Don Piano

Real AGI would just be a person. As in: Jeff from accounting, but artificial. Not particularly dangerous, but also very much not on the path of LLM development on account of being qualitatively different.

So… yeah.

If folks find the tools useful, by all means stop.

Wasn’t it evertim?

To be clear: the direction I’d like to see isn’t ignoring larger-scale changes but embracing that these things are linked. Companies don’t burn fuel for fun, but for profit (or non-capitalist modes of resource allocation - if the central party committee decides to satiate the people’s hunger for meat and cars, that’s also a problem). And the profit there comes from all of us, individually as well as collectively. So action against that probably should also happen on both levels.

I’m being a bit annoying about it because the companies don’t burn all that crap for fun but, as you laid out, for our collective consumption patterns. I developed the impression that the whole “x companies do y% of emissions!” thing, similar to “no ethical consumption” reminders tends to fulfill a function not aimed at motivating larger-scale changes (e.g. banning animal agriculture wholly instead of making an individual choice to not consume em; banning ICE cars from being produced/sold while creating comprehensive public transport instead of merely biking to work yourself) but at detaching oneself from the role we do actually play in society. (Also, smaller/individual scale weirdoes are a good source of activists that can radiate social structures out into general society)

No, why do they produce all the emissions?

Huh, odd, why do they do this?

I did that last week, I’m pretty sure I put the resulting sign on my door

No. It’s human output.

Or you could pour river water down your socks, it’s quick, it’s easy, and it’s free!

See also: Skoptsy

Friend Computer is your friend. Don’t allow the mutants and secret societies to tell you otherwise!

Correct, some contracts cannot be done. Depending on the specifics, the illegal parts of the contract are unenforceable, or the entire thing is null and void. Employers often hate this, because it limits the exploitation they can enact upon employees.

https://en.wikipedia.org/wiki/Weber–Fechner_law

No, it’s not. Standard error of the mean estimate is SD/sqrt(n-1), so it depends on the signal/effect to noise ratio how much you can learn from it. SE of the mean difference for a between subjects design is slightly different but not miles off from that. You can squeeze even more info out of it by using within subjects designs.

What makes you think a sample size of 50 is irrelevant?

You used a Germanism. Do you have a Betriebsrat?

Also, don’t make introversion your entire personality.