"As a writer who makes his living on the internet, here’s how my purchasing thought process went: the more serendipity, the more ideas—the more ideas, the more publishing—the more publishing, the more views—the more views, the more money. $59 for more ideas? That’s a no-brainer investment."
For myself, , my critical path is not ideas, I have dozens of essays planned, the bottleneck is the time to write and revise. I am hesitant to adopt AI tools in that regard.
So many things to say here as someone who has been obsessed with these speculations since the Google Glass program.
Let's start with Ted Chiang (and Nicholas Carr) and ask some questions about what augmented eidetic memory will do to human cognition and interaction.
This issue has been treated by “The Lifecycle of Software Objects”.
This is the firehouse of information problem: you need to drink clean water and it is refreshing and good, but you also must drink it one sip at a time: a firehose blasting into your mouth is not just sub-optimal, it’s harmful.
In the abstract maybe it sounds good to record these details of the every day, but in practice you’re just filling the fire truck with untold gallons that no one will ever be able to absorb -- except a LLM
So then you have to be conscious of what you’re really doing, which is converting your whole life into training data. Maybe you think: if LLMs will have power in the world, I want my experiences to be part of the data that shapes that power: okay, fair enough. But what do you give up in exchange?
My experience tells me that the nature of my life is different depending on whether I’m “doing the things” on record, for others, or for me and my family and friends. The quality and substance and subjective feel of life is inferior when the camera is clicking or the recorder recording or my brain judging the unfolding events for share-value, with this last being the worst.
So just knowing that my life has pressed the “record” button is enough to drain it of some of its quality and subjective value -- and the value of the data I would be creating would be not just very tiny, but too great for me to gulp, and therefore the gain of only some “AI” company built on the toxic mud from the mining of rare earth minerals, low-paid human labor, and clouded by the carbon it takes to run the machines.
And as we know, those machines need a LOT of water -- they will gulp and gulp at firehose speed and not leave us a sip.
I wish I had conversations worth transcribing :D but yeah, I agree, that would be exciting thing to try, and I am surprised it’s that cheap, relatively
Also. Has the world forgotten Pleo the Dinosaur?
Here's a piece from 2011: