club kid deserves all the awards. its the most moving film i’ve seen all year, and its so deeply emotional in so many ways.
taking a step back and appreciating the gorgeous art in our lives reminds me of the privilege of being human.
@gitghxst i’ve acknowledged that the way my non-tech friends use AI at work is 3 years trailing, the way my SWE friends use AI at non-tech companies is at least 18 months behind.
@gitghxst my instinct told me that my muse wasn’t creative enough, so i had claude make a duet for us to play but there weren’t enough dots to go around. i settled for doing tasks instead.
@v3rnolo attention is a zero sum game. the capture of attention is already amongst a consolidated group of platforms.
do you think the future of the attention economy continues to democratize further or simply continue centralizing as most efficient capitalistic markets do today?
agents finish their task at a faster rate than i can review their output, digest, and re-iterate.
i’d say request for decisions when i lack the context and information to make them, thus review time is slowed by my own ability to absorb and decide together.
parallel agents compound this as i tend to give attention to agents doing low hanging fruit and tasks i have deep context on, over collaborating with the agents working on harder problems
as context switching demands more of us, i feel we are accelerating toward even shorter attention spans.
managing 20 agents has reduced my attention span far more than short form content ever has.
i’ve been tracking my time at work recently, on task percentage etc.
it’s been eye opening to see how often and frequent i get distracted.
some observations:
- on a good day ill hit 50% of my non-meeting time on things i intended to work on
- most of the time the things i intend to work on still occupy the largest sum of time spent
- the number of threads in my day grows as the number of parallel agents i am running grows
closing my laptop gets harder each day as the number of agents i have running grows. moving things to remote harnesses helps but there are some things that i still start locally.
@maiab the distinction between agentic discovery and purchasing hasn’t been made clear yet. agents are pretty good at finding things i probably wouldn’t have but if i know what i want it’s almost always faster to buy it myself. chat interfaces as a checkout ux still sucks.