one small village resists
no robots in the gaeltacht
This started off as one of my annoying joke opinions, but I then started taking it seriously – I often find interesting angles by taking other people’s metaphors literally so I guess the same thing shouldn’t surprise me with my own.
It’s basically a thought arising from the really clear and increasing danger of people cooking their brains with AI. We have the clear edge cases of Chatbot psychosis, and these are becoming really hard to ignore. There are stories in which it’s very clear that if a human being had been on the other side of the conversation, they would be looking at a criminal charge with a 14 year maximum sentence. Pharmaceutical products which had serious side effects like this get emergency withdrawal notices. And yet people suing the manufacturers seem to be treated like a curiosity.
Away from the edge cases, it’s also getting hard to ignore that lots of people in public life are getting weirder. Without reaching diagnostic levels, it’s a well known phenomenon that senior managers of all kinds have to be careful of the tendency for their judgement to get much worse and their idiosyncrasies magnified just because they’re surrounded by people who agree with them. (A while ago, I floated the theory that your “effective decision-making age” is your actual age, minus twelve times the number of months since someone last said “no” to you). And as several people have said[1] the chatbots and their ever-so-agreeable, tuned-to-maximise-interaction nature mean that now everybody can have the sensation of being surrounded by sycophants!
Industrialising a process matters. Things which can be ignored at small scale and in artisanal production start to become Big Bloody Problems when the difference between “marginal cost is close to zero” and “but it’s not actually zero” becomes significant.
And as the linked posts in the last paragraph argue, there’s a precedent for this, if we’re going to take seriously the idea that AI is a new industrial revolution. The last Industrial Revolution was great, you’ll get no argument otherwise from me. But it had a long transition period which was not so great. In particular, the industrialisation of food production was an absolute public health disaster, which was only noticed when whole areas of Scotland were no longer capable of supplying military recruits, because of a diet that had shifted from porridge and small beer to jam, tea and white bread.
So, there might be some similar transition period to get through, during which the AI is not yet able to replace human decision making, but where its abilities and widespread use are enough to significantly degrade our own. Which is problematic, because unlike physical health, if your mental health is getting worse you don’t necessarily realise it. This is something that’s haunting me at the moment, that there is something specially dangerous about the industrialisation of decision-making, because it’s bringing the costs and benefits of industrialisation in general into an area where they are very likely to undermine their own safety features unless very carefully designed indeed. And we are not doing much to make sure that they are carefully designed.
Which brings me back to the original joke. The future may belong to the Irish and Welsh. To Euskal Herria, to Malta and to Iceland. Simply because these are the languages which haven’t generated enough machine-readable text to train a language model on[2].
[1] (ie, enough that I’m too lazy to check up who I actually pinched this point from, sorry. Claim credit in the comments!)
[2] Actually, on this basis, you would probably also be backing an Indian future, as it’s a large country where the main everyday language isn’t the business language of the IT professionals and which has the additional protection of Unicode being annoying to work with.
In the world of engineering materials, there's a thing called "low-background steel", which is steel that was made before widespread testing of nuclear explosions lofted radioactive material into the atmosphere, contaminating everything made after 1945 with trace amounts of it. Low-background steel, scavenged from pre-war shipwrecks, is important when building shielding or enclosures for sensitive scientific instruments that need to detect trace amounts of radiation from an experiment, and not from the environment.
I feel like the AI boom is going to have a similar effect, there will be "low-LLM" cultural artifacts that we can be sure are not products of AI -- books and articles from before the 2020s, and as described here, uncontaminated languages like Maltese and Irish and Welsh and Icelandic.
Alas, India has now fallen prey to the University of Whatsapp in terms of misinformation, so I wouldn't bank on it...
The thing about tea, jam, and sugar is an interesting point on both ends--the average British person's diet got less healthy, if more addictive, while the crucial ingredients were supplied to Britain by slave labor, indentured servitude, and colonial exploitation.