I wonder what other applications this might have outside of machine learning. I don’t know if, for example, intensive 3d games absolutely need 16bit floats (or larger), or if it would make sense to try using this “additive implementation” for their floating point multiplicative as well. Modern desktop gaming PCs can easily slurp up to 800W.

“I need some more [input] sanitizer for my eczema script, the console is red and inflamed whenever I check it.”

the police say they are targeting the criminals responsible but cannot “arrest their way out of the problem”. They also say manufacturers and tech firms have a bigger role to play.

Even though I fully expect the police here aren’t doing as much as they could (I mean come on, are they expecting phones to come with wiimote hand straps?) , I’m at least glad their public rhetoric is that they can’t “arrest their way out of the problem”.

I imagine that’s poor compensation when you’ve just had your phone snatched, however.

The audacity to tout classism and ableism as reasons as to why people should “get to” use LLMs for their “write a novel in a month” challenge…

Even when someone’s inability to write a novel in a month is because of their class or disability, I somehow doubt they want to let a machine write their novel for them. I mean, it’s not like NaNoWriMo is a way to put food on the table or something, right?!!

This feels like the arguments Mid journey fellators fanboys were spouting a year ago (or has it been two?) on how not everyone can afford a school of fine arts 🙄

Their starting aside is pretty great as well;

And I’m using that term throughout this post because it’s the commonly accepted descriptor, but we all know it’s not really artificial intelligence, right? I also want to distinguish it from actually-useful and ethically-produced technology like what gets used in the medical field to help humans examine and analyze impossibly huge datasets in the service of doing things like curing cancer. We’re talking here about the plagiarism machines like ChatGPT, everything it underpins, and all of its conceptual mirrors.

Leave no wiggle room for the AI sycophants.

Having just watched the lecture, the only classified info I can recognize is the capabilities of 80s era satellites.

Given that, I think it’s quite a shame that the whole thing is only now available. Rear Admiral Hopper seems to have been someone who deeply understood both computers and people. The prescriptions she gives regarding “systems of computers” and “management” vs “leadership”, to name just two, are spot-on. Her lecture is quite grounded in what I’d call “military thinking”, but that’s just because she’s in a room filled with people who are of that life. In my opinion, everything she talks about is applicable to communities and businesses.

The general gist of the entire ~90mins reminds me of Project Cybersyn in its perspective on how computers could serve society.

The idea is neat, and there is a certain precedent for the approach in .htaccess files and webserver path permissions.

Still, I worry about the added burden to keeping track of filenames when they get used as stringed keys in such a manner. More plainly: if I rename a file, I now have to go change every access declaration that mentions it. Sure, a quick grep will probably do the trick. But I don’t see a way to have tooling automate any part of it, either.

Bitcoin enables the fantasy of becoming a rent-seeking plutocrat that cannot have their wealth seized by a government nor a popular uprising.

An oldie but a goodie : https://csszengarden.com/

“With respect to content that is already on the open web, the social contract of that content since the 90s has been that it is fair use. Anyone can copy it, recreate with it, reproduce with it. That has been freeware, if you like. That’s been the understanding,” said Suleyman.

Ugh. Social contract, free use, freeware - those all mean very different things. I don’t think the head of a department like that should be blabbing to the public if they’re going to mix up terms like that. Do they not have PR and legal departments that are versed in anything beyond Microsoft’s historical business methods (lie, steal, and fearmonger about open source)?

Not to mention that in some places, you cannot give up the IP rights over code you write.

Not to mention “fair use” is primarily for artistic endeavors.

Not to mention “freeware” is for programs, not written word blog posts or images.

etc…

Lol. What, are we going to be installing Candy Crush on our robots? Expecting to be able to project recurring revenue from a humanoid robot based on smartphone numbers is a new kind of ignorance.

I just got back from a trip to a substantially less developed country, and really living in a country, even for a little bit, where I could see how many lives that money could improve, all being poured down the Microsoft Fabric drain, it just grinds my gears like you wouldn’t believe. I swear to God, I am going to study, write, network, and otherwise apply force to the problem until those resources are going to a place where they’ll accomplish something for society instead of some grinning clown’s wallet.

Amen. We always need more insiders who are ready to take up the cause of not doing stupid shit with the ungodly accumulation of resources our society has permitted, especially when we are currently leaving so much of the world to play catch-up while we continue to leech them dry.

Not necessarily cash, but definitely a bit of luck. Some lawyers, if they think a case is guaranteed to go your way, will do the work for free in exchange for receiving a portion of the damages the final judgement will award you. Even rarer, some lawyers care enough about some issues on a personal level that they’ll work for free, or reduced rates, on certain cases.

In this case, I’m not sure there are any damages whatsoever to award to OP - a “win” is forcing the company to abide by the GPL, not pay up money. The EFF and the FSF, as others have brought up, are probably the best bet to find lawyers that would work on this case for the outcome instead of the pay.

@> By the end of the decade, American electricity production will have grown tens of percent

… don’t worry guys AI will totally save us from climate change! /s

You’re right, I should have been more specific.

If you’re already storing your password using pass, you aren’t getting 3 factors with pass-otp unless you store the otp generation into a separate store.

For services like GitHub that mandate using an otp, it’s convenient without being an effective loss of 2fa to store everything together.

I already use pass (“the unix password manager”) and there’s a pretty decent extension that lets it handle 2fa: https://github.com/tadfisher/pass-otp

Worth noting that this somewhat defeats the purpose of 2fa if you put your GitHub password in the same store as the one used for otp. Nevertheless, this let’s me sign on to 2fa services from the command line without purchasing a USB dongle or needing a smartphone on-hand.

Maybe it’s just because it’s all actually real, but there’s a mind-melting yet quiet horror to this article that I feel Lovecraft could only dream of imagining.

Like some sort of Ionesco’s Rhinoceros mixed with a reverse Kafka’s Metamorphosis, grotesquely stretched over Primo Levi’s If This Is A Man.

That’s probably too convoluted of a metaphor to mean much to anyone else, but, fuck, I don’t have the energy or the will to unpack it further than that.

To think, when I first read the post title, that my initial reaction was “here we go, techtakes has finally gotten too edgy for me”.

I need a cigarette and a scalding hot shower.

You may also interact with countless bots without ever knowing, because creating fake identities is free.

Maybe. Bots don’t seem currently capable of holding a conversation beyond surface level remarks. I think I tend to engage with thought-provoking stuff.

On the off chance that I reply to a bot, it is as much for my reply to be read by other humans viewing the conversation. So I don’t understand how interacting with countless bots is supposed to be such a big downside.

Plus, I don’t see how public/private key pairs prevents endless “fake” identity creation/proliferation. It’s not like you need a government-issued ID to generate them (which, to be clear, still wouldn’t be great -just got other reasons).

Fair, some people value their identity.

To be clear, I’m talking about online identities. In which case, I would argue that if you value it so much you should not delegate it to some third party network. My IRL identity is incredibly valuable to me, which is why I don’t tie it up with any online communications services, especially ones I have no control over.

For average people nothing changes, the app can hold their key for them and even offer email recovery.

…so then the app can post on my behalf without me knowing? And it’ll be signed as if I had done it myself. I don’t understand preferring this if you’re not also self hosting.

That’s something having signatures and a web of trust solves.

But as I wrote in my previous message regarding gpg signing circles (a web of trust), that doesn’t “solve” things. It just introduces more layers and steps to try and compensate for an inherently impossible ideal. Unless I’m misunderstanding your point here?

Besides, you fail to see another problem: Whichever centralized, federated site you use can manipulate anything you read and publish.

I just take that for granted on the internet. It’s true that key-signing messages should make that effectively impossible for all but the largest third parties (FAANG & nation-states). But you still need to verify keys/identities through some out-of-band mechanism, otherwise aren’t you blindly trusting the decentralized network to be providing you with the “true” keys and post, as made by the human author?

Anyway, if you don’t see a need for tools like nostr you don’t need them.

Maybe I’m not expressing myself properly; I don’t see how nostr (and tools like it) effectively address that/those needs.

Sort of like how there was (arguably still is) a need for cash that governments can’t just annul or reverse transactions of, yet bitcoin and all cryptocurrencies I’m aware of fail on that front by effectively allowing state actors (who have state resources) to participate in the mining network and execute 51% attacks.

It weirds me out that most of the arguments for nostr I come across are around how “you can’t loose your identity, it’s just a private/public keypair!”. Maybe I just don’t get banned enough to understand the perspective, but to me the real problem is the content/discussions being lost, not usernames for some corner of the web.

I really don’t care about loosing my identity on a social media website; I’ve found it healthier to view social media accounts on the same level as my customer account at my isp and power utility. When I change ISPs, the old account is closed down and I start up a new one at the other ISP. What’s important to me is the service getting delivered, not that it remembers that I’m the same person from however many years ago. It’s still the same me here in my body, interacting with the web. I know what I need from it, it doesn’t always need to remember who I am (and sometimes I’d rather it forgot or never knew in the first place).

My final point is a bit of a troll, but also kinda serious: how decentralized is it when your identity is “centralized” in your key pair? Loose your keys or loose your password to the key, and your identity is similarly effectively gone. Even worse in this case, no-one can restore it for you. Which is why I don’t tie my identity that much to any online service, especially ones I don’t host. The only thing that truly preserves my identity is the flesh-and-blood body that I inhabit (and even that isn’t fail-proof).

I’ve interacted with GPG signing circles before. So many people are losing access to their keys. So many more are considering some of their keys as compromised. In either case they’re regularly generating wholly new keys, essentially rebooting their “identity” from scratch. When they do so, they always rely on flesh-and-blood interactions to have their new identity verified and trusted by others.

Maybe it’s a question of which circles we’re involved in; mine are already regularly hopping accounts, without being forced to by bans or server outages. I’m used to interpreting the tone & content to recognize “people”, and ignoring usernames. On top of that so many people regularly change their display names on social media for vanity and expression purposes that I can’t reliably use them anyways for recognizing accounts.

Right?

“Protecting vulnerable individuals” - they must mean Putin and Bibbi, not actual victims of political intimidation, sabotage, etc.