Recent submissions

Comment · 4 days ago
i find this whole blog very dubious, with apologies to the author.

a bunch of people did psychedelics and then found that reality seemed more inclined to listen to their volition and respond to their intention or attention.

this was an n of some small number, and the original person who reported it could not reproduce it.

i don't believe this is good science. the test is self made by the subject while under the affect of the drug. the subject has a bias as well.

this is a trip report.

Comment · 27 days ago
well, you are the expert!

more seriously: how would it be able search for something it has no experience seeing at all

Comment · 2 months ago
what i would do for some formatting on this piece.
Comment · 2 months ago
in some way i guess thats always been an issue with the messaging on climate change and the issue in general.

its always a problem and its slow moving enough and cumulative enough that it will always be better to fix it sooner but the real dangers are many years off.

when the jet stream collapses overnight then we'll be forced to deal with it

Comment · 2 months ago
at the end of the day 'personal changes' as a fix for climate change are just another lie big corporations fed us to make it feel like the onus was on a personal level.

the amount of co2 that industries pollute and that various regulation would be able to change is multiple orders of magnitude compared to regular personal responsibility.

Comment · 3 months ago
seems like a pretty clear-eyed analysis to me.

we're rapidly approaching the paperclip maximizer

Comment · 4 months ago
i guess the point is that any fusion is better than any single model and a fusion of the top two models is obviously the best? for cost though i guess you could just duct tape together 10 open source models and then thats comparable?
Comment · 4 months ago
great game! well done.
Comment · 4 months ago
certainly you should understand the context of the comment then- the commenter was saying we had a 40 year moratorium on research because of the governments decisions at the time. There is a lot of research again now, but research takes time.
Comment · 4 months ago
its funny how long some people wait to decide the company they work for has lost their moral compass. "dont be evil" is already 8 years dead at this point.
Comment · 5 months ago
the market wins
Comment · 7 months ago
that explains why i kept hitting limits this week
Comment · 7 months ago
thats only because kimi 2.5 was trained using data stolen from claude. it wouldnt exist without riding claudes coat tails. none of the so called 'open source' models would
Comment · 7 months ago
quick make a derivative version of something people were hyped about for 5 minutes 2 months ago!
Comment · 8 months ago
its pretty clear that especially because AI have become worker-level tools that the negative externalities are passed to the consumer (worker). This means that we will be expected to keep up with an increasing pace of work because johnson in the cubicle next door is pumping out tons.

and better yet for the corporate overlord, the worker becomes more replaceable as well. their knowledge is less valuable and their overall skill is also less. they are in a sort of perpetual race to the bottom where they are just the meat monkey in the loop that is only valuable for keeping their agent(s) on task.

profits go up but the company knows in their darkest nights that they are also easily replaced. all of capitalism is just a house of cards waiting for the next model to pull the final card out from beneath us and then there will be truly nothing left except a handful of frontier ai companies who are fully integrated with government powers for all 'lawful purposes.'

the riots in the streets are easily quelled without a human in the loop.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton