Dec 10, 2022·edited Dec 10, 2022Liked by Ian Leslie
Very interesting entry. One of those that will stay with me. Two thoughts:
1. The rise of "lines to take" for all manner of public facing entities (e.g. politicians) is surely in part a manifestation of the rise of data-driven/modelled decision making - to the extent the lines are formulated based on models of how messaging will be received. Maybe this leads to a similar effect as with Marvel films (i.e. perfectly satisfactory, but intentionally not introducing new or risky ideas), but in more obviously consequential terrain for wider society.
2. There are likely to be interesting second-order problems of relying on models to do our thinking for us e.g. a) we may become worse at making decisions where we can't form a useful model (because e.g. the situation is in some way unique or contains too many unknown or unidentified variables to be sufficiently tractable for formal modelling), or b) recognising when the model we're relying on is either simply wrong or being used wrongly (especially if we are partly using a model-driven approach to a skip debating what assumptions should go into it).
What an expansive, interesting, fun and imaginative piece of writing. I loved the section on key changes and bridges in particular. What’s in your coffee and where can I get some. Great stuff Ian.
Another lovely article, although I did have a minor tic when I read:
"Machine learning programs work by labelling objects and assigning them to categories. An image in a dataset either contains a car or it does not - it can’t be car-ish."
Au contraire, many ML classifications models do work like this - they assign probabilities to various outcomes. An image can be a bit car-ish, a bit dog-ish.
I do get your point about modern life reducing the space for nuance.
This post sat in my mailbox for a while before I got to it - glad I didn't leave it languishing longer. It's a great piece of writing. I don't have any great insights other than to say it resonated with me. Thanks, Ian.
The question of how to avoid becoming “structures, abstract entities with all the messiness of humanity scooped out” made me think of Hannah Arendt and what she wrote in the Human Condition regarding labour vs ‘action’. For her the way out of the predicable patterns of human behaviour that arise from labour and satisfying out basic necessities came from action, where freedom and the miracle of something truly new happening were possible. I think her fear was that we were losing not the capacity to act, but the opportunity, with labour in a society of consuming jobholders becoming the greatest good, with all else enslaved to this aim. The voice you describe in this article sounds similar to me as the ideas she develops for action, where it is only through action, and by being acting beings, that we reveal truly who we are as opposed to merely what we are.
The thing for me is having written this, I still feel scared I have no idea what she’s talking about and all I’m doing is pretentiously parroting what I’ve read. I loved the Human Condition, but the concept of action was so utterly alien to me it makes me wonder what our society has already become sixty years after she wrote it (although it could also be I’m also just not very well educated!). But if she was in any way right, it does feel like we are slowly renting out any hope we have for action and freedom to AI and models, and that we’ve lost sight of any of the principles (such as honour or glory or even solidarity) that Arendt thought inspired action.
Fascinating as always, and one thing in particular struck me. You say "everyone must be either pro-woke or anti-woke", but then go on to define a third state of woke-sceptical which, at least on first reading, seems to be the one that most people I speak to in real life would identify with. Supportive of the principles that underpin the Deadly Woke Agenda, but wary of the quasi-religious fervour that surrounds it in online discourse.
I'm not sure that changes your conclusions in any real sense, other than perhaps to offer a small chink of light that - maybe - those human qualities still exist in abundance, they're just harder to articulate
Hello Ian - thank you for another thought-provoking piece. On the change in pop music, I thought you might enjoy this thorough and enlightening analysis from The Pudding (which also proffers all kinds of other fascinating attention wormholes) https://pudding.cool/2017/05/song-repetition/
And on the relentless desire to categorise and label, and the loss of nuance, I wondered if you'd read (and had thoughts on) Iain McGilchrist's 'The Master and his Emissary'?
Right on time and on the nail. Excellent post. Thank you.
Very interesting entry. One of those that will stay with me. Two thoughts:
1. The rise of "lines to take" for all manner of public facing entities (e.g. politicians) is surely in part a manifestation of the rise of data-driven/modelled decision making - to the extent the lines are formulated based on models of how messaging will be received. Maybe this leads to a similar effect as with Marvel films (i.e. perfectly satisfactory, but intentionally not introducing new or risky ideas), but in more obviously consequential terrain for wider society.
2. There are likely to be interesting second-order problems of relying on models to do our thinking for us e.g. a) we may become worse at making decisions where we can't form a useful model (because e.g. the situation is in some way unique or contains too many unknown or unidentified variables to be sufficiently tractable for formal modelling), or b) recognising when the model we're relying on is either simply wrong or being used wrongly (especially if we are partly using a model-driven approach to a skip debating what assumptions should go into it).
What an expansive, interesting, fun and imaginative piece of writing. I loved the section on key changes and bridges in particular. What’s in your coffee and where can I get some. Great stuff Ian.
Another lovely article, although I did have a minor tic when I read:
"Machine learning programs work by labelling objects and assigning them to categories. An image in a dataset either contains a car or it does not - it can’t be car-ish."
Au contraire, many ML classifications models do work like this - they assign probabilities to various outcomes. An image can be a bit car-ish, a bit dog-ish.
I do get your point about modern life reducing the space for nuance.
This post sat in my mailbox for a while before I got to it - glad I didn't leave it languishing longer. It's a great piece of writing. I don't have any great insights other than to say it resonated with me. Thanks, Ian.
The question of how to avoid becoming “structures, abstract entities with all the messiness of humanity scooped out” made me think of Hannah Arendt and what she wrote in the Human Condition regarding labour vs ‘action’. For her the way out of the predicable patterns of human behaviour that arise from labour and satisfying out basic necessities came from action, where freedom and the miracle of something truly new happening were possible. I think her fear was that we were losing not the capacity to act, but the opportunity, with labour in a society of consuming jobholders becoming the greatest good, with all else enslaved to this aim. The voice you describe in this article sounds similar to me as the ideas she develops for action, where it is only through action, and by being acting beings, that we reveal truly who we are as opposed to merely what we are.
The thing for me is having written this, I still feel scared I have no idea what she’s talking about and all I’m doing is pretentiously parroting what I’ve read. I loved the Human Condition, but the concept of action was so utterly alien to me it makes me wonder what our society has already become sixty years after she wrote it (although it could also be I’m also just not very well educated!). But if she was in any way right, it does feel like we are slowly renting out any hope we have for action and freedom to AI and models, and that we’ve lost sight of any of the principles (such as honour or glory or even solidarity) that Arendt thought inspired action.
Fascinating as always, and one thing in particular struck me. You say "everyone must be either pro-woke or anti-woke", but then go on to define a third state of woke-sceptical which, at least on first reading, seems to be the one that most people I speak to in real life would identify with. Supportive of the principles that underpin the Deadly Woke Agenda, but wary of the quasi-religious fervour that surrounds it in online discourse.
I'm not sure that changes your conclusions in any real sense, other than perhaps to offer a small chink of light that - maybe - those human qualities still exist in abundance, they're just harder to articulate
Hello Ian - thank you for another thought-provoking piece. On the change in pop music, I thought you might enjoy this thorough and enlightening analysis from The Pudding (which also proffers all kinds of other fascinating attention wormholes) https://pudding.cool/2017/05/song-repetition/
And on the relentless desire to categorise and label, and the loss of nuance, I wondered if you'd read (and had thoughts on) Iain McGilchrist's 'The Master and his Emissary'?
This one is important! Great stuff
Well, I try my best to be just like I am
But everybody wants you to be just like them
They say "Sing while you slave" and I just get bored
Ah, I ain't gonna work on Maggie's farm no more
- Bob Dylan