Post

One another Bluesky lesson is that people want to put their head in the sand wrt certain kinds of lines of reasoning, and it tends to work out well when you give them the ability to do so. Sorry you got muted. If you want emotional satisfaction, put them on "puts me on stupid lists" list, works 4 me

Hell that's entirely possible, you might've had previous interactions aplenty for all I know, but it looks entirely uncharitable to your own arguing chops in the context of this thread, especially given it shows up at the top in threading view. (also, 🙏 listifications can be mind poison, seriously)

Turns out decentralization wasn’t enough and we might need to fix the UX if we’re going to fix social. The 🙉🙈 experience is definitely proving to be a failed experiment.

@merien.com AI-apologists zijn mensen die AI-gebruik verdedigen zonder de risico's te erkennen. Ze minimaliseren schade en verwaarlozen schadereductie.

These same people then constantly wonder why they lose so much.

I think one of the reasons this argument doesn't always persuade tech folks (putting aside the shills and those with vested interests) is because coding is one of the areas where it *does* very often increase productivity.

the vast majority of software that you use every day doesn't make a profit. most of it is "built" by non-profits, but more generally random nerds. see: xkcd.com/2347/ I'm running most my inference locally fwiw. on my own hardware.

You do realize that, the popping of the dot com bubble was just an educational moment. E-commerce is still quite popular. As will be AI.

Repeating and teaching them the truth that they are being manipulated and indoctrinated with their education undermined for an existentially destructive climate threatening scam that cannot be stopped from regressing further into slop. Those "millions of people" will get over losing it at that cost

Blocking ppl who are not intolerant, bc they voice details, context & nuance wch is offered to reduce harm, is the real "Blueskyism." It's Nietzsche's idea abt fighting monsters/staring into the abyss. Bluesky is the abyss & some start seeing monsters in everyone, thus becoming monsters themselves.

Have you ever spoken to anyone in the field of, say, addiction treatment?

At what point are folks going to try literally any other tactic than condescending rants? Does it matter that LLM apps are at the top of virtually every app store nearly every day because individual people are choosing to download them, and the criticism hasn't been effective in slowing that?

💯

Defining some development standards for A.I projects in your company is helpful. Things like archiving prompts for genai builds, preserving non a.i APIs , requiring a.i specific pentests , having a flow chart for when and when not to a.i. etc..

this a shit take, bro nobody goes rightwing who wasn't already aimed in that direction blaming that on any pushback which is less than *perfectly polite* is tone policing and respectability politics, writ large I'd expect a *journalist* to know the fucking difference

I'm genuinely not following what you're saying here. Where did I say something like that?

Nobody is saying that people shouldn’t be scolded for their choices. What is being said is that scolding people for their choices is a piss-poor way to get them to change those choices.

Yeah again, by screenshotting this person you are in fact scolding them with moralizing about their choices

This def sounds like something an “AI apologist” would say. Did all the school boards buy YouTubes without understanding it, force teachers to use it without knowing whether it’s at all efficacious in the classroom?

If a whole industry has become committed to its proliferation, there's really little people without the cash needed to make alternatives can do apart from saying it's a piece of shit. You can tell folks about sensible uses, but it's sure not what the service providers are constantly pushing.

Welcome to the party!

You and Masnick are both are pushing technology you KNOW to be dangerous to people. You can pretend you want some non-dangerous version, but there's no such thing. You guys just want to be on the hype train with your buddies, and you don't care who gets hurt. Classic silicon valley shitheads.

I found this from Lee Vinsel to be very relevant criticism of the tech criticism community. peoples-things.ghost.io/why-do-peopl...

Sorry but we are allowed to rant about our family consuming fast fashion, eating too much meat, and using GenAI, without necessarily having to engage with them every single time. And also: it's not their fault, but the system's. You might be misguided about what's "effective" and what isn't.

Thats crazy

There is value in the overall tools for doing the drudge work of, for example, creating a survey form, which is my primary use for it. But the resources going into saving me ten minutes may not be worth the squeeze here.

I actually agree you're not gonna scold most people out of using AI, but I do think in many ways the extreme hype is almost as silly as NFTs were, in that neither is organic

I basically refuse to talk about the subject online (I'm breaking my rule now) because of how toxic the conversations get. Which I think is sad, because I think there are interesting debates to be had, but just not here.

You did not! But the "AI Apologists" moderation list Anil is referring to is indeed the kind of thing often used for that, and that is the AI discourse we seem to have on Bluesky: entirely lacking in nuance.

I ended up on an AI block list for saying something along the lines of AI isnt going anywhere, we should be looking for useful applications in things like robotics instead of just trying to shut it down. I'd rather steer the ship the right direction than sink it.

There was a larger conversation going on prior to that; I am assuming you likely encountered that post in isolation because somebody wanted to dunk on it. It’s far from the entirety of the point I was making, or even a particularly significant part of it.

I mean, Look at who the CEOs are cozying up to. It doesn’t matter how much foundational work or research is done if the ones directly in control of weightings want a techno-fascist regime. Look at what Meta’s AI is doing with minors. The space for agentic AI and LLMs is tainted the same as crypto

😬

No the leaders of silicon valley being that is actually the problem.

I feel there is some selection bias occurring. I try to showcase both AI wins and where it has gone wrong in medicine. There’s a bigger appetite for anti vs pro AI news. That’s why I’m trying to build more spaces for community to grow and have nuanced conversations, and avoid animosity and vitriol.

Why do you require people to discuss their research on social media? Some of us keep that stuff to ourselves - and I feel like we should be allowed to.

I don't know.

Lot of folks also don't fully appreciate how much progress has been made on the NOT data-center-hungry version of AI. You can run a surprising amount of stuff on your own machine for basically the cost of playing on an Xbox.

The actual use cases for LLMs are so small, though. And the companies trying to shoehorn them into everything are just not helping. They are the people you need to talk to, really.

But you aren't trying to get rid of it. Instead you're promising a non-dangerous version and that doesn't exist. You are putting a sugar coating on a poison pill. No one's saying you have to go yell at AI bros, but you're actively supporting it, and that's different.

Ok but, regardless of which tactic is actually superior, in what way is the maximalist critique stopping anyone from building these so-called user-empowering AI tools? Also this is conjecture but I think ghost absolutely benefits from people making up rude names for substack users.

You're a guy who's hyping a technology we know is dangerous. A guy who makes his money covering new tech, who stands to benefit from the technology. Which one of us is less biased, do you think?

Agreed. It’s really demoralizing for the people fighting for responsible AI within these companies. And if they get bullied and shamed into leaving the industry, things can and will get MUCH worse.

I hope you do write that. Those people are tiresome and reality-divorced.

The idea that “better tools” can solve massive social problems like LLMs seems like a category error. The problem with LLMs isn’t that people aren’t empowered enough - so what on earth would empowering people solve?

I consider all of those things to be AI. You have correctly pointed out that the marketing strategy used by silicon valley to push this new crop of LLMs by making them synonymous with AI is part of the problem here. Of course AI itself was first coined as a marketing term itself.

lol blinkered

Some guy elsewhere in this thread was talking about firebombing data centers and preventing people from using LLMs “by force.” AI always seems to cause people a few hinges short of a working door to lose the ones they have instantly.

It certainly has had a chilling effect. Some of the most fascinating AI posts come from @emollick.bsky.social, but he has mentioned a few times he’s been wary of posting here due to some of the responses he’s received.

Yeah that reaction is driven by the ways that this current batch of LLMs have been pushed across all sectors of society. If you want to blame someone for that blame Sam Altman et al because they are the ones who created this backlash.

Yes. That's bad. So let's do something better.

Yeah I mean, even just the alt text labeling tool here on Bluesky, or the abusive content detection, are both very useful positive things.

I mean, there are two use cases for AGI, one is to destroy jobs and transfer wealth from labor to capital, and the other one is to enhance labor productivity. Guess which one the "big four" are focused like lasers on right now? I get it.

Word and excel destroyed millions and millions of jobs, we should have banned them when we had the chance

Could you give an example of one?

Well, no, not if someone will come and write my iPhone apps for me, a retired sysadmin who can tell the AI what to do, more or less like a jr dev in another country, and it does it. I assemble the pieces and have a program I could never ever have created on my own. I feel pretty fukcing empowered!

Hopefully that is captured by your article, but if not, maybe rethink your premise!

Nobody with any sense thinks it’s going to exist like it does now long term.

I know what to do with it.

While I agree with your first sentence, your second sentence is very annoying to me as a long-time AI hater that you have made me look bad by association.

I'm sorry you have come to my thread to support the original authors point about non-empirical anti-ai no thoughtism. Oh look I went on google scholar and found an article about proposing classes of drugs to manufacture and research to find particular beneficial effects in patients.

I'm begging you just give me 6 more months on the juice 😜 all AI is gonna do is make you scream less at your stupid computer and all these dipshit companies that think they can replace their customer service department with AI will eat an entire mountain of shit I can't replace you. period.

The vehement, absolutist arguments against AI probably aren't going away, any more than AI is! So the 2nd-order appeal you're making may not persuade. I'm fwiw as LLM-hostile as most academics, but there is much I don't know, and I learn useful stuff from you and Mike M. Thanks.

Local governments don't really like it when you go in and pay people US-standard wages because they don't want all their doctors, government workers, etc to quit and become data annotators

out of all the people in the world being exploited by Western hypercapitalism, "data labeling" is probably something I'd tackle last...

whataboutism has nothing to do with the problem at hand

Don't scold people using AI. It's the politicians who are letting this happen. Scold them and vote out anyone who is not putting up severe guardrails that cannot be easily rolled back. And most importantly, democratize the hell out of it. Let's not repeat the same mistakes of social media.

(The shrug being quite literally a shrug emoji!) & this doesn't even go into the nervous feeling I get about how we're a nonprofit & I can't even imagine what's going to happen when these subscription costs start to reflect what it actually costs to provide it with investor returns.

They're using it because it's free and losing tons of money. I don't get how it's going to exist if you can't make money on the stuff.

No? Are you really having that much trouble following the conversation?

I mean, no? He's not at all?

AI is dying and it's reputation is poison. So I'd say that it is working. What "harm reduction" are you suggesting