That is fucking bleak.

He could have saved the world 50 years ago by just not opposing nuclear. And he also could have admitted that at any point during the past 50 years.

But instead, what do we see in the comments every time we talk about nuclear? “It would be too slow to save us in time.” Do you think we should’ve believed that argument in 1976?

Archive link

Please: Permanent standard time so that we have a consistent baseline that aligns with better health outcomes and with IANA conventions, plus regulation to guarantee a legal minimum number of daylight hours before/after work.

50 degrees where you are right now, or where your workplace is?

Depends what you mean by social media.

Digital socialization is not necessarily bad.

Easy self-publishing at scale is not necessarily bad.

A marketplace for human attention, where products are optimized for “stickiness” and “eyeball-hours” is bad, but doesn’t necessarily need to be social media. (Buzzfeed comes to mind, and Ricki Lake before that.)

The social media that we ended up with is essentially a chokehold on the means of digital socialization, forcing users to perform for an audience instead of connecting to individuals, and capturing the economic value of that performance as an asset for the platform owners.

We could’ve ended up with something quite different.

Here’s a hot take though:

Products that are built under these conditions will inevitably tend towards a shape that reflects the logic of the market, and they continue to reflect that logic even when financial incentives are removed.

For instance, Lemmy is very Reddit-shaped, and therefore reflects some of the market logic behind Reddit, even after you stick it on the fediverse.

Edit: Barry’s Economics has an excellent video on exactly what sucks about the social dynamics baked into these platforms: https://youtube.com/watch?v=Dqh17U3tymU

HN comments are predictably calling them terrorists, saying they’re funded by shady NGOs, and acting surprised that AI would be targeted for the shitty state of the world. Stay fashy, HN.

I’m not sure we really “skipped over” TUIs. It seems like they only got practical in the past 10-15 years.

But the broader point stands: The conditions do seem right for The Age Of The TUI. Go for it, homeslice.

“The time you enjoy wasting is not wasted time”

To be clear, I’m explicitly focusing on the moral subject here, not the object of their actions. When I say “abusing an LLM could injure the abuser’s humanity”, I’m making absolutely no claim about the LLM.

As I said in the previous thread about this:

LLMs aren’t conscious. They can’t feel pain.

But you should treat them kindly. Because even in a consequentialist or deontological frame, your morals rely on developing good habits; but in virtue ethics, moral habits is really all there is.

Edit: Very Bad Wizards actually covered this pretty well in this episode from 2018: https://verybadwizards.com/episode/episode-150-paul-bloom-insisted-that-we-talk-about-sex-robots

We already know that habitually using an LLM will weaken your critical thinking.

It would not be surprising to learn that habitually abusing an LLM would weaken your empathy.

For me it went from “didn’t know it existed” to “not interested in playing it”.

https://archive.ph/eOu09

Those are some good picks. I’m gonna copy them.

Like an airplane speeding down the runway, running out of time to generate enough lift and already way past the option of safely aborting

So um… They makin one of these? https://pantheon-amc.fandom.com/wiki/Svalbard_data_center

Wow, Sony’s innovations include rootkits on music CDs you purchased, taking away your ownership of your games, and turning your entertainment devices into PoS kiosks for Sony Group Corporation.

They’re always looking for new and exciting ways to say “fuck you” to their users.

It really illustrates how far gone this regulatory regime is — that it’s even possible for a company to stay profitable while displaying such unending contempt for their customers.

I mean… kinda?

You can control what actions it can perform (cuz really “it” doesn’t perform any actions, you’re just parsing its output) and you can control what is allowed to add to the context.

But like… at the end of the day, it’s a stochastic engine that can’t differentiate between instructions and data. You can’t protect it from prompt injection, you can’t prevent it leaking context (or training/tuning data), you can’t stop it from telling the harness (or the operator) to do dumb stuff.

Edit: It’s kinda like saying “plinko is just a board and pegs — it can be controlled”. Like yeah, you can control how you build the plinko board and how you interpret the outcomes and how you let people play it… but none of that means you can control where the chip falls.

There are genuine challenges here. It’s just… none of them are “looming superintelligence threat”, and none of them let the vendors and/or operators off the hook.

Oh they edited

As you wish