Rayman Pirate-Community

Funny story about that, which I just realized : I've recorded some sessions as a voice actor for IA training over the past two months. I just now make the link that the recording sessions ended three weeks before that AI friends sheananigan was released to the public.

The client for the training was Meta Inc. So perhaps those IA friends may have featured my voice in it. I guess we'll never know.

On an unrelated note : does anyone want to hire a freelance voice actor?

Pirez wrote: Thu Jan 09, 2025 9:52 am Tech CEOs are bowing up to Trump and his tendency to lie all the time and bully anyone who want to restore the truth.

Business is business. Some of those companies are under heavy regulatory pressure and they've realized they cannot afford to continue alienating a significant proportion of the population by disproportionately censoring conservative content and heavily promoting woke agendas.

There is a lot to say about Trump, but as far as those recent policy changes are concerned, I think it's a rather healthy change towards more freedom of expression and less political correctness. Of course, the risk is that it will help fake news and hateful content spread faster, but I hope they'll manage to strike a right balance this time.

Hunchman801 wrote: Tue Jan 14, 2025 9:07 pm

Pirez wrote: Thu Jan 09, 2025 9:52 am Tech CEOs are bowing up to Trump and his tendency to lie all the time and bully anyone who want to restore the truth.

Business is business. Some of those companies are under heavy regulatory pressure and they've realized they cannot afford to continue alienating a significant proportion of the population by disproportionately censoring conservative content and heavily promoting woke agendas.

There is a lot to say about Trump, but as far as those recent policy changes are concerned, I think it's a rather healthy change towards more freedom of expression and less political correctness. Of course, the risk is that it will help fake news and hateful content spread faster, but I hope they'll manage to strike a right balance this time.

I think this will also have a very healthy effect on AI, as I personally would like AI to stay away from the manipulataions of any ideologies. A good AI is an objective AI in my book, I don't want it to spew out a skewed answer.

Greengoop wrote: Thu Jan 09, 2025 6:08 pm Do you mean art he generated or art made by his fans?

Art he generated.

Elite Piranha wrote: Wed Jan 01, 2025 10:01 pm Something that caught my attention:

Facebook Thinks AI Bots Will Be Your New Social Media Friends

Connor Hayes, Meta’s vice president of product for generative AI, told the Financial Times last week that the company expects AI bots will “exist on our platforms, kind of in the same way that accounts do.”

“They’ll have bios and profile pictures and be able to generate and share content powered by AI on the platform,” Hayes added.

This made me think that I would be amusing to see a Razorbeard AI account with a bunch of Henchman bot accounts following him, being constantly annoyed by Teensie accounts making memes of the Robo-Pirates. Also, penguinz0 video about it:

I think this will surely blow up in Meta's face, I don't think people are enthousiastic about this at all. Also it's funny to see a Charlie video here, I watch his videos and especially his tier lists. I definitely believe most comments and top comments are always bots, when you click on their profiles they look too artificial.

Hunchman801 wrote: Tue Jan 14, 2025 9:07 pm

Pirez wrote: Thu Jan 09, 2025 9:52 am Tech CEOs are bowing up to Trump and his tendency to lie all the time and bully anyone who want to restore the truth.

Business is business. Some of those companies are under heavy regulatory pressure and they've realized they cannot afford to continue alienating a significant proportion of the population by disproportionately censoring conservative content and heavily promoting woke agendas.

I know this response is coming late, but are they really under that much regulatory pressure? Or have they been under-regulated for years and hoping that it stays that way by throwing their weight behind a regulation-skeptical* administration? As for censoring conservative content, I'm not sure that lines up with my experience of YouTube recommendations over the past few years, but then I do have watch history and profiling deliberately turned off, so who knows…

* Yeah, I know that's the American spelling but the European one makes no sense to me…

It ain't a good day for NVIDIA...
Any thoughts about DeepSeek? Did any of you give it a try?

Yeah, I heard about this from a colleague today. He told me he heard it was MIT-licenced, and indeed it seems they've released both the model and the weights it uses under that licence. I haven't had a chance to go through it properly though.

I also heard it can explain its reasoning instead of just producing a bunch of content like other LLMs (but doesn't Copilot cite sources?), and that it's more power-efficient. But maybe that's just hype, I'm not in a position to critique that right now (or ever, really, not being an expert…)

PluMGMK wrote: Sun Jan 26, 2025 8:49 pm I know this response is coming late, but are they really under that much regulatory pressure? Or have they been under-regulated for years and hoping that it stays that way by throwing their weight behind a regulation-skeptical* administration?

Let's just say they are under heavier regulatory pressure. Whether it is much to begin with is rather subjective and depends on one's approach to regulation. Among the various ongoing investigations and trials, I find some rather justified and others a bit otherworldly and not in the interest of the consumer, so it's hard for me to make broad statements about this!

PluMGMK wrote: Sun Jan 26, 2025 8:49 pm As for censoring conservative content, I'm not sure that lines up with my experience of YouTube recommendations over the past few years, but then I do have watch history and profiling deliberately turned off, so who knows…

Well, many have claimed that, and that was also my impression, though I must admit I'm not aware of any trustworthy studies on the matter. For example, I know of this one Instagram account that was sharing facts related to mass immigration and woke culture, without comments, and yet was suspended for a long time until Meta's sudden change of policy. But again, isn't everyone's perspective always a little biased?

DaveRattlehead wrote: Mon Jan 27, 2025 8:12 pm Any thoughts about DeepSeek? Did any of you give it a try?

Nope, though it must be quite good as its models are now respectively 4th and 8th in the Chatbot Arena leaderboard. However, the sell-off seems rather excessive to me. Its training budget is a very efficient figure (low OPEX), but comparing it to Meta's Llama is irrelevant: experts deem that OpenAI and Google's latest models were trained on similar OPEX budgets. I've also seen favorable price comparisons with OpenAI's ChatGPT-4o, but they overlook the fact that Gemini 2.0 Flash is cheaper than both. A correction was probably due, but I don't see it changing the dynamics of the AI economy in the long term.

Given the Chinese ownership of DeepSeek, there are naturally certain things it can't talk about

Hunchman801 wrote: Tue Jan 28, 2025 6:13 pm

PluMGMK wrote: Sun Jan 26, 2025 8:49 pm I know this response is coming late, but are they really under that much regulatory pressure? Or have they been under-regulated for years and hoping that it stays that way by throwing their weight behind a regulation-skeptical* administration?

Let's just say they are under heavier regulatory pressure. Whether it is much to begin with is rather subjective and depends on one's approach to regulation. Among the various ongoing investigations and trials, I find some rather justified and others a bit otherworldly and not in the interest of the consumer, so it's hard for me to make broad statements about this!

Yeah, I'll take that point. Traditionally I've gotten nervous when people talk about more regulation for web platforms, since such things could conceivably make it harder for the two of us to run this place the way the community has come to expect… But the super-platforms that exist now are able to amplify lies and incitement to violence way more than traditional broadcast media ever could, and yet don't seem to be subject to the same level of regulation… As for the investigations ongoing, I can't bring any to mind right now, but yeah, I seem to recall some having me going "really, of all the things to look at, you're looking at that? "

Lately I've been wondering what the world would be like if SOPA had passed in 2012 and created the precedent that scared people so much at the time… Probably not that different at all, but you never know…

Recently I noticed that someone I knew replaced their Facebook profile with an illustration of them in the style of a Studio Ghibli movie, and then I learned about this:



This is the video from 2016 of Hayao Miyazaki that was mentioned:



The quotes that really stuck with me:

An artificial intelligence could present us grotesque movements, which we humans can't imagine.

This one aged pretty well.

I strongly feel that this is an insult to life itself.

Well, we would like to build a machine that can draw pictures like humans do.

I feel like we are nearing to the end of times. We humans are losing faith in ourselves...

We are heading towards a future where AI is going to be present in our lives (whether we like it or not), and the sooner we come to terms with this, the better. I already know a couple of people who have been completely replaced by artificial intelligence, and they've had to find another job because it's going to be pretty hard for them to get back into practice. It's going to be a matter of adapt or die. Jobs will disappear and new ones will appear, just as they always have.

Regarding art? I do believe that an AI cannot yet catch up with human creativity. However, give it some time... Maybe they will end up regulating it, but most likely the artists of the future will have to combine their skills with the use of AI in order to continue practicing.

As for my work, I have to say that it is quite unlikely that it will be able to replace us in the short term (they would have to invent quite advanced robots to make experimental work in the lab), but it's incredibly useful for the computer work. The most recent example I have is in a bibliographic search for a specific type of reaction. What once took a whole morning to find what I wanted, the AI solved it in a matter of seconds with an advanced search. That is honestly a brutal change, since last year it didn't even know how to do basic calculations in our field.

WOW, I just saw a major government agency (don't know the correct wording) use, what it looks like AI generated Studio Gibli images, to promote their services .

What's going on?

I mean, I know that Trump and other politicians have also posted AI generated images, but for me, using them to promote official information provided by governments, seems kind of bizarre, but maybe it is just me.

I hate how AI has kinda ruined awesome photography, because now everyone questions whether AI has been use to enhance the photo.

DaveRattlehead wrote: Mon Apr 14, 2025 7:49 pm I already know a couple of people who have been completely replaced by artificial intelligence, and they've had to find another job because it's going to be pretty hard for them to get back into practice.

It always angers me when I hear this, because not only can their service not be replicated and it is quite literally robbing people of an income, but AI is way too early into development to actually give any valuable input.

Elite Piranha wrote: Wed May 07, 2025 3:49 am WOW, I just saw a major government agency (don't know the correct wording) use, what it looks like AI generated Studio Gibli images, to promote their services .

What's going on?

I mean, I know that Trump and other politicians have also posted AI generated images, but for me, using them to promote official information provided by governments, seems kind of bizarre, but maybe it is just me.

Elect a clown, expect a circus.

WOW, I just saw a major government agency (don't know the correct wording) use, what it looks like AI generated Studio Gibli images, to promote their services .

The thing is though, it can only do this because Ghibli (and any style heavily influenced by it) is reinforced in human consciousness.

Theoretically, artists have an advantage wherever they can do anything novel, because these models will not be able to reproduce it accurately. One might contest that it forces everyone into their own niches and rewards the kind of creativity that takes dramatic leaps outside the familiar and known.

But if true AI emerges that can form its own conceptions of things from primitive elements, all of that changes. I can't imagine what it will look like because it depends largely on the shape it takes, who builds it and for what purpose and so on.

My now-former employer rolled out a new PowerPoint template a few months ago, with several options for section-divider slides with different pictures. All of them were really cheap AI slop. One of them included "solar panels" that had random noise across them instead of a regular grid of traces

Also, this article came out in the last few minutes: https://thedailywtf.com/articles/ai-the ... d-the-ugly

PluMGMK wrote: Wed May 07, 2025 12:38 pm My now-former employer rolled out a new PowerPoint template a few months ago, with several options for section-divider slides with different pictures. All of them were really cheap AI slop. One of them included "solar panels" that had random noise across them instead of a regular grid of traces

Also, this article came out in the last few minutes: https://thedailywtf.com/articles/ai-the ... d-the-ugly

There is this saying I've heard this week when listening to this woman's retelling of being human trafficked (yeah heavy, I know) that really touched me and seems to be applicable in all manner of ways to our current society, especially when it concerns loneliness.

"If we don't stop seeing each other as disposable, we'll build a world where none of us matter.

This hit me like a truck.

I recently found about that a big spanish One Piece forum is having trouble because of bot attacks, possibly chinese AIs, causing an overload of 10000 active sesions and filling their space with activity logs.

Source: https://x.com/PiratekingES/status/1933620052923674929

The thing is, these situations aren't even necessarily "attacks", they can just arise from greedy bots scraping websites for training data, as noted by Drew DeVault here: https://drewdevault.com/2025/03/17/2025 ... on-me.html

It's happened here a couple of times, which is why guests now have to deal with annoying CloudFlare messages a lot of the time. A clear instance of "this is why we can't have nice things"