Brian Clegg's Blog, page 25

The Capture (Series 2) - BBC iPlayer

I don't usually review two seasons of a TV show separately, but I'd like to follow up my review of the first series of The Capture, partly because the second series is better than the first (though like the first it had some credibility issues) and partly because the political context (as opposed to criminal law) brings into sharper focus the way this kind of information technology can be misused.

As before, writer Ben Chanan proves good at giving us shocking scenes where deepfake videos and audio are used both to conceal what's really happening from security cameras and police radios and to replace broadcast video, totally twisting the content of a political interview. This seems to be intended to ruin rising political star Isaac Turner's career - though in reality things are far more complex.

We get the same state actors - MI5, CIA and SO15 - as in the previous season, plus Truro Analytics (surely not a name that is supposed to bring Cambridge Analytica to mind?), a political media manipulation agency.

Apart from a rather saggy fourth episode, Chanan keeps up the pressure and makes this a very engaging thriller that has significantly more at stake than the first season. Featuring as it does a government minister, the stakes are higher and the addition of the diva Newsnight presenter Khadija Khan (as before, hardly any of the characters are likeable) is a brilliant move.

It was, without doubt, my favourite TV watching of the last few months. In terms of its political message, I just wish it had explored the shades of grey more. The central technology to do those deepfakes, Correction, was being used by the government to manufacture evidence where they had good intelligence that there was terrorist activity, but not evidence that could be used in court. This is clearly the start of a very slippery slope (as this season shows) - but when the MI5/police operative DSu Gemma Garland is trying to sell it to our heroine DCI Rachel Carey, she just says what they are doing and why, which isn't convincing.

It would have been much more interesting to produce a hierarchy of moral/ethical decisions. Would Carey support killing terrorists in a war? Assassinating terrorist leaders (e.g. Osama bin Laden)? Police shooting to kill where it's necessary to preserve lives? Security services shooting to kill when it's necessary to preserve lives? Faking evidence to convict terrorists where there is clear evidence such as phone taps that can't be used in court for some reason? I'm not saying Correction is a good idea - but rather that knowing exactly where to draw the line is genuinely difficult decision.

However, that sort of moral consideration is perhaps too much to expect from a fast-paced thriller. The fact remains, this was excellent TV.

I'm going to finish off by pulling apart the last episode in some detail, as this is where the credibility issues that did irritate me loom large, so don't read on unless you've watched it all or don't care about spoilers.

See all of Brian's online articles or subscribe to a weekly digest for free here

===== SPOILER ALERT ======

For me, there was a feel of desperation about the final episode, as a sort of happy ending was shoehorned in at high speed - and it also left a lot unexplained. One problem was trivial - the head of Truro Analytics, Gregory Knox was seriously beaten up by Carey in the process of arresting him - but in all shots after this he didn't have a bruise on him. But the bigger problem was the way the denouement worked.

The show ends with an interview between the Newsnight presenter Khadija Khan and Isaac Turner - but Turner pulls out and the interview goes ahead with a deepfake of Turner being interviewed (with Khan's compliance), which is then used to unveil Correction to the world when Turner is shown to be actually in Piccadilly Circus, not in the studio.

One oddity here is that Turner is not in on the trick - so what would have happened if he hadn't pulled out of the interview? Of course, the whole thing could have then been a deepfake, but they couldn't do the reveal in Piccadilly Circus.

A second problem, once it's all explained is the incredible ability of Truro Analytics, which goes far beyond anything even vaguely feasible. The very premise that somehow you can improve someone's ratings by showing different aspects of them to key target audiences is fine on social media, but doesn't work on broadcast media. But also, it seems highly unlikely that Knox's company, which was no Google or Apple, could both do the social media manipulation and create perfect deepfakes in real time. Most painful was the company's ability to do full blown psychohistory - predicting what will happen down to the behaviour of individuals. This shows a total lack of understanding of chaotic systems.

And finally there's the big finish. The fake interview is taken over by replacing a huge chunk of the fake Turner's script. Carey is shown to be introducing the change with a click of the mouse. But we have been following her pretty well in real time, and at no point did she have the chance to write this script or somehow insert it into the system. What's more, once it is in play, it is somehow locked in - yet in an earlier scene they showed how easy it would be to edit the script on the fly - where has this ability to lock it in come from?

This didn't entirely ruin the ending - we did get our 'goodies triumph over the nasty intelligence people' moment. But it left it feeling a bit contrived.

Has the BBC not heard of paragraphs?

The BBC is, of course, a broadcaster rather than a publisher - yet its website, and particularly its news website, should surely be treated as if it were a publication. That being the case, it's time they got someone involved who knows how to lay out text. All too often their articles look as if they've been taken straight off the autocue.

Look at the start of the 'article' from political editor, Chris Mason illustrated above. The whole thing is more than twice as long as the extract shown - and pretty well the only paragraph that has more than one sentence in it is a quote.

It's horrendous. I find it really difficult to read. Please, someone at the BBC, get your act together. 

Alternatively you might like to approach the Royal Literary Fund. They're really good at helping students and organisations with their writing skills - it's clear that the BBC needs some assistance as soon as possible.

The Sandman - Netflix

Although I'm a big Neil Gaiman fan, I've never read The Sandman, as I just can't relate to comic books and graphic novels. I'm not a very visual person, and I like a story to contain (far) more words. Nonetheless, I was aware it was considered something of a big thing in the genre, so watched the Netflix version with interest.

The first episode is distinctly on the slow side - I know at least one person who gave up part way through - but it's worth persevering to make your mind up, as things certainly change gear after a while. However, I do think the series has an underlying problem in the way it has apparently been painstakingly based on the individual comics. This gives an extremely episodic approach, making it almost impossible to produce a really top notch drama series.

The problem is not having mostly separate stories in each episode - I like that. But if a series is take that approach, it needs two other things to support it - a group of good recurring characters and a strong story arc across the series. It's even better if it can also incorporate humour despite the subject matter - that's why Buffy the Vampire Slayer was so brilliant. That had all these elements. But The Sandman lacks most of them.

I really thought things were improving with the introduction of Jenna Coleman's character in the third episode. Johanna Constantine is a perfect foil for Tom Sturridge's well played, but frankly morose Dream/Morpheus. There was enjoyable interaction between main characters and that humour gave useful contrast with the darkness elsewhere. And then she was gone for the rest of the series (except a cameo as her ancestor). I also wanted to see far more of Vivienne Acheampong's excellent librarian, Lucienne. Unless Netflix fixes this with more recurring characters that appear in most episodes, this series won't hit the spot.

It's not that there can't be one-off episodes dedicated to a specific event and key character.  Episode 6, featuring Dream's sister Death and a human who the pair grant a free pass from death in the fourteenth century, who Dream meets up with once every hundred years, is brilliant. But, for example, the fifth episode, primarily featuring a secondary character in a diner is just dull. And when a continuing story kicks in across episodes 7 to 10 (still not a proper series story arc) it has too much material not set in the real world, is convoluted and lacks the humour and energy of the Constantine episode. It's plodding by comparison.

Overall, the CGI is okay, but lightweight in places. The buildings in the Dreaming never looked real, and though you could argue that's how they should look in the Dreaming, it didn't appear to be intentional. The biggest CGI disappointment was the gargoyle Gregory in the Cain and Abel sequence of episode 2 - it was good in close up, but when flying it looked like something out of a video game. I do have one other moan on the visuals - the nighttime sequences, of which there are a lot, are very low contrast - even with my TV in dynamic mode it was sometimes difficult to make out what was going on.

Overall, interesting concept, occasional good episode, but needs a serious overhaul.

See all of Brian's online articles or subscribe to a weekly digest for free here

Drowning in SEO spam

If you don't have a website, you are probably lucky enough not to be bombarded with emails promising to improve your SEO (Search Engine Optimisation). Here's a typical example that I received recently:

I have looked at your website and realize your website is great design but your website ranking not good all search engine Google, AOL, Yahoo, and Bing. 
Do you want more targeted visitors on your website? 
We can place your website on Google's 1st page. Yahoo, Facebook, LinkedIn, YouTube, Instagram, Pinterest etc.). 

We also offer the most competitive rates for this service, May I send you quote?

What is a bit scary (apart from the grammar) is the sheer volume of this stuff I receive. Thankfully, my spam filter is pretty good at clearing them out of my inbox, but one day this week, for example, I received 20 such emails, and rarely get fewer than ten a day. 
We can only assume that some companies respond to this bombardment, or why would the spammers bother? And when they do, it can result in potentially respectable companies being hauled into the mire. Because, sadly, some of the less salubrious SEO companies try to get better search engine ranking by... spamming.
Specifically, they attempt to put comments on blog posts like this one with links to their customers' websites, as a result, in theory, pushing the clients' sites a little up the search engine ranking. Traditionally, those spam posts have been for something generic, such as the ones in the image at the top (shown larger below so you can read some of their purple prose (I particularly like 'thanks for coprating profit automation'.)

But lately I have been getting a few for what appear to be ordinary, presumably honest, UK companies that have been taken in by these spammers. When I do get a dodgy comment linked to what appears to be a genuine company I always email them to point out what is happening - but I  am yet to have a reply from any of those I have alerted. This is the kind of thing:

The trouble is, if you want to be, say, the 'most trusted air conditioning & heating engineer service' anywhere, it's not a good look if someone is spamming in your name. Here's another example:

Note, by the way, the sheer irrelevance of the comments to the subjects of the blog posts. It's not just comment spamming, it's bad comment spamming.
So if your company pays for SEO and you get someone complaining about this happening, don't ignore them. Apologise and sack your SEO company. You know it makes sense.

Hydrogen? No thanks...

Image by Anne Lund from Wikipedia
reproduced and modified under  GNU 1.2I've always found the old anti-nuclear badge hilarious. It might say 'Nuclear Power? No Thanks', but its central image of the smiling sun portrays the biggest nuclear reactor within several light years. But I can't resist adapting it as a comment on the over-enthusiasm for some applications of hydrogen in green energy.

Don't get me wrong, hydrogen is a wonderful element - literally the number one. There is a specific irony in my use of that image because, of course, hydrogen is the fuel of that vast nuclear reactor that keeps us alive - and, for that matter, it's one of the two component elements of the water that is essential for life.

I am also not denying that hydrogen has a useful role in the future green economy. It's often mis-described as an energy source. It's actually an energy store, and as such could be useful as a way of storing away renewable energy. For example, solar energy can be used when the Sun is not shining by producing hydrogen at power stations that is later cleanly burned. It is also a potential fuel for large, specialist vehicles if batteries are not a sensible option, such as long range buses, lorries and possibly aircraft.

However, there are two 'green energy' applications of hydrogen that are still mentioned but that have serious problems.

One is in cars. The Grand Tour presenters (formerly of Top Gear) have frequently praised hydrogen as a petrol replacement, in opposition to electric cars. But for me it has three issues. One is it's even less safe to handle and live with than petrol. A second that it has lower energy density than petrol, so you get less range for the same tank size. But the biggest one is that hydrogen ties us to fuel stations. We already have a much bigger distribution network for electricity. When, many years ago, I had a Mondex card, the original experimental cashless card, its biggest plus was I could top it up at home - I didn't have to go somewhere to get cash. The same applies to electric cars. Of course they need to be cheaper and with better range, but that will come. But who would want to keep filling stations if we could get away from them?

The other problematic application is using hydrogen as a replacement for gas heating in homes. The idea is that you put hydrogen down the same pipes we currently put natural gas down - so, like electricity, the distribution network is already there to the point of use. Sounds great. But I was always worried that hydrogen, which is better at leaking than natural gas would find far too many ways of escaping... and I'd rather get away from pumping highly flammable substances into the home.

Now, though, there's a second and even more convincing reason - it's not a cheap solution, nor is it even particularly green. This isn't a comment from an alarmist source, it's the outcome of a review of 32 independent studies, reported in Chemistry World. As the original study notes 'it takes about five times more electricity to heat a home with hydrogen than it takes to heat the same home with an efficient heat pump, either individually or as part of a district heating network.' This is not the green future.

This has been a Green Heretic production.

See all of Brian's online articles or subscribe to a weekly digest for free here

The Capture (Series 1) - BBC iPlayer

Though hugely flawed in some ways, The Capture proved to be one of the most gripping TV shows I’ve watched in a while, thanks to the regular invocation of clever twists that make the viewer think ‘How is that possible?’

The focus throughout is video surveillance - specifically how, and if, it can be misleading or tampered with to make something that didn’t really happen appear to be the case.

I’m specifically reviewing the first series - I haven’t started the second yet, but will give it its own review. The focus initially is a soldier accused of killing someone without the need to do so in Afghanistan. (This was why I didn’t watch the series when it first came out as I tend to avoid military topics.) But although the soldier in question becomes a major character throughout, it isn’t really his story.

It was clever and did more than entertain, really giving the viewer an opportunity to think about the underlying moral dilemma (the details will have to wait until after the spoiler alert). The other main character, a fast-tracked female DI, temporarily assigned from SO15 (counter terrorism) to murder was well filled by Holiday Grainger, emphasising both the initial resentment of her by the team and her sometimes ruthless urge to get on, as the negative balance to her positive unwillingness to let go, even when ordered to do so.

— SPOILER ALERT —

The most unlikely thing for me was that to make the whole thing work required not one, but two conspiracies. Firstly there was the CIA/MI5/SO15 grouping referred to as Correction and secondly a more unlikely grouping of human rights activists, who decide to use the mechanism employed by Correction to show that a terrorist’s conviction relied on fake video evidence. This did make for some intriguing twists, but seemed one conspiracy too far.

The Correction aspect was also overplayed in the sense that the CIA outfit particularly seemed far too effective. But this is where the moral dilemma card is played, even if it seems highly unlikely in practice. The Correction involves faking video evidence to convict terrorists when the security services know that they are guilty, but this knowledge is based on data that can’t be used in court. The details seemed a little hazy - it would only be theoretically justifiable if things like communication intercepts were always inadmissible, which I didn’t think was the case.

The other other-the-top aspect was the way that the security services reacted to the fake video produced (with extremely unlikely skill) by the human rights activists - the botched and extreme measures taken seemed logically unnecessary. I’d also say that the soldier’s introduction to the activists was like something out of Poliakoff’s excellent fantasy The Tribe - it was far too slick for the reality of who they were.

So, a lot of stretching of likeliness is needed to make this work - but it's worth it, both for the gripping drama and the possibility that something like this might be done, whether by security services or rogue states.

 

The mnemonic trap

As an author, it's not uncommon to get emails or letters correcting something in one of my books. Sometimes these corrections are useful, at other times, the correspondent misses the point. But I recently had one from Ronja Denzler that was not only correct, but also highlighted something really interesting about mnemonics.

These phrases to remember something can be genuinely handy - most of us can still recall those for rainbow colours or planets (often still incorporating Pluto) from school, while I distinctly remember a woman called Ivy Watts from my physics class. But the most elegant are the numerical mnemonics, where the numbers of letters in each word represents a digit. This form reaches its zenith in the mnemonics for pi - so much so that the art of producing these has its own, distinctly tortured, name of 'piphology'.

When I wrote Introducing Infinity - a graphic guide in collaboration with the excellent illustrator Oliver Pugh, I asked if he could use a fuller version of the mnemonic I vague recalled from school. The bit I could remember was 'Now I, even I, would celebrate in rhymes unapt the immortal Syracusian...' - Oliver extended this further in the image below:

It was this illustration that Ronja wrote to complain about - because the thirteenth decimal place is incorrect. The reason for this is that the rhyme was originally dreamed up by one Adam C. Orr of Chicago. Being American, his idea of how to spell 'rivalled' was not the same as the British one. Although Oliver is entirely accurate in his illustration that rivalled should have 8 letters, unfortunately the US spelling only has 7 - and that's what the 13th decimal place should have been.
The moral of this story is that if you are designing a numerical mnemonic in English, make sure that you don't use any words such as rivalled, or travelled, or colour, or labour where US and British spellings deviate - otherwise, someone is bound to be misled. 

See all of Brian's online articles or subscribe to a weekly digest for free here Find out more about Introducing Infinity here.

Don't knock our queues

Photo courtesy of Eva AmsenI was somewhat bemused by a thought piece by James O'Malley suggesting that we've got it all wrong with our pride in Britain's ability to put on a good queue, as exemplified by the queue for the late Queen's lying in state.
O'Malley tells us 'It’s also a completely bonkers, wrong-headed way to think about Britain. When we see a queue, we should feel embarrassed.' The reason for this, he suggests, is that a queue means that we aren't being dealt with efficiently - our precious time is being wasted. He goes on to say 'when we see a queue we should want to celebrate it - we should want to eliminate it.'
As someone who worked in Operational Research, and who has done quite a lot of work with queuing theory, I would respectfully suggest that he is wrong. Not in the suggestion that we should want to minimise queuing - of course no one wants to waste their time. But outside of a fantasy world we have to consider the reality that lay behind Macmillan's response to being ask what the greatest challenges faced by statesmen were: 'Events, dear boy, events.'
There were always be circumstances where resources are limited compared to demand - and in such circumstances, the queue in its multiple guises is often the fairest way to deal with the situation. The British pride in its queue forming is not in the need for the queue, but rather dealing with circumstances where there is that resource/demand imbalance in a good tempered and orderly way.
It's absolutely true that we should engineer queues out of existence if possible. I recently sang the praises of Marks and Spencer's new system where I can scan things, put them straight into my shopping bag and pay on my phone as I walk towards the door without ever going near a till. That's a queue successfully deleted. But you can't always do that engineering. In some circumstances this is because it takes time to engineer solutions and we may need the queue now. In others, the cost of the resources to remove the queue is greater than the cost of the queuing time. 
That's not to say that O'Malley is wrong in saying that virtual queues such as booking systems have a lot of value - but it's naive to suggest that one could have been set up for the scale and nature of queue required in the Queen's lying in state. O'Malley suggests we could have modified the Covid jab booking system. But this would have involved attempting to apply a massively multiple queue, multiple server system to a single queue, dual server physical reality. The only way a booking system would have significantly changed things is if it was used to drastically reduce the number of people who could attend, which isn't really helping.
So, accepting there are circumstances where queues are needed, I think it's perfectly acceptable to be proud of a culture where, when they are required, they form naturally and easily, rather than in some countries, where any pretence of queuing rapidly collapses into a free-for-all. I was particularly impressed a few years ago when I went to a row of three cashpoints with a good few people waiting. I'm not sure how O'Malley would envisage dealing with this kind of queue. Putting at least ten cashpoints in each location, most of which would spend all their time idle? Making people book ahead to withdraw cash? But the queuers were a beautiful sight to see. Instead of forming separate queues behind each cashpoint, with no guidance they formed a single queue, multiple server formation, providing the fairest division of waiting time. I was truly proud of them.

When will the Green Party go green on energy?

I despair of the Green Party here in the UK. They are still pumping out the same old knee-jerk reaction of the ex-hippies to nuclear power. It's as if they didn't realise how important climate change was. It's not just one item in a green buffet of options. It's the big one. 
We need massive change to deal with the climate emergency - and that includes moving to a mix of energy generation that doesn't produce greenhouse gasses. Yes, there must be plenty of wind and solar (and ideally wave/tide too) - but we also need generation that doesn't depend on the weather and sun - for which nuclear is the obvious option.

Caroline Lucas of the Green Party put out this tweet on 1 September 2022, and I don't know where to start.

Of course, a new nuclear plant is not 'the solution to the cost of living crisis' (though nuclear energy is a lot cheaper than the current price of gas). It's part of the long-term solution we desperately need to put in place to complete the move to zero carbon energy generation. There is no magic energy source that will deliver energy security with zero carbon instantly. We need to build the appropriate infrastructure. Of course we need immediate action to slash bills - and that isn't about building new power stations - but that's no excuse for putting of building essential infrastructure to give us energy security in the future. To make matters even worse - Caroline Lucas was saying exactly the same thing a decade ago - and from that perspective we definitely did need help in 10 years time.

Unless the Greens get their heads out of the sand, they remain a barrier to coping with climate change. They are as green as a company trying to get us to use more fossil fuels and should be avoided at all costs.

This has been a Green Heretic production.

See all of Brian's online articles or subscribe to a weekly digest for free here

The Gap in the Curtain - John Buchan ***

John Buchan is best known for his thriller The 39 Steps, filmed melodramatically by Alfred Hitchcock in 1935. The Gap in the Curtain is another period piece, first published in 1932, but it is anything but a conventional thriller. It's sometimes presented as science fiction, but it would more reasonably be described as fantasy: although the events it covers are supposedly triggered by the work of a scientist, the mechanism is pure fantasy.

We begin at a country house party, where a random selection of toffs are encouraged to take part in an experiment by the mysterious Professor Moe. By obsessing over a particular section of the Times newspaper for a while (plus the administration of a mystery drug), seven participants are set up to have a second's glance at a small section of the newspaper from one year in the future. In practice, two of the experimental subjects, including the narrator, don't undertake the final part, so five people are given a brief glimpse of the future. Some study a section of the paper relevant to their role - an MP, for example, finds out something about the future government - others seem to have picked part of the newspaper pretty much at random. The rest of the book then covers the subsequent year five times, once for the experience of each subject. 

So far, so good - a nice, high concept plot. Although I like the Hitchcock film, I've never read any Buchan before - I found his style, frankly, rather clunky. The country house setting is like P. G. Wodehouse without the humour, and the first section progresses distinctly slowly. Things get rather better with the five individual sections. In each case, although the predicted text from the Times is correct, the outcome is different to the one that the participant expects. I did particularly enjoy the MP's section - there was a strange similarity with 2022 British politics, with the country facing dire economic straits and some MPs who were clearly only in it for what they could get for themselves. Although the people were fictional, it was interesting to get a feel for what early 1930s politics was like as a Labour government struggled with the problems of the day.

For amusement's sake, the cover of my
1962 edition (no idea how I got itI did find it hard to warm to any of the characters - I love Wodehouse, and his lightweight protagonists are simply fun. Here, the entitled gentry, who make up pretty well all of the characters in the book, often feel repulsive to the modern viewpoint. I don't know if this was Buchan's intention, or whether he felt this was what upright British folk should be like. But this aspect wasn't too much of a problem as it really wasn't necessary to engage with the people - it's the plot that dominates.

All in all, though never a thrilling read (and the denouement of the last section was a distinct let-down), the period setting and the underlying idea made it well worth the read.

-- SPOILER ALERT --

I do want to moan about a couple of aspects of the plot, so if you don't want a spoiler, don't read further.

Two of the five active participants in the experiment focus on the obituaries section of the paper - and both see their own death listed. This requires a ridiculous pile-up of coincidences. Firstly, why would they choose to look at this part of the paper anyway? Everyone else looked at a topic that interested them. Why didn't they look at a sporting page or whatever? More to the point, what are the chances that, purely accidentally, both of them were predicted to die the day before the day the paper was published? There was no suggestion that the experiment was the cause of their deaths - they simply had a glimpse through the titular 'gap in the curtain'.

One of the two death cases was handled quite well. The person effectively scared himself to death, with a little twist where he avoided the way he 'should have' died, but still died. The second, though, involved the  subject escaping his fate. That wasn't a bad idea - but the way that Buchan made this happen was to have an obscure relative die on the predicted date, whose name was the same, whose regiment had been incorrectly reported and whose birthdate had been misheard and put in wrong. A whole pile of unlikeliness. How much more interesting it could have been if someone (the narrator, for example) had sent a fake death notice the Times (inevitably with various obstacles to this happening). The piece would have appeared, but wouldn't really be true. Managed well, it would have been a far better ending than the pathetic list of coincidences and accidents Buchan uses. Moan over.

See all of Brian's online articles or subscribe to a weekly digest for free here You can order The Gap in the Curtain at Amazon.co.uk, Amazon.com and Bookshop.org
Using these links earns us commission at no cost to you