It's better that we have them than not, but I do think line graphs and the empirical analysis they represent sets in motion deeply irrational thought patterns in humans. Sometimes it's because one is representing something very complex as a simple scalar and it only matters that it goes up and down. Sometimes it's because it's inviting you to tell stories, to form expectations about what will come next, ie what's further down the time axis, or what a particular piece of past history "means".
Capital takes this from being simply something humans are funny about to being a battleground for our shared notions of what future humanity is headed for, and what futures we should want to work toward. And I think a critical consciousness you must form, to avoid being misled constantly, is to understand which part of a given curve you are looking at. When someone has a vested interest in the line going up forever, they will do anything in their power to get you and everyone else to believe that a past period where the line went decisively up is the important indicator, a story that will continue happening indefinitely into the future.
We saw this quite vividly in the last crypto bubble, where shady finance bro interpretations of candlestick charts congealed into what was pretty clearly a kind of mysticism. And we're seeing it again with LLM hype: ~3-5 years ago a few companies threw a lot of capital and data at some ~10 year old research and there was this new product that didn't exist before, and it has completely broken a lot of peoples' brains. Suddenly something as fuzzy and baggagey as "intelligence" became this line that we are arguing and speculating about. And the investors have dropped a cinder block on the accelerator pedal, they need their ROI and don't care what absurd shit they have to tell everyone to get them to believe it. And obviously, the truest deepest social danger is not what some shiny new piece of code can or can't do, but what those capitalists will do to everyone once they have the power to. I think what we need to cultivate is a sense of informed skepticism and a willingness to push back on narratives that exist primarily to advance a financial interest.
The line isn't magic. Stare at it hard enough and it will tell you whatever story you, or someone looking to exploit you, wants to hear, instead of the stories that we might actually want to have directing the future of our society.
Remember "Big Data"? It was a huge deal five to ten years ago. The idea was that the more data you could collect, the better decisions you could make for your business and even predict the future behavior of your customers. Despite this idea never being proven, investors were pouring billions of dollars into companies to grow users at any cost. Because the more data you have, the bigger your valuation. But the only people actually making money in this gold rush were those selling shovels and beer, i.e. boring middleware providers.
Anyway, the top of the S-curve of "Big Data" dovetails nicely into the bottom of the S-curve for "artificial intelligence." What if the problem is just that we need to find a better to process all the data we already have? I'm sure this won't have any negative externalities at all. 😊