Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

Reality

13 (65%)

Universe

4 (20%)

The World

3 (15%)

Total Members Voted: 20

[King Zultan]

Guess that means I need to work on my handwriting because right now it's really shit and hard to read.

[martinuzz]

The New York Times decided today to explicitly forbid the use of it's archives for training AI.
They changed their user agreement so that anyone using their archive for AI training purposes will face fines or other unspecified legal punishment.

I am not sure if I can agree with this.
If all media with at least some journalistic quality standards deny access to AI training, we will end up with AI trained by 4Chan

[dragdeler]

Yeah one would hope something commonly accepted as the "paper of record", would be prime training material. By the time society will be done curtailing this "threat", it will be worse than toothless it will be spam.

[MaxTheFox]

Well there are always opensource models that can be distributed through torrent. Yo ho ho and a bottle of rum!

[EuchreJack]

The New York Times decided today to explicitly forbid the use of it's archives for training AI.
They changed their user agreement so that anyone using their archive for AI training purposes will face fines or other unspecified legal punishment.

I am not sure if I can agree with this.
If all media with at least some journalistic quality standards deny access to AI training, we will end up with AI trained by 4Chan

Ah, then it might have ethics! The HORROR!

[McTraveller]

So this was 7 years old; I wonder if there is anything more modern? Would it be better?

Sunspring

[Folly]

Today I tapped on a clickbait news article recommended by my phone, despite being fully aware the folly of such endeavors. Just as I started skimming through the nonsense, searching for anything that resembled useful information, I noticed a prompt near the bottom of the screen; Google wanted me to let their AI summarize the article. A few seconds later some 10+ pages of excessively padded bullshit had been boiled down to 3 short bullet-points.

This is honestly the best thing since ad-blockers. I mean, it's terrible that we've come to a point where we even need something like this...but we do need this, and now it's here.

[lemon10]

So this was 7 years old; I wonder if there is anything more modern? Would it be better?

Sunspring

Yes, the difference between now and even 3 years ago in AI is massive and categorical. For instance these AI generated, voiced, and drawn south park episodes.
Modern AI can easily hold conversations on a wide variety of topics in such a way that say, five years ago they would have passed the turing test.

[methylatedspirit]

AI is an ill-defined field that encompasses everything from statistics, to the algorithms that make your phone's capacitive touchscreen register the correct inputs most of the time, to the current hype-train AI that operates intelligently but can't work mechanically or know when to give up, to silicon-based people.

If you ask me, it's Humanity that'll save us from AI, just because nobody fucking knows what it means. The field needs to grow up and learn what it actually wants to be, because right now, all I'm seeing is a very loose conglomeration of words that serve the status quo. I despise OpenAI-brained garbage because its problem domain is unbounded yet its input resources are bounded, but I have sympathy for projects like Cursorless (voice recognition for code, but one that works and needs you to speak in a domain-specific language) because they know what they want out of the tech while happening to use the same generative pre-trained transformer ideas that the OpenAI clowns use.

[Loud Whispers]

I hope AI manages to survive the future of copyright litigation because listening to Plankton do Franki ValliSquidward doing Frank Sinatra covers is just right

[EuchreJack]

Today I tapped on a clickbait news article recommended by my phone, despite being fully aware the folly of such endeavors. Just as I started skimming through the nonsense, searching for anything that resembled useful information, I noticed a prompt near the bottom of the screen; Google wanted me to let their AI summarize the article. A few seconds later some 10+ pages of excessively padded bullshit had been boiled down to 3 short bullet-points.

This is honestly the best thing since ad-blockers. I mean, it's terrible that we've come to a point where we even need something like this...but we do need this, and now it's here.

Yeah, but sometimes the AI lies. It's entirely possible that article didn't say what the AI said, but rather the AI figured you would want it to say that, and just gave you what you wanted.

[Robsoie]

https://old.reddit.com/r/ChatGPT/comments/zkbkgu/chatgpt_is_very_touchy_and_doesnt_like_me_asking/

[lemon10]

Sometime in 2023 I noticed an odd phenomena, google search was getting worse. At first I thought nothing of it, maybe it was just me. Then on finding an article about how google was changing results secretly to similar ads without saying it was an ad I figured that it was just usual enshittification.
(Although on going back to look at the article it turned out to not be quite true.)
But I am now convinced that this is not the case and that there is another culprit. And that culprit is two separate factors, the first is extreme SEO optimization and exploit finding. There is a war as websites attempt to exploit googles SEO to get their sites to the top. So even if something is better its tough for google to tell and they instead give you the worse but better optimized result.
The second is more interesting, AI. In fact fully half of the internet is now AI.
If you have been remotely following the internet you will know google is not alone in having issues sorting it all out, amazon is filled with fake AI generated books as well.

And its going to get so, so much worse. Currently it isn't cost effective to fill forums with GPT 4 spambots or their more capable descendants, but it will be within a few years as the prices will drop exponentially.
Soon without a fundamental change there will be no way to tell if the person you are talking to on the internet is a real person.
There are a few ways to combat this that I can think of, the most obvious of which is getting rid of the anonymous internet as it exists entirely. This would have everyone have an Account linked to their real name as well as having pages and websites be tied to real people (or corps with real people) behind them. Companies wouldn't know *who* they are necessarily, but they would know they have a real person behind them and can't just make a thousand bot/AI accounts.
Something like this *will* need to happen by the way, the internet as it is will simply not function in the face of AI advancements and cost reductions.
---
These days when searching the only way I have to reliably tell what real people are thinking is is to look up reddit posts since those are still real people. As I thought at the time the API changes were 100% the right play, and they make reddit a place that still has value against the coming tide.

[KittyTac]

As someone who is currently under an authoritarian fascist regime: nope. Nope nope nope. I'd rather be in a bot-filled Internet than an Internet that would allow cost-effective surveillance.

And ngl, I find it very easy to tell someone real from a GPT bot. GPT has a very specific manner of responding, and doesn't have very much of a memory for distant events. It's not that I don't think some kind of solution is necessary, but de-anonymizing the Internet is not an acceptable one. It would create more problems than it solves, and is also logistically implausible to implement.

Not to mention that, afaict, the fidelity of text AI seems to have plateaued. It will never be a truly passable imitation of a person.

I suppose a possible solution is better detection, + legislation against fully AI-generated sites. I'm confident that the quality of the slop will not increase in a meaningful way for the foreseeable future, so just slapping these websites as they go up should be decent enough. And spambots should be up to forum admins to deal with (I suppose these detection tools should be made open-source).

[Laterigrade]

Soon without a fundamental change there will be no way to tell if the person you are talking to on the internet is a real person.
There are a few ways to combat this that I can think of, the most obvious of which is getting rid of the anonymous internet as it exists entirely. This would have everyone have an Account linked to their real name as well as having pages and websites be tied to real people (or corps with real people) behind them. Companies wouldn't know *who* they are necessarily, but they would know they have a real person behind them and can't just make a thousand bot/AI accounts.

this is not an internet I would like to be a part of

These days when searching the only way I have to reliably tell what real people are thinking is is to look up reddit posts since those are still real people. As I thought at the time the API changes were 100% the right play, and they make reddit a place that still has value against the coming tide.

I hadn’t considered that that might have been why they made the API changes, which makes them make a bit more sense. But this really isn’t true, there are plenty of fake accounts on reddit — comment-stealers, mass-upvote accounts, product-advertising bots. Reddit is a prime example of a bot-infested shitshow, if only somewhat less than most designated social media.