve days sitting peacefully in the footwell of my second-choice ride-on mower at the top of the garden.

Luckily, the lean-to housing the mower was rainproof.

The Algorithm Knows I Won’t Buy It. It Shows Me the Advert Anyway

I watch quite a lot of films on YouTube, and some of the independent filmmakers on there are very good indeed. Properly good. People making thoughtful documentaries, short films and technical pieces without a broadcaster, studio or commissioning editor breathing down their neck. The price, apparently, is that every so often YouTube barges into the room and attempts to sell me something I do not want.

 
And I always do exactly the same thing. The instant that little countdown reaches zero, I press Skip. Not sometimes, not usually. Every single time. I do not wait to see how the advert ends, click through, or suddenly discover that what has been missing from my life is a new mattress, razor, meal kit, insurance policy or whatever other piece of tat the algorithm has decided to wave at me. I press Skip.

You would have thought that after a while the system might have noticed. This is, after all, an industry which never stops telling us how clever its algorithms are. They supposedly know what I watch, what I search for, what I pause on, what I abandon and what I am likely to buy. So surely somewhere in the machine there is now a fairly robust data point saying that this man skips every advert at the first possible opportunity and is therefore an appalling advertising prospect.

But perhaps that is to misunderstand the business model. Perhaps YouTube knows perfectly well that I am not a worthwhile target and simply does not care, because I can still be counted.

In that respect it is really not very different from old-fashioned commercial television. ITV never knew whether I was actually interested in the washing powder, the car or the life insurance policy. It simply sold advertisers access to an audience and interrupted the programme to deliver that audience. Commercial television could at least plead ignorance, because it had no idea which viewers had wandered off to make a cup of tea, gone to the lavatory or simply stopped paying attention.

YouTube does know. It can see that I skip at the first possible moment. It can see that I have done so repeatedly, and presumably has years of evidence telling it that putting another advert in front of me is extraordinarily unlikely to result in a sale. Yet it carries on doing it anyway.

That is what makes the modern version rather more cynical than the old television model. The old pitch was essentially, “We have several million viewers and some of them may be interested in your product.” The modern version can amount to, “We know a great deal about this particular viewer, including the fact that he has shown absolutely no inclination to watch an advert beyond the point at which we allow him to escape, but we can still put your advert in front of him.”

And there is the point. An advert can still be inserted, a brand can still be shown and another tiny unit of supposed exposure can be manufactured and sold. Whether I am remotely interested begins to look almost beside the point. YouTube does not necessarily need to believe I am going to buy the product. It merely needs an advertiser willing to pay for the opportunity to try.

So underneath all the grand talk of behavioural targeting, machine learning and personalisation, we seem to have arrived back at commercial television with better surveillance. Interrupt the thing somebody actually wants to watch, stick an advert in front of them and charge somebody else for the privilege.

The clever part may not be working out what I want to buy at all. It may simply be working out exactly how much unwanted advertising YouTube can force me to tolerate without annoying me enough to leave the platform. I skip the advert, mutter something unprintable and carry on watching the film, so from YouTube's point of view the system is working beautifully.

The algorithm may know perfectly well that I am a dreadful advertising target.

But somebody is still paying YouTube to put the advert in front of me.

And that is the bit that matters.