> Wikipedia quotes a paper saying people make / recycle around 50 kilograms per day. That would imply that people make around 0.5787 grams per second. But this is in tension with the idea that people use 100 watts at rest. Since that 100 watt number seems to be more strongly established, I think 1 gram per second is a better estimate.
The cited paper in Wikipedia might be accurate and the author is just expecting "at rest" to mean awake but mostly idle. I'd wager that during sleep, we use less than 100W. For the sake of argument if you take out 8 hours (asleep time) from a 24 hour period, you end up with a rate of ~0.8681 grams per second, much closer to 1g/s. Anyway, just a little fun with numbers.
Yeah, I'm starting wonder if this is bs. They don't have a twitter account like the pledge form indicates. The youtuber that purportedly started the campaign only has like 36k subscribers, so it's not clear where such a high level of pledges would be coming from and the rpc endpoint response vs page count is pretty suspicious.
To make the analogy proportional (according to the article), the difference would be something like "do they notice the difference between 1h 40m and 5 hours" (i.e. 3x more). My money would be on yes, they'd very much notice.
> The most common category is that the models make wrong assumptions on your behalf and just run along with them without checking. They also don't manage their confusion, they don't seek clarifications, they don't surface inconsistencies, they don't present tradeoffs, they don't push back when they should, and they are still a little too sycophantic.
Does this not undercut everything going on here. Like, what?
> You might be thinking "just disable updates, man" or "just install LTSC", or "just run some random debloat script off of GitHub". Why? Why would I jump through all these hoops? I'd rather put in the effort for an OS that knows what consent is and respects me as a user.
The absolute choice quote here. Tattoo it on your forehead.
Ah, yeah, I'd suspect they'd only have plugins. Def understand the need for mobility. I never want to miss a chance to share with folks about the existence of EPTL, though. :P
They have 5 electric leaf blowers available right now. You mentioned you already bought one, but perhaps you could share this info with some of your friends or neighbors to save them the cost. :)
> An effect that’s being more and more widely reported is the increase in time it’s taking developers to modify or fix code that was generated by Large Language Models.
And this is where I stop reading. You cannot make such a descriptive statement without some sort of corroborating evidence other than your intuition/anecdotes.
I'd wager it's the first (51.1%). Mostly because I simply cannot imagine that nearly 70% of kids in 2005 enjoyed reading. That seems too high. Based on..? My vibes.
I don't want to beat a dead horse, since sibling commenters have covered this, but I'd implore you to imagine the spectrum of reactions which Meta _could_ have had when discovering their research indicated they were having a negative impact on people.
Some of those reactions on that spectrum would lead to greater human flourishing and well-being, others of those reactions would lead to the opposite. Now think about the reaction they actually _did_ have. Where on the aforementioned spectrum would their actual reaction fall?
Zooming out, how have they reacted to similar circumstances in the past when their own internal research or data indicated a negative impact on people?
The continued "outrage" is that they've exhibited a recurrent pattern across myriad occurrences.
The cited paper in Wikipedia might be accurate and the author is just expecting "at rest" to mean awake but mostly idle. I'd wager that during sleep, we use less than 100W. For the sake of argument if you take out 8 hours (asleep time) from a 24 hour period, you end up with a rate of ~0.8681 grams per second, much closer to 1g/s. Anyway, just a little fun with numbers.