The Braindump Blog

This week has been manifesto week for the biggest political parties in the UK, part of the lead up to the forthcoming surprise general election.

Links to the ones I found so far, in alphabetical order:

• Conservatives (80 pages)
• Green (48 pages)
• Labour (not easy to see how many pages it is in total)
• Liberal Democrats (117 pages)
• Plaid Cymru (72 pages)
• Reform: I don’t think they’re bothering with one. No comment. The closest I could find was “Our Contract With You” (32 pages).
• SNP - not sure, couldn’t quickly find any info on this one.

I do actually intend to read at least a few of them. Wish me luck.

In terms of my tendency to want to find a reason to vote for Labour, I’m wasn’t thrilled to hear in advance via The Guardian that their offering is going to be “a business-friendly manifesto targeted at former Conservative voters.”. Big eye-roll. But at least they have an amusing line on the Conservative plans, calling them the “most expensive panic attack in history”.

It seems there was a recent point in time where neither OpenAI’s ChatGPT nor Google Gemini chatbots appeared to be capable of giving an answer as to who won the last US presidential election.

I don’t think it’s that the robots joined QAnon, because they also wouldn’t reveal prior results of other countries' elections. But it’s clearly deliberate, further reinforcing the idea that if we’re going down the route of a world where people increasingly rely on these robot brains to inform them about what is true about the world then the creators of these products will have a disturbingly high influence on what facts will be considered worthy enough, or ‘safe’ enough, to reveal.

In the case of LLM chatbots and their continuing tendency to make things up, ‘safe’ probably has to mean ‘safe even if we give the wrong answer’, which I imagine (I hope!) is why the bots were programmed to keep their mouths shut on these matters given the prevalence of now AI-ingested source content available on the internet that is…misinformed.

This isn’t a new problem of course. Reference material that claims to answer questions about what’s going on in the world has always had more than its fair share of gatekeepers.

In practical terms, what people learn via e.g. Google search is very highly influenced by what results Google designs their engine to return first (and it’s increasingly a mess). But at least they present options, of a kind. The sort of one-stop-shop certainty and “you don’t need to look elsewhere” authoritativeness that the chatbots have been designed to project when answering questions I think potentially might make things worse. That said, of course, even the Encyclopedia Britannica can hardly be said to be free of bias; if nothing else then in the sense of what facts its creators deem as worthy of inclusion vs not.

In the past few days though it looks like ChatGPT has been altered so that it now will reveal the result - Biden won very definitively, in case you were unsure, even though Trump sure continues pretended he did - all 50 states, would you believe? (no).

Gemini still won’t respond to those types of query though, and instead suggests you use Google Search to find the answer - which these days I’m not entirely sure is always the most reliable way to find the truth of political matters either.

The BBC found a rare example of a Reform party candidate for the forthcoming UK election who doesn’t appear to have unhealthy levels of obsessive love for Winston Churchill. Meet Ian Gribbin.

The only problem is which political figures, past and present, he is infatuated with. As a friend noted, some of the Reform-tending folks seem to have new heroes now.

…Mr Gribbin posted on the Unherd website: “Britain would be in a far better state today had we taken Hitler up on his offer of neutrality…. but oh no Britain’s warped mindset values weird notions of international morality rather than looking after its own people.”

From a more recent point in time:

In the run up to the invasion of Ukraine, he praised President Putin, writing in January 2022 that he had “shown a maturity of which we can only dream of”.

And whilst he’s in love with whatever the vicious authoritarian violent dictator of the day is, he does manage to loathe at least the slight majority of the rest of the world’s population

‘Men pay 80% of tax – women spend 80% of tax revenue. On aggregate as a group you only take from society….Less complaining please from the ‘sponging gender’.

…

He added that women are “subsidised by men to merely breath (sic)”.

He’s seemingly pro-equality in some ways. But of course it’s the darkest version of equality imaginable.

“Square that inequality first by depriving women of healthcare until their life expectancies are the same as men, Fair’s fair.”

What about women who do put themselves in harm’s way in order to defend his way of life? Umm, no, not a fan.

In December 2021 he wrote female soldiers “almost made me wretch (sic)” and were a “total liability”.

By many measures, Reform are doing incredibly well in the polling. There’s no suggestion that most of their supporters share Gribbin’s opinion. But it is good to know something about where the folk we’re voting for are coming from, no matter who that is. For what it’s worth, it doesn’t seem like the party is considering taking any action against him.

The Gunning fog index is an algorithm that aims to indicate how readable your writing is by calculating the (roughly) the number of years of formal education you need to have had in order to understand the text after a single reading.

It’s quite simple, you can calculate it by hand. Based on Wikipedia’s description it works like this:

1. Pick a continuous passage of the text of around 100 words. Don’t include partial sentences.
2. Calculate the average number of words per sentence.
3. Count the number of ‘complex’ words. Complex words are those that:
   • Have at least 3 syllables. Do not count common word endings like -es, -ed or -ing as syllables. For example “created” counts as 2 syllables.
   • Are not proper nouns, familiar jargon or compounds of simple words such as “sunflower”.
4. Add together the average sentence length and the percentage of complex words
5. Multiply the result by 0.4.

Of course you don’t have to calculate it by hand. You can paste some text into this website amongst others and it’ll do the calculation for you.

It must be said that the various website options don’t all give you the same answer, which I’d guess reflects the slightly subjectivity needed when determining common word endings, compound words and the other exceptions to the general rule. Or of course faulty programming.

In summary, the formula looks like:

0.4 * ((count of words / count of sentences) + (100 × count of complex words / count of words))

As an example, the first few sentences of this post would thus be calculated as:

0.4 * ((113 / 9) + (100 * 11 / 113)) = ~ 8.9

The estimate is then that you’d likely was 9 years of formal education to easily understand it, making you, in US terms, a high school freshman.

It’s one of these things that can only ever be a rough approximation, and is presumably targeted only at the English language. But I can see it being of some use in checking whether what you write is vaguely suitable for the audience you intend it for, as well as providing some insight into what makes text readable.

It is only one of very many proposed formulae for calculating readability - it just happens to be the first one I came across. The Readability Formulas website gives a lot more info on a lot more formulae of that type, including an opinion on which to use when and an online calculator that can calculate several at once.

Conservatives halt digital ad campaigns after daily spend plunged

The party has halted all its campaigns on Google and Meta platforms

I mean, good for them for not giving these platform incredible amounts of funding any more. But it also feels like further evidence towards the increasingly less insane sounding theory that Sunak doesn’t even want to win the election.

A couple of years ago, back when British energy prices were skyrocketing to unprecedented and, for many, totally unaffordable consumer prices, I took some time to try and understand which appliances I should try to really cut back on if I wanted to save myself from future bankruptcy

Catherine from gocompare.com, a site that allows you to compare the prices of many types of products and services (although you probably already know that if you live in the UK - they make lots of adverts that feature officially the most irritating character on television: Gio Compario), recently wrote in to let me know about the energy cost calculator that they now have on their site.

The idea is that it lets you easily compare the approximate costs of running different devices for a given amount of time. For example, if you’re someone who uses the oven a lot and is considering options when getting a new one, it currently reveals that using an electric oven for 2 hours a day is likely to cost you around £1.04 a day vs £0.24 for a gas oven. Which, to go on a tangent, makes it not all that surprising that some of the ever-increasing numbers of folk that depend on food banks to stand a chance of getting their next meal are having to turn down some of what’s in theory freely available to them on the basis that they can’t afford to cook it.

The calculator’s values are only approximations of course. They don’t ask you to input any information about the specific appliance models you have or the terms of any specific deal you have with your energy supplier - although they note that most people are anyway on a standard variable tariff that will usually follow the maximum allowed by any fluctuations of the government’s energy price cap. The energy consumption per device data is taken from the Centre for Sustainable Energy (which is also worth a read!) and the prices come from a British Gas article.

So, especially given there’s not a tremendous amount of variation between most people’s energy costs per unit these days, the calculator is certainly a useful tool to give you a quick and easy understanding of what using your appliances is likely to cost you and, if you have options within a category, which choice is likely to cost you less.

Those of us concerned with the environment - which if you’re a human you should be - might also be able to use it as an indication of which types of appliances and devices are more energy efficient, and hence less polluting to use, on the basis that higher energy costs within a type of fuel obviously correlate with churning through more energy resources. Perhaps that’s a feature that they could add in the future if enough data is available; a more explicit indication of the energy usage in each scenario you try and its typical environmental implications.

Notorious liar Nigel ‘Vile’ Farage - never has a man with a greater or more desperate thirst for fame and validation been seen - has ‘changed his mind’ once again. He’s treating us to the horror-show spectacle of him standing as an MP again in the upcoming general election. Pity the poor voters of Clacton. Even if this will be Farage’s 8th attempt to become an MP, the first seven happily ending in nothing but failure.

Last month he ruled out standing as an MP for the weird billionaire’s favourite party Reform. This month of course he is actually going to stand as an MP for Reform.

At least this will provide some rationale for the propensity of the British media insisting on featuring him on an extraordinary, unconscionable, amount of election related programming - about an election that until this announcement in theory he was not really involved in.

I can’t help but think that the equally fragile Trump must have turned down the idea of sharing the spotlight with him. So instead he’s staying on our shores, more’s the pity.

On the upside, someone took the time to throw a milkshake at Mr Farage yesterday. Again.

I recently noticed Freegal, a music service that does free streaming and even DRM-free downloads for members of various IRL libraries. All 100% legal.

It is limited to a few hours of streaming a day and a handful of track downloads a week - perhaps library dependent but it’s 3 tracks a week for me - but it’s hard to complain about much given it’s free.

It’s also not quite Spotify or Apple Music when it comes to track availability. For example, the only Taylor Swift you’ll find is her entries on a few compilation albums. But they do apparently have a good few million songs to work your way through, old and new. There are definitely some reasonably mainstream 2024 albums available.

The couple of downloads I made so far arrived in 256 kbps MP3 format. I’m sure the purists wouldn’t be satisfied by that, but many of us can live with it. It’s substantially higher kbps quality than Spotify’s free plan gives you, which is a mere 160 kbps.

I did struggle to figure out how to extract the downloaded MP3 files themselves from the iOS app without having to link my phone up to a computer. This could well be a me problem. My easy workaround for now was to use their (mobile) website rather than the app.

In general, libraries have a ton of unexpectedly wide-ranging online services for their members that don’t require you to ever even walk through their doors beyond what’s necessary to sign up for free in the first place. At least in the UK, I’d assume the same is true elsewhere.

There’s been various iterations of digital offerings at my local library. Right now they’re offering free digital access to ebooks, audiobooks and magazines via Libby. Although that content is DRM locked so you do have to use their app to engage with it - unless of course you live somewhere where it’s OK to remove DRM from files.

Mark Logan becomes the third Conservative MP to flip his allegiance and join the Labour party. He crossed the floor last Thursday.

This is the third MP in just the last few weeks to do so. He follows Dan Poulter and the (controversial) Natalie Elphicke.

This is quite rare behaviour! Even when MPs do quit or get expelled from their party but don’t want to stop being an MP it’s more usual for them to stand next as as independent rather that switch directly to another party.

Apparently only one Conservative MP left their party before 1992. And the first defection specifically from the Conservative to the Labour party wasn’t until Alan Howarth did so in 1995.

It’s extra rare to see someone leave the currently-governing party to the main opposition. Alan Howarth’s 1995 defection was once again the last instance of that happening.

📚 Finished reading Lost in a Good Game by Pete Etchells.

Since childhood the author, Professor Pete Etchells, has deeply enjoyed playing video games. Some now have special meaning to him. And now he’s a Professor of Psychology and Science Communication he’s written a book that is an intriguing mix of tales of his personal gaming experience along with a more objective review of what the science so far tells us about the effects of gaming on humans, good or bad.

Gaming certainly seems to have brought him solace throughout some challenging parts of his life, which he goes into in detail. He lovingly writes about certain games; what they’re about, how he participated, who he played with and so on. Anyone who has played many of the mega hits over the past few decades may well identify fondly with these parts. Anyone unfamiliar with, or wary of, videogaming in general might probably learn a lot about the hobby - and perhaps be surprised at the wide range and depth of the full gamut of things we call computer games. It’s not all gem-laden Match 3 phone apps.

On one hand, as he mentions, there’s a school of thought that suggests his love of gaming might make it hard for him to research the topic objectively. But the opposing school of thought, to which he belongs, argues that in fact people who are experienced gamers can do better research because they already have the necessary background to understand the nuances of content and context involved within the activity.

Although not always thought of this way, games are a fairly unique form of art in that we are (often) in control of them rather than passive observers. This can give us a sense of personal investment and immersion that we don’t find in other artistic domains. Games might give us a sense of exploration, provide escapism, or connections to other people. It might vary by game, by person, and over time. All in all, they can be a very personal experience that leave us with memories of hopefully good times and on occasion maybe teach us something about ourselves.

We also learn a lot about the history of video games, which is longer than one might expect. The first computer that was capable of playing a game against a human was apparently “Nimrod”, which was exhibited in 1951. By the 1960-70s games were starting to move from the computer lab into arcades. Pong, the first digital arcade game release, came out in 1972.

He then looks into why we play video games. The upshot is 1) that different people play for different reasons at different times, and 2) it’s hard to get an unbiased view on those reasons. This uncertainty and unknowingness is a key theme to which we return to a lot when it comes to evaluating the science of gaming.

Perhaps the most commonly used classifications of why an individual plays games are those based on Richard Bartle’s work of 1996. Bartle observed four main groups:

• Achievers: players who aim to score points, collect treasures, level up their characters.
• Explorers: players who like to find out more about the virtual world.
• Socialisers: Players who aim to communicate with people who they share an interest with.
• Killers: players who like to fight or otherwise annoy other players.

But Bartle makes never claimed that these are particularly scientific. It’s just what he picked up after talking to countless players over the years.

More generally regarding motivations, Etchells touches on self-determination theory. That’s an idea from psychology that suggests our behaviours are driven by both intrinsic and extrinsic motivation. The intrinsic side of things is probably why we participate in most forms of play and sports, and hence is likely relevant to video games. Three key components of intrinsic motivation emanate from our human desires for competence, autonomy and relatedness - all of which (some) games can provide.

There’s a lot of confusion we need to concern ourselves about in the literature and beyond. Firstly what is the definition of a video game? Call of Duty? Candy Crush? Both? And if both, do we risk conflating too much when we talk about “video gaming”. Why should one produce the same effects as another? In which case, it’s questionable how useful it is to talk about “the effects of videogames” as a whole.

Computer games are also a newish form of art. Today’s grandparents, and possibly parents, we’re unlikely to have grown up playing them. This leads to a very understandable fear of the unknown, exacerbated by media scare stories. I think the author sees this kind of sentiment as the latest in a historical line of “these new things are ruining your children” moral panics.

We then learn about the bad scientific practices that infuse historical video games research - and indeed psychology in general. Conclusions emanating from bad science are rarely useful. This section is worth reading to learn more about what makes for good or bad science even if you have no interest in video games.

As well as statistical trickery, there’s also the fundamental questions of what research questions are asked. The author believes that the vast majority of research has pursued a very limited number of questions, mainly around the harms of video games and in particular their link with violence. A study will never report any positive effects of video gaming if it is not set up to look for them. A habit of critical thinking is to be encouraged.

In addition to any outright fraud, there’s a wide range of what have come to be known as questionable research practices to consider, some of which historically are extremely prevalent in psychological research. Here we’re talking about both individual-researcher controlled behaviours things like p hacking as well as more structural issues such as the file drawer effect and so on.

So what does Etchells conclude as being the verdict of whether video games harm us, make us violent and so on? What mainly comes across from this book is the sad fact that we just don’t know very much. There are certainly studies that claim to show harm. But there are other studies that claim to show no harm, or even some benefit. Many, perhaps almost all, of these studies are suspect, or at least incomplete, in one way or another. It’s not even clear that we know how to measure something like aggression in a way that we should actually care about in terms of negative real word outcomes.

His take, which makes sense to me, is that the mixed messages coming out of this body of work mean that it’s unlikely that the games studied so far have a huge effect either way on outcomes such as real-world violence. Certainly there is not nearly the ubiquitous evidence for harm you would assume there is if you read only the moral panic driven newspaper headlines on this topic. But this is of course not evidence that games can’t cause any harm to anyone in any circumstance. The author finds it likely that some people are more prone to harm from video games than others, and some more prone to benefit - the same as almost any other such activity humans get up to in life.

He is particularly concerned about the more recent “innovations” in gaming which have, necessarily, been studied less. Examples include the added immersion caused by virtual reality and the all-too-present prevalence of what are essentially addictive gambling mechanisms - or indeed actual gambling - that infuse many of today’s most popular games, especially on mobile devices.

In 2018 gaming addiction was classified as a mental health disorder by the WHO, although there is apparently much debate in academia as to whether this is a good thing or not. The author reports that there’s certainly no agreement over how prevalent this diagnosis is or where its boundaries should lie, which are problems in themselves when it comes to making sense of or acting on such a potential problem.

Games usually show up on screens, so one chapter in the book deals with the perils and pleasures of screen time. Sadly the same kind of conclusions on the science show up here as we saw in the critique of gaming science. There’s no universally accepted 100% definitive science showing that screen time is bad or good, and little in the way of reliably measuring or classifying it in the first place. Once again, perhaps not all screen time is the same or likely to have the same effect, any more than all videogames are the same.

He cautions us to think of screens as merely a tool. As with any tool we should learn to use them properly. We control them, not the other way around.

We later learn about how games can produce data that aids science, leading to more understanding about the human body and brain works. We end a chapter on e-sports, which even 6 years ago was a $900 million industry, and one on loss. That’s loss both of the personal type but also in the sense of how we could or should preserve games themselves. This proves to be a much greater challenge than simply putting some 30 year old disk in a safe somewhere and hoping for the best. Apart from any technical issues, there’s more to gaming and gaming culture we should want to share and remember than a series of magnetic 0s and 1s.

His final conclusion is not all that satisfying. But it is what we apparently have and hence should drive how we should think about the topic when we’re making our own decisions about gaming. Known unknowns are, after all, better than unknown unknowns.

Are games good or bad for us? The honest answer is that we don’t convincingly know either way, and it’s probably a bit of both…

…

As much as they’re a form of entertainment, they’re also a tool: one that must be treated with respect and responsibility, in the full knowledge that if used improperly or without due care and attention, they may cause as much damage as good.

📺 Watched Beyond Paradise season 1 and 2.

This is ‘Death in Paradise’, fully UK edition. It literally has the one of the same detectives - 2017’s Humphrey Goodman.

But by now, he’s moved from exotic tropical paradises to Devon in the south west of England in order to settle down with his girlfriend who’s in the midst of setting up a wine bar. Of course he also somewhat accidentally manages to retain his responsibilities of bumbling around being an awkward Englishman in charge of a local police branch. Which is again situated in a small town with an incredible crime rate.

The weather and the accents might be different, but there’s still a good amount of light hearted murder and other such cozy crimes to go around.

In comparison to the original, perhaps there’s a bit more variety of crime going on but also, at least in the first series, a bit too much relationship drama for my taste. For the most part I prefer to watch seemingly impossible crimes get solved by improbable flashes of magical insight over some poor soul’s marriage slowly falling apart. Didn’t stop me finishing it though.

I felt compelled to check. Happily, it turns out that if the now-certified-criminal Trump was to become US president once again then he could not pardon himself for the crimes he was found guilty of yesterday.

There seems to be some legal debate on the general topic of “can a president pardon themselves?”. But for these particular cases it’s quite clear that he could not. He was being tried by a state, New York, not the federal government. That means he would need the state to agree to pardon him. Seems unlikely.

From the US Office of the Pardon Attorney:

Does the President have authority to grant clemency for a state conviction?

No. The President’s clemency power is conferred by Article II, Section 2, Clause 1 of the Constitution of the United States, which provides: “The President . . . shall have Power to grant Reprieves and Pardons for Offenses against the United States, except in Cases of Impeachment.” Thus, the President’s authority to grant clemency is limited to federal offenses and offenses prosecuted by the United States Attorney for the District of Columbia in the name of the United States in the D.C. Superior Court. An offense that violates a state law is not an offense against the United States.

Of course that’s not to say he won’t just say he magically pardoned himself anyway. And there’s always the risk of untoward support from the worryingly activist Supreme Court justices. And as some of the many, many pending charges against him are for federal crimes, those ones might be a lot more at risk. But it’s not something he could (legally) do as-is with regards to yesterday’s verdict.

The same applies to any US president of course. For example Biden has no power to pardon Trump of yesterday’s crimes, even if for some bizarre reason he wanted to.

I just learned that Microsoft released some older versions of MS-DOS as open source.

Whilst it shows my age, I’m glad that my early computer experiences involved a lot of command line stuff, even if I recall the struggles against config.sys and autoexect.bat. Feels like it set me up for nerd success somehow.

Donald Trump has been convicted on all 34 counts of falsifying business records in his historic criminal trial in New York

From BBC News

This is the first time the US has ever had a former president receive a criminal conviction.

(Please let it remain ‘former president’; presumably he’ll still be the Republican nominee.)

The study behind the trial that Novo Nordisk stopped early last year for efficacy reasons has been published in the NEJM. And sure enough, it’s reporting that semaglutide (aka Wegovy, Ozempic) is helpful, even life-saving, for patients with type 2 diabetes and chronic kidney disease.

To date, chronic kidney disease has no cure, whilst affecting more than 1 in 7 Americans.

From Nature’s reporting:

…those who received semaglutide injections weekly were 24% less likely to have ‘major kidney disease events’, including kidney failure and dying owing to kidney complications

…

Participants who received semaglutide were also 29% less likely to die from heart attacks and other major cardiovascular incidents than were those who got a placebo, and 20% less likely to die from any cause during the trial period.

Of course it’s not the first time that this drug has been shown to save lives.

Seems it is impossible to delete your old Google Gemini chatbot chats if you have a Google business or education account. So be extra careful what you ask it I guess!

This seems…suboptimum. Even if there’s some weird legal reason this is universally necessary, beyond Google just wanting to leverage your data forever, it’d be nice to be able to delete them from the interface so as not to totally clutter up your workspace.

📽 Watched Civil War.

Set in a presumably near-future time - I suppose 2025 would be a reasonable guess - we see that the US has descended into another civil war. Not a figure of speech, but an actual war with armies, militias and the like.

A definitely-not-Trump President sits in the White House broadcasting to everyone about how he’s about to have the greatest military victory ever seen, everyone says so. Whilst elsewhere, several well-equipped but uncoordinated secession movements - at least Texas, California and Florida if I remember correctly - fight what’s left of the federal military, presumably to gain their independence and remake America into what they think it should be. Which is often a fairly unpleasant vision.

Anyway, we follow a team of war photo-journalists (Reuters and the NYT still exist in this world, at least to some extent) who set off to try and make it to the White House to try to photograph and interview the president for the first time in several months. Washington DC has been made basically impenetrable at this point. And traveling through the rest of a violent, impoverished, suspicious United States isn’t an easy journey.

There’s not a whole lot of backstory. Of course if you remember the January 6th invasion of the US Capitol building you might not really need one, although personally I was a little disappointed to not get to see the buildup and transition, political and social.

The film does seem to take some care not to pit an army of obvious goodies against one of cartoon villains; few people come out looking all that good or virtuous. One wouldn’t necessarily think a California secession movement would politically have a ton in common with a Texas one as it stands today.

In practice, it actually reminded me quite a bit of a non-zombie version of a show like Walking Dead. Very much an action film, albeit one with emotional resonance given the very well done depiction of a war torn US which, apart from being disturbing in its own right, wasn’t a million miles away from scenes one sees from the current horrific IRL wars in Ukraine, Palestine and elsewhere.

The Labour party’s chief of staff Sue Gray has created a risk register - less formally known as the ‘shit list’ - that details some of major problems they will have to face if they win the forthcoming general election as expected.

Any of these could apparently’upend the political calendar', irrespective of what the winning party actually would like to do.

• The potential collapse of Thames Water
• Public sector pay negations
• Overcrowding in prisons
• Universities going under
• NHS funding shortfall
• Bankrupt local councils

It’s not pretty. Maybe that’s why Sunak appears to be doing his best to lose the election.

🎭 Excitedly awaiting ‘Standing at the Sky’s Edge’, which claims to be ‘the best British musical in decades’.

A friend reports that it is indeed very good. Plus it has some personal novelty value.

It’s set in a big imposing block of flats in Sheffield called Park Hill, which I basically lived opposite to in earlier life. It was an imposing 1960s block of an architecturally brutalist nature. Sadly I never witnessed much in the way of singing or dancing emanating from that direction back then.

In fact, the flats were unfortunately a little infamous at the time. Although certainly not as bad as the area apparently was pre-skyscraper-construction in the 1930s, at which point it was spitefully known as ‘Little Chicago’ due to the excessive rate of violent crime.

Roundabouts the time I left the area the residents had mostly been dispersed. Unfortunately not all that willingly for some I understood back then, so perhaps ‘evicted’ is a better term in some cases Some fancy developer was busy enacting its vision for a bright new (I assume gentrified) future for the structure, although it being a listed building they couldn’t do too much to its overall size and shape.

Nonetheless, I don’t know what hallucinogens the re-designers had taken when constructing their modern day Park Hill mood board; but I think I remember some artists' impressions of what we were to supposedly look forward to featuring much in the way of rainbow colours and the occasional flamingo roaming throughout the grounds of their modern-day take on utopian streets in the sky.

Apple Music tempts fate / goes full engagement growth hack by releasing a list of the top 100 ‘greatest albums ever made’.

To save you clicking, the top three are:

3: Abbey Road by The Beatles

2: Thriller by Michael Jackson

1: The Miseducation of Lauryn Hill by Lauryn Hill

Predictably, certain sections of the internet explode in indignant rage at the absence of their preferred artists or genres (or at its least pleasant, that the number 1 album wasn’t recorded by a White man). Personally, I’m finding a few gems amongst them that I haven’t previously given a go, so it’s certainly not all bad.

I believe the list was voted for by a set of people deemed to have some above average insight into the topic, using these criteria:

• Albums that represented a cultural moment for the artist or genre.
• Albums that were complete thoughts, not just collections of hit songs.
• Albums that thoroughly represent culture in production and lyrics.
• Albums that inspired a generation to want to create more music.
• Albums that represent the BEST in storytelling, musicianship, recording and production.
• Albums that are timeless and reached far beyond the genre categorization.

But of course taste is a personal thing. No one else’s top 100 list is going to be yours. Not to mention that your favourites are probably not predicated on whether they ‘reached far beyond the genre categorisation’.

Apple is far from the only company to attempt such a feat. Rolling Stone has done a top 500 which they’ve updated on multiple occasions.

It’s quite fascinating to see how the ‘best ever’ list changes, or doesn’t, over time and over judging panels. The Pudding has made a fascinating interactive visual essay about this, highly recommended.

More former big hitter Tory stars are abandoning the sinking ship. Perhaps the most surprising to me is Michael Gove, who, just a few days after proclaiming ‘Who dares wins!’ in reaction to the bizarre announcement of a general election, apparently doesn’t dare and won’t win.

We’re now up to 80 Conservatives standing down rather than fighting the next election. Can we make it 100?! For what it’s worth, the comparable figure of Tory standowns before the 1997 Labour landslide general election victory was a mere 75.

If some of this is really due to the prospect of Labour banning MPs having second jobs as The Spectator suggests, well, double good riddance. Ruling the country really should be something you focus on for more than a few minutes a day.

with nary an umbrella in sight, announces that the UK is to have a general election on July 4th. The date seems symbolic of…something. Perhaps his Americaphilic nature?

In case you’re wondering what the background noise is around 2 minutes in, it’s anti-Brexit campaigner Steve Bray masterfully broadcasting “Things Can Only Get Better”, a song which for Britons of a certain age and level of political interest is vividly connected to Labour’s massive 1997 General Election victory, as well as, let us hope and pray, being true at a surface level. Good stuff.

I must admit being somewhat surprised at the timing. Was he drunk? Or was it really that there was one whole week of non-catastrophic economic and migration data and, horrifyingly for the rest of us, that’s as good as he imagines it’s ever going to get.

Apparently I’m not the only one. It seems many of his own Members of Parliament, even the big famous important ones, are a bit taken aback. There seems to be rumours, some coming via Nicholas Watt, of yet more submissions letters of no confidence about Sunak and/or lots of Tory big beasts deciding they’re probably going to stand down rather than face the likely humiliation of trying to get elected when, despite on occasion Labour’s best efforts to disappoint even those of us who want to find a reason to vote for them, you’re still 20+pp down in the polls.

In fact at least 67 Tory MPs have already announced their intentions to do so - mostly well before yesterday’s announcement of the election’s date to be clear - including Theresa May, Sajid Javid, Ben Wallace and Matt Hancock.

An incredible story from the New Yorker about the case of British nurse Lucy Letby, a nurse - nicknamed the “angel of death” by some - who was convicted last year for the murder of 7 babies that were in her care, and the attempted murder of another 6.

But what was such a serious conviction for such a sickening crime based on? Well, if one takes the New Yorker reporting at face value then…honestly not very much. The main piece of evidence appears to be statistical in nature - basically she was in the right place at the right time, or at least most of the places at most of the right times - and highly subject to the Texas sharp-shooter fallacy.

I have no take on whether she’s guilty or not - how could I? - but the evidence seems absurdly lacking if one takes this story at face value.

Although anyone in Britain will struggle to even see the story as that page of the New Yorker website has been blocked from being accessed by any UK users due to a court order.

MP David Davis is not pleased about this, suggesting that such a block is “in defiance of open justice”. As is always the case with these sorts of blocks, the effort in any case seems a bit weird and ineffective in these days of the global internet. It’s easily viewable via a VPN or on one of the many archiving sites. This for instance is the sort of link I’m sure none of us would ever contemplate clicking on. Or, you could just buy the paper version of the New Yorker.

The randomizr package for R (and apparently Stata) provides some nice simple functions to help automate the process of randomly assigning participants to groups in for instance randomised controlled trials.

Common designs include simple random assignment, complete randomization, block randomization, cluster randomization, and blocked cluster randomization. randomizr automates all of these processes and assists scientists in doing transparent, replicable science.

Last week OpenAI, a company who never shies away from inventing something at least adjacent to the torment nexus, demonstrated their latest generative AI model: GPT-4o. The “o” stands for “omni”; whether this may be in the sense of omniscience, omnipresence, omnipotence or omnibenevolence (or all 4) is left as an exercise for the reader.

TLDR: It’s faster, cheaper and can respond to text, audio, image, and video inputs with text, audio, and image outputs in a much more “human” way.

Here it is in action:

It is of course amazing and magical, as is much of the stuff that they produce. But I’m not sure it’s entirely wise.

I can see use-cases for models that can infer something about a user’s emotions, although I would worry about even that given the emotional state of users is something that mainstream social media networks actively try to exploit.

But is it really necessary to give the model the ability to sound like it has emotions? When giving vocal responses it laughs, it jokes, it acts curious and interested, like it cares. Sometimes it seems excited. It flatters, it teases. But why? To obfuscate its artificialness?

I’m not convinced that a primary goal of AI development should be to try and produce products that are increasingly indistinguishable from humans. It’s not like the crew of the Starship Enterprise ever seemed to complain that the ship’s computer was too unemotional. The one time it did end up programmed with an emotional-seeming “personality”, Captain Kirk turned it off as soon as he could.

Once again, perhaps there are legitimate use cases for this - other than to run scams on a mass scale. But I suspect they are rare, should be carefully studied in advance, and aren’t an essential component of providing most humans with the potential everyday benefits of this technology without introducing an unnecessary layer of confusion.

It’s the least original observation available I know, but the OpenAI performance seems rather modelled after the movie Her. This film was in fact created after Spike Jonze read an article about a website that let you chat to an AI program - over a decade a go! Since then we have certainly seen people develop strong emotional attachments and possibly even love for some AI models far less “omni” than GPT 4o.

I dare say that for now the OpenAI verison will have more restrictions than “Samantha” did - but it seems similar in mode and temperament at least.

I’m sure it’ll only be a matter of time before someone runs Theodore’s lines from the script of Her through GPT-4o. I can’t say I’d be incurious about the results. Hopefully a bit more “As an AI model I cannot….” will be involved.

Here, for the committed, is the full 26 minute announcement of what’s new in the world of OpenAI: