postultimate

Journalists have been on the grievance-grifter gravy-train for a long long time, so we can expect unrelenting hostility to anything that redirects attention away from the lucrative "bias" narrative to any other issue.

> AI ethicists and researchers such as Timnit Gebru and Meredith Whittaker

Ooooh, I love their work !

Reading these histories together, we find that Babbage’s proto-Taylorist ideas on how to discipline workers are inextricably connected to the calculating engines he spent his life attempting to build. From inception, the engines — “the principles on which all modern computing machines are based” — were envisioned as tools for automating and disciplining labor. Their architectures directly encoded economist Adam Smith’s theories of labor division and borrowed core functionality from technologies of labor control already in use. The engines were themselves tools for labor control, automating and disciplining not manual but mental labor. Babbage didn’t invent the theories that shaped his engines, nor did Smith. They were prefigured on the plantation, developed first as technologies to control enslaved people.

Doesn't "everything Plato said was stupid" count as philosophical progress since Plato ?

This is a decent exploration of the third-worst possible outcome of AI, but it's a bizarre dismissal of the second-worst, even though he explains the mechanism himself:

> The doomsday scenario is not a manufacturing A.I. transforming the entire planet into paper clips, as one famous thought experiment has imagined. It’s A.I.-supercharged corporations destroying the environment and the working class in their pursuit of shareholder value.

Well, yes. If the AI is better than human it will be put in charge of corporations, and will be given the goal of maximizing shareholder value. Since those shareholders can be corporations, there's no reason why this has to involve human preference at any point. As a single-minded optimizer indifferent to humans, it would need to be successfully restrained in order to be merely oppressive - by default, it's a Paperclipper.

> when a straightforwardly “I’m a Nazi” Nazi showed up in the beta, people used the report function, and the Bluesky team labeled the account and banned it from the Bluesky app

"Don't like the feudalism of Mastodon ? Come enjoy the monarchism of Bluesky !"

Among all the fuss about typography, the authors have overlooked the most important detail about the PHD thesis, which was that Dennis Ritchie was the inventor of Brainfuck.

No, this is just question-begging by treating GPT's access to the world as being external to it, but your own as being part of you.

If we fix this by treating your senses as external, then we can imagine a copy of you with its senses rewired so that artichokes* taste like icecream (and vice-versa). (plus we lie to you about which is which.) The resulting imtringued2 is identical to you, but doesn't like ice cream despite it saying it likes ice cream. Just like IceCreamGPT2.

* Or some equally disgusting "food".

When GPT mentions ice-cream, it does so because it was in the corpus. When it occurred in the corpus, it was as a reference to actual ice-cream. So GPT has just as much intentionality as you do.

You might claim that you've eaten ice-cream, and that that makes a difference. But if we assume that your senses aren't lying about what your senses do, then what they do is produce information - indications of difference without any indication of what it's a difference of. That puts you in the same epistemic position GPT is in. GPT knows just as much about ice-cream as you do.

1 GPa is less than half the yield strength of the best steels. All you would need to maintain the superconducting state is to put the material in a steel pipe.

Good for him. Nobody's racism should be unscientific.

Welp, HN has nerdsniped me again: After some unpaid pondering, here's my suggestion for a middle ground on downvoting:

1) Reduce issue-related downvoting by only allowing some random subset of the current active users to downvote, a different set per post. Don't let the user know their vote won't work in advance, but tell them after.

2) Reduce vote sockpuppeting by recording which pairs of voters downvoted a post, then disallow that pair from downvoting a post again. This includes votes that failed due to 1).

Yes, tell everyone how much you hate white people too. Your act of piety is surely guaranteed to keep the inquisition from your door.

'Originators assumed that “liberals” would react negatively to such fliers and condemn them or take them down, thus “proving” that liberals did not even think it was “okay" to be white.'

...and indeed, it proved exactly that.

No, that's not what |I'm supposing. The 80386 got used in vending machines because technology moves on, the new stuff gets cheap, and because it's more generally useful, it gains in popularity while the old stuff loses it, even in cases where the old stuff is adequate. The same process will probably happen for AGI - unless non-AGI has capabilities that AGI can't replicate, AGI will probably replace it.

1974: "No-one would put an AGI in a bomb, that's just ridiculous"

1986: "No-one is going to put an 80386 in a vending machine, that's just ridiculous"

"The Yellowstone magma chamber is sitting there like an enormous angry zit. What's our plan ?"

"We're going to stick a very big pin in it"

But the lack of a positive refcount means that it is, so this solves the problem the article was complaining about.

> task groups (nurseries from trio) can't arrive soon enough in the stdlib

Please, no. Asyncio is horrible, and bodging it to make it less horrible just means we will be forced to live with the remaining horror. Far better to replace it with something that works properly (yes, Trio).

This one seems quite easy to fix - just have the scheduler check that task objects have a positive refcount before running them, including the first time.

Did they change their minds about the waste in the pools at Fukushima ?

If those had collapsed (and at one point it looked as if they might), it would have fatally irradiated over 10,000 people.

Copyright offences on an industrial scale.

That one and much more here:

https://michaelbach.de/ot/index.html

A larger corpus.

Just on the offchance that anyone cares, the PLUCHINO/BIONDO/RAPISARDA paper quoted in the article is a complete fraud.

They use a model where, in each step, wealth is halved with p=0.5, and doubled with p=0.5*talent, where no individuals have a talent above 0.9. In other words, in the absence of randomness, everyone has zero wealth.

Does randomness cause wealth in a model designed so that lack of randomness causes zero wealth ? The answer may surprise you.....

It's a radioluminescent paint company, so about six and a half thousand years.

These findings — from [..], the Anti-Defamation League

An organization that was created to help a Jewish rapist/murderer escape justice said so ? Well that's an unimpeachable source !

Yes,

Google CEO Jeff Weiner's Quest to Turn the AI Resource Engine Into an Effective Leader

is a very worrying development.

Hurricane Florence will likely “peel off” the US Atlantic Coast, US Coast Guard spokesman says

Getting the right amount of memories right – the culture of memory buffing in bats

The art of lithium-ion dynamite (1995)

We built the game that makes it harder to punch people

I posted this on Hacker News, and now it's being cut off from my website (2021)

Show HN: I built an acrylic web interface that resembles my laptop

Ask HN: How come not other languages?

80-bit integers get twice as big

EBay fired me for praying during my contract negotiations

Oregon Community College Students with Common Hyperparasites (2018)

New body cameras that didn’t see a shooting

Show HN: Awk – The ultimate search engine

The Complete Code of the IWK-01 Space Gun

Write It Down: Probably

NASA's Hubble observatory detects earliest known appearance of crude grease on objects around the world

Bitcoin collapses below one BTC

Show HN: I made a web app for making to-do lists

Hollywood is turning to Russia as a model of “positive cultural integration”

--

Ok it's just interviewing for The Onion at this point.

Microsoft considered killing the startup community, says CEO

Tesla’s strange new CEO: ‘This is not the right place for me to be successful’

Puerto Rico is receiving $6

Earth’s largest cloud is the remains of a comet (2019)

A deep dive into C++ features that make your life harder

Windows 10 is basically iOS without the app drawer

Never trust a trusted set of data (2013)

The Rust Tax

Terabot: A remote-controlled toy train that trains riders to do your work

4-Hour Workweek Passes UNH’s Students

Why does Brexit happen?

NASA Launches High-Definition to Broadcast Its Infomercials

I made a living building software that no one needs

Scientists have created a new kind of blood clot

Two weeks without Instagram

Life on the ocean floor and how humans inhabit it

Taxing robots is immoral, right? Because it’s a) easier, and b) it would vastly reduce human labor

Why are people so glad to have their memories wiped?

Does the US lead the world in online harassment?

A 50-year-old man shops for lamb chops on the Internet for the first time

They said it was impossible to dig for cobblestones without being coated with mud

--

Ok, that's it. Time to abandon HN and go over to this site.

Actually, scratch that - I'm just moving to the universe this bot is from.

If the company bans you, they are publicly claiming you have performed a disreputable action of some kind. If you have not, why not sue them for defamation ?