Social Media's evolution was inevitable-repost from 2018

Please stop teaching psychology

…  or why Darwin and Dawkins are always right.

This article from 8 years ago seems more relevant now and introduces the

idea of a psyhcological-niche obeying the laws of natural selection evolution.

It could also be titled 'don't blame the techies for your mess'

In my role as a Computer Science teacher I needed to prepare some classes for my students about social media. It’s in the syllabus, it was in the news; Facebook and data harvesting being the media’s ‘mot du jour’ . So I ‘did’ SnapChat, Instagram, Twitter, WhatsApp and Facebook. But what struck me was I hadn’t got much to teach my students. The computing stuff was old and simple; TCP/IP comms, a bit of up to date compression and optional encryption; it was a bit all 1990. As a techie, its core was of little/no interest.

So who is interested in this stuff? I don’t mean the users of social media, obviously they are interested, but which ‘gods’ are watching over these applications?

A few years ago I looked at the UK graduate stats by subject. Without too much dissection it is safe to say that psychology graduates rather outnumber computer graduates by a factor over tenfold and much more than that if you only count females. Psychology is now rather hidden in the official stats as it is bundled into a category ‘subjects allied to medicine’. I suspect it was becoming embarrassing to discover how many budding ‘shrinks’ were in the pipeline.

So, take a bit of simple tech ( remember MS Instant Messaging, Skype?) add a generation of business and psychology graduates and what do you get?  Social Media, the love child of a million psychology graduates looking for a job in selling anything from plastic frogs to conspiracy theories.

What does psychology qualify you for with regards to selling? … manipulation is the obvious answer. However to manipulate effectively you have to know your target. To know your target as an individual does not scale well, so called micro- targeting works really not on individuals but  on quasi-homogeneous groups. It’s so much more than your browsing history and postcode

Here is how it’s done. Below is a homage to the great Richard Dawkins, author of the Selfish Gene and creator of the gene’s abstract replicator, the meme.

Social media platforms have something in common, the ability for a communication to be approved by a group or deprecated; more familiarly ‘liked’ or ‘disliked’. This is profound and if we borrow the logic and terms of evolutionary biology we can do the following steps:

A post, (maybe a picture/movie or an opinion) gains in value if it is liked and loses value if disliked. The direct analogy is with a selection pressure deriving from an environment. So an ecological niche corresponds to the social media group ( It may be big or small and highly specialised or less so just like nature). Being ‘liked’ is a bit like having the right beak to open the local nuts.

If we regard a post not just as one post, destined to survive or die, think of it as an instance of the user themselves; an abstract organism that is made manifest via it’s posting-phenotype. A subsequent post is actually a reproduction from that user. It’s not the same post obviously, but it inherits properties from earlier posts. ‘Liking’ a post makes it fitter to survive in its environment and subsequent posts will then inherit characteristics that caused it to be liked. Ideally being ‘liked’ even more becoming a more valuable property of the organism.

Vice-versa for dislikes you may think but not quite so. Consider a group of say 20 folk, a given post you notice is liked by 15+ on average, but consistently 1-3 dislike it. They do not have a selection pressure, they simply get ‘unfriended’ or whatever the term is in the software. If 15+ quite often don’t like your post then you must change or migrate out of the group. The latency built into the system increases with the size of the group but the principle remains true.

However you do it within your ecological nice ... psychological niche is the right phrase, ‘fitness’ to survive reflects an accretion of adaptations. And as in Nature, the drive is towards homogeneity, but without sex to stir the pot or an unyielding indifferent environment to provide external limits to change. A psychological niche is user-defined and so has no limits only homogeneity.

Now we have it, a group occupying a psychological niche can be characterised by the watching ‘gods’ through a process called data-harvesting. Once you know the psycho-nice well enough you can manipulate it, sell to it.

So that is where all those psychology graduates went. All they demand from the techies are more tools to harvest data from daft quizzes, geospatial data, face recognition and so on. Then it’s back to the ‘how the brain works’ books on behaviour and job done.

Micro-targeting bad examples: don’t try to sell ex-Nazi era thermal underwear to folk living in sunny Israel nor books on liberal democracy to an ISIS whatsapp group. They won’t buy.

So don’t blame the computer scientists for the evils of social media, the tech is not rocket science or even new. Blame an army of psychology graduates looking for a job who took our IP protocol and made a monster. All of this by the way is fully predicted by Scott Adams in his 1999 book The Dilbert Future: “Thriving on Stupidity in the 21st Century”. 

Statins and Mitochondria update.

 

I've posted about the effect of the cholesterol synthesis blockers known as statins on mitochondria before. This week though a couple of updates on the national news seem worthy of comment.

The first was a a publication from the University of Columbia detailing the source of the muscle pain side-effect associated with statin use. The work was very detailed at the molecular level and demonstrated clearly that the pain was associated with in infux of calcium ions into muscle cells. 

The second was a finding that  most of the very many side effects of statins as described on the crib-sheet supplied with the drugs were actually not attrubutable statins at all. The muscle pain side effect though was.

The above findings are interesting at the scientific, the linguistic and legal levels equally. The first finding regarding the influx of calcium ions is important because it is well known that mitochondria rapidly and unselectively absorb these ions, swell and may burst as the outer membrane gives way. 

Nearly fifty years ago my first paper showed that mitochondria became more fragile to swelling with age. Then we did not know so well that their rupture and release of weakly bound Cytochrome C ( this binds less tightly with age) triggers cell death ... we know it well now and it is the source of muscle wastage (sarcopenia) in old age. So, Columbia's research shows why a side effect of statins is muscle pain but it's findings infer that all mitochondria  will swell as a result of the drug. It's just that it is not always fatal to the cell! Do this to damaged, senescent mitochondria in aged cells and the cell will die as the mitochindria rupture.

The second finding's announcement was laughably 'straw-man' and linguistically manipulative.  The imaginary side effects listed on the drug sheet turn out to be, well, imaginary so can be crossed off,  and the confirmed side-effect (see above) is very rare so please ignore

... so the authors conlclude that the drugs are actually 'safer' than we thought and their role in tackling 'bad-cholesterol' is an even better option for now reassured people. 

Outrageous, my dim GP will no doubt be persuaded by this Pharma's ChatGtp bot generated blurb but this week's message is actualy the opposite. It's that statins cause severe change in calcium flux in muscle cell and this side effect is confirmed as real, ...  and the reasearch provides a clear causal path to sarcopenia in the vulnerable populations prescribed the wonder drug.

To get the message above though you will need at least a PhD in mitochondrial biochemistry and a working ability detecting thematic meme reasoning. 

Or in the opposite scenario, to be reassured, you will need a) a trust in big Pharma struggling to sell its wares, b) faith in your GP's scientific knowledge and c) an admiration for the Mandelson level of manipulation reached by generative AI