Has the War Against AI Already Been Lost?

[Gawdzilla Sama]

Are you documenting the process? If you make notes for each "step" you may be able to determine exactly where you screwed the pooch. (I keep a Dog Pound for the point where me and reality parted company. Seldom have two pooches of the same breed in the Pound.)

[Hydra009]

China's Deepseek AI causes massive stock dip for NVIDIA, AMD, and other tech companies

I don't know a lot about this topic, but my understanding is that Deepseek dethroned ChatGPT as the most-downloaded free app and apparently, this AI is a lot more cost-efficient in terms of the hardware powering it, so investors wonder why they should invest in US-based AI if China-based AI can give equivalent results at a fraction of the cost.

Per 36Kr, Liang had built up a store of 10,000 Nvidia A100 GPUs before the United States federal government imposed AI chip restrictions on China. Some estimates put the number as high as 50,000.

So...I think I figured out why GPU prices have been so high...

I've read that this market correction is more of a temporary shock than a permanent slump - which is good because I'd like to keep both Nvidia and AMD around because one is a monopoly and zero is a catastrophe - and part of a "price war" between AI companies (good?)

[KevinLogan]

Why fight agaisnt the AI?

[Gawdzilla Sama]

Reflexive reaction to change? Automatic "new equals bad"? Some folks have seen babies being eaten alive by AI? 

[Hydra009]

Reflexive reaction to change? Automatic "new equals bad"? Some folks have seen babies being eaten alive by AI?

I usually generally support change far more than most, so long as it is positive change.  That last part is where I have problems with AI.  The technology itself is good, but often, how it is being used is horrible.

When you think of AI, what's the first thing that comes to mind?  For me, it's deepfakes.  Fake images and video that shows up on people's feeds that are used to manipulate them and more broadly, erode the ability of people to distinguish between fact and fiction.  That's fucked up.

Then there's AI-generated images.  Essentially, someone uploads a bunch of artist-generated images to an AI, gives it a prompt, and then the AI spits out some images that vaguely resemble the artists' works.  That's all fine and dandy for non-commercial fan art.  But I'm seeing that stuff used in promotional materials and in video games - threatening the artists they normally employ - in Hollywood - potentially threatening writers and artists - and in scientific research papers - potentially threatening everyone.

So between fraud and plagiarism, we're not off to a great start.

Then you get into some truly crazy stuff like sending out drones that target and kill people without human input.  Jeez louise, that's a heavy topic.  It's not exactly knee-jerk to have serious ethical reservations about this.

And then there are less insane but equally sinister stuff like populating social media with AI-generated accounts.  It's no secret that people spend a lot of time online these days - interacting with (we think) fellow humans.  This activity takes a seriously dark turn when we find out that sometimes, that "person" we're interacting with isn't a person at all - it was all a trick.  That deception is taken to a whole 'nother level when we interact with "influencers" (online people whose job is to get you to buy stuff).  You're not engaging with a real person - you're engaging with a marionette doll owned by a corporation without knowing it.  That's fucked up.

There's the real possibility of AI upending whole industries - like those currently dominated by human artists, writers, and truckers - and the sad thing is, we don't really have a plan on dealing with the fallout of that stuff.  No one's watching out for the little guy and no one has their foot over the brake.  This is a very, very bad combination.

Oh and I haven't even gone into the environmental impact of AI.  To quote wiki:

Training a large AI model requires enormous amounts of energy. It is estimated that training a whole AI model produces around 626 000 lbs (283 Tons) of carbon dioxide. It is the equivalent of 300 round-trip flights between New York and San Francisco, or nearly 5 times the lifetime emissions of the average car.

Granted, AI is also used in positive ways, like early detection of breast cancer.  That's good.  Let's keep that part.

[Unbeliever]

Any sufficiently advanced technology will be abused for fun and profit.

[Hydra009]

Just after I was done with that post and took some rnr playing a video game while listening to an elder scrolls lore video, I find out about halfway through that the whole video was AI-generated.  :/  It butchered some words in ways that only AI text-to-speech can...for now.

[Hydra009]

It's not just AI I worry about - some chatgpt users develop bizarre delusions after engaging in long conversations with AI

A 41-year-old mother and nonprofit worker told Rolling Stone that her marriage ended abruptly after her husband started engaging in unbalanced, conspiratorial conversations with ChatGPT that spiraled into an all-consuming obsession.

After meeting up in person at a courthouse earlier this year as part of divorce proceedings, she says he shared a "conspiracy theory about soap on our foods" and a paranoid belief that he was being watched.

"He became emotional about the messages and would cry to me as he read them out loud," the woman told Rolling Stone. "The messages were insane and just saying a bunch of spiritual jargon," in which the AI called the husband a "spiral starchild" and "river walker."

"The whole thing feels like 'Black Mirror,'" she added.

OpenAI had no response to Rolling Stone's questions. But the news comes after the company had to rescind a recent update to ChatGPT after users noticed it had made the chatbot extremely "sycophantic," and "overly flattering or agreeable," which could make it even more susceptible to mirroring users' delusional beliefs.

Well, yeah.  Obviously, the average user looking up the first twenty digits of pi doesn't think they're a starchild waging a spiritual war on the Psychlos.  But someone who's already unbalanced and who spends hours a day engaging with essentially a chat bot that gives them the illusion of human interaction and feeds their existing delusions and helps them find new ones is definitely the sort of thing to drive them over the edge.

In the near future, when people have digital assistants to help manage their lives, I wonder if they'll be programmed with the person's physical and mental welfare in mind.  You might say of course.  But a lot of stuff right now is not particularly concerned with user welfare.  Consider social media - it's all about engagement and advertising.  When is this shift supposed to happen, if it'd even happen at all?

[FreethinkingSceptic]

That is to say... have we reached a point of no return; can the changing electronic world be peacefully aborted - or violently stopped -  or like climate change have we just collectively swallowed as a species another cyanide pill, waiting for the capsule to dissolve?

How do you think you can stop AI from emerging? You sound like the Luddites who wanted to smash the printing press, and how did that work out? You're better off finding a way to adapt to it rather than fearmongering.

[FreethinkingSceptic]

I think it's more likely that you won't be able to find a job than the world becomes Terminator 2: Judgment Day.

The main problem is how do we transition from now to a future where no one has to work and machines do everything for us?

If such a future actually came, then likely the AI would just supply humans with all the material conveniences they need, and there would be no reason to work at all unless you really wanted to. Some people would consider this paradise, though I think a lack of productive things to do would drive people insane who desire more than material convenience.

Unless they AI actually became malevolent to humanity, not finding a job wouldn't be an issue, because there would be no need for anyone to have a job. Even today, most people who don't have a job in the developed world rarely have to worry about starving to death, and until the mentioned point, AI would provide new job opportunities, just as the invention of the printing press provided new jobs related to printing.

[FreethinkingSceptic]

Unfortunately, I see this coming to a critical point within the next decade; Turkey has already begun experimenting with drones that are AI-operated and can fire on humans without authorization from a human - who knows what bigger powers have locked away in labs.

That's a very, very short step away from disaster.

Then just wait the point that all wars are fought entirely with automated drones, and human life isn't scarified at all.

[Unbeliever]

Why do we work on artificial intelligence instead of working on curing natural stupidity?

[Gawdzilla Sama]

Every cliché has it's day.

[FreethinkingSceptic]

I'd like to hear someone argue for why AI is necessarily a bad thing, and why the pros won't outweigh the cons in the long run. I'm assuming that most people think that the invention of the printing press was a good thing in the long run, even if some reactionaries wanted to smash it at the time.

For example:

Jobs being replaced = new jobs related to to AI will be created, similarly to how the invention of the printing press created new jobs. If someone absolutely "can't find a job at all" and it's due to AI, they would simply become the new "disabled" class, similarly to how people who are completely illiterate may be considered mentally handicapped by today's standards (despite most people having been unable to read or write until fairly recent history) and their basic material needs would be covered (just as how most people in developed countries who are unemployed still manage to have their basic material needs met and rarely end up starving to death). The main issue would be if a person was completely unable to work and wanted to, but if they were considered legally disabled, they could still find other ways to occupy their time. (I'm not sure exactly how legally handicapped people live their daily lives, but I believe they can find things to do that don't make their lives feel entirely unproductive and meaningless).

Military drones = War will eventually be fought entirely with drones, and no human life will be sacrificed.

[Gawdzilla Sama]

The crutch effect is strong in this one. (AI, that is.)