@FaceDeer@fedia.io
@FaceDeer@fedia.io avatar

FaceDeer

@FaceDeer@fedia.io

Basically a deer with a human face. Despite probably being some sort of magical nature spirit, his interests are primarily in technology and politics and science fiction.

Spent many years on Reddit and then some time on kbin.social.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

j4k3,
@j4k3@lemmy.world avatar

From my experience with Llama models, this is great!Not all training info is about answers to instructive queries. Most of this kind of data will likely be used for cultural and emotional alignment. At present, open source Llama models have a rather prevalent prudish bias. I hope European data can help overcome this bias. I can easily defeat the filtering part of alignment, that is not what I am referring to here. There is a bias baked into the entire training corpus that is much more difficult to address and retain nuance when it comes to creative writing. I’m writing a hard science fiction universe and find it difficult to overcome many of the present cultural biases based on character descriptions. I’m working in a novel writing space with a mix of concepts that no one else has worked with before. With all of my constraints in place, the model struggles to overcome things like a default of submissive behavior in women. Creating a complex and strong willed female character is difficult because I’m fighting too many constraints for the model to fit into attention. If the model trains on a more egalitarian corpus, I would struggle far less in this specific area. It is key to understand that nothing inside a model exists independently. Everything is related in complex ways. So this edge case has far more relevance than it may at first seem. I’m talking about a window into an abstract problem that has far reaching consequences. People also seem to misunderstand that model inference works both ways. The model is always trying to infer what you know, what it should know, and this is very important to understand: it is inferring what you do not know, and what it should not know. If you do not tell it all of these things, it will make assumptions, likely bad ones, because you should know what I just told you if you’re smart. If you do not tell it these aspects, it is likely assuming you’re average against the training corpus. What do you think of the intelligence of the average person? The model needs to be trained on what not to say, and when not to say it, along with the enormous range of unrecognized inner conflicts and biases we all have under the surface of our conscious thoughts. This is why it might be a good thing to get European sources. Just some things to think about.

What is the Legal copyright on a Lemmy Post?

Most instances don’t have a specific copyright in their ToS, which is basically how copyright is handled on corporate social media (Meta/X/Reddit owns license rights to whatever you post on their platform when you click “Agree”). I’ve noticed some people including Copyright notices in posts (mostly to prevent AI use). Is...

stinerman, (edited )
@stinerman@midwest.social avatar

IANAL. However I know a bit.

which is basically how copyright is handled on corporate social media (Meta/X/Reddit owns license rights to whatever you post on their platform when you click “Agree”).

Yes, this is how it works. You give them a license to your posts.

I’ve noticed some people including Copyright notices in posts (mostly to prevent AI use). Is this necessary, or is the creator the automatic copyright owner?

The creator automatically owns the copyright. People can put in license terms, but they’re effectively useless in this context. Let’s say OpenAI violates the copyright on your post (it’s still an open question whether or not training AI on copyrighted data constitutes copyright infringement, but we’ll assume it does). Your only recourse is to sue them if they do this. Because you never registered the copyright, you’re limited to recovering actual damages – if you do register the copyright you can get statutory damages, which are up to $150k per violation. So how much money did you lose on the ability to commercially exploit this post that OpenAI took away from you by copying your posts? Less than the cost to bring the suit, I’m sure.

So the TL;DR here is that the anti-AI licensing thing is only effective if you’re registering the copyright on your posts/comments. And even then, that’s only true if AI training is considered to be copyright infringement.

Balinares,

The Spathi warplane is coming along nicely.

Police make arrests in killing of B.C. [Canada] Sikh activist Hardeep Singh Nijjar (www.cbc.ca)

Canadian police have arrested members of an alleged hit squad investigators believe was tasked by the government of India with killing prominent Sikh separatist Hardeep Singh Nijjar in Surrey, B.C. last June, CBC News has learned....

Doesn't the need for a permit fundamentally contradict the US's ideals of free speech?

I went to some palestine protests a while back, and was talking to my brother about the organizing, when revealed something I found pretty shocking, we (the protesters) had acquired a permit to hold the protest. Apparently this is standard policy across the US....

southsamurai,
@southsamurai@sh.itjust.works avatar

It’s a matter of when and where.

A protest isn’t inherently able to be made anywhere any time. Public property is shared property, so in order for one group to use that to the exclusion of others takes planning. On private property, nobody is obligated to allow you there at all, and not all universities are public.

The street you block in protest isn’t your street, it’s our street, so you have no more or less rights to it than I do. A permit, in theory, is just a fancy way to make sure that anyone else that would be using that place can have alternatives. Traffic being rerouted, etc.

There are also health and safety issues like overloading bleachers being a risk, or a space not being able to be evacuated in an emergency because of too many people.

You’re right, if permits are not granted equally, it’s a huge problem, and abuses can occur. They do occur. But as long as the entity (be it a government or a university or business) isn’t placing undue obstacles, grants permits equally, and everything is done without corruption (bribes or such), that’s how the most basic peaceful protest has to work.

This isn’t saying that there can’t be other forms of protest. There are many forms available. But a simple awareness protest? It needs to be peaceful, cooperative, and only marginally inconvenient. Remember, that kind of protest is about awareness, of making sure people know and possibly gaining support. It’s a flies with honey vs vinegar.

Now, illegal protests, disruptive protests, and even violent protests have their place too. As do rallies, which can be seen as a type of protest, but isn’t really the same imo. But you have to choose when and where to apply those tools effectively. Some protests, the entire point is to cause disruption, get arrested, and use that as a tool to achieve a goal. But a cooperative legal protest can do things that type can’t.

Remember, the right to free speech does not guarantee you can exercise it everywhere. Your right to free speech ends at your neighbor’s door, so to speak. You can say what you like, but she doesn’t have to let you say it on her porch. You can be asked to leave, and should you refuse, be removed after some hoop jumping along the way (which varies by state and municipality).

A public space like a street is the same basic thing. You have a right to use it, and so does everyone else. Your freedom and their freedom may come into conflict. Having some kind of system for resolving that issue is necessary. Permits work for parades, parties, and protests.

Even with the kind of anarchy that most people identifying as anarchists espouse, there’s a social contract involved in that kind of thing. It’s only a matter of the mechanism involved.

Besides, there’s always been limits on speech. Collectively, there’s an understanding that there are limits where, when, and how they can be expressed. The whole “fire in a theater” trope is an example. There’s “fighting words” laws on the books. The only question is what any given social structure has decided those limits are.

Generally, whether it works correctly or not, representative democracy promises that the limits are agreed on via those we elect to do so. That has flaws of course, but that’s the US on a simplistic level. We’ve all agreed to the limits in one way or another, or can choose to try and change those limits (or the social structure itself).

But if that slippery slope slides too far, well, that’s when protests become revolution, in theory. Assuming enough people agree and work together. It doesn’t seem to happen very often, but that’s the ultimate safeguard against a democracy failing.

Generative AI can turn your most precious memories into photos that never existed (www.technologyreview.com)

Dozens of people have now had their memories turned into images in this way via Synthetic Memories, a project run by Domestic Data Streamers. The studio uses generative image models, such as OpenAI’s DALL-E, to bring people’s memories to life. Since 2022, the studio, which has received funding from the UN and Google, has...

"Permission is Hereby Granted" -- MIT License text becomes viral “sad girl” piano ballad generated by AI (arstechnica.com)

On Wednesday, prompt engineer Riley Goodside tweeted an AI-generated song created with the prompt “sad girl with piano performs the text of the MIT License,” and it began to circulate widely in the AI community online.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • tester
  • DreamBathrooms
  • thenastyranch
  • magazineikmin
  • osvaldo12
  • ethstaker
  • Youngstown
  • mdbf
  • slotface
  • rosin
  • ngwrru68w68
  • kavyap
  • GTA5RPClips
  • provamag3
  • cisconetworking
  • InstantRegret
  • khanakhh
  • cubers
  • everett
  • Durango
  • tacticalgear
  • Leos
  • modclub
  • normalnudes
  • megavids
  • anitta
  • lostlight
  • All magazines