Father, Hacker (Information Security Professional), Open Source Software Developer, Inventor, and 3D printing enthusiast
Maybe take pictures of them instead?
This doesn't make sense when you look at it from the perspective of open source models. They exist and they're fantastic. They also get better just as quickly as the big AI company services.
IMHO, the open source models will ultimately what pops the big AI bubble.
If only they would actually spend it on things that benefitted the economy instead of:
- Buying land/houses
- Buying successful businesses in order to loot them (typical Private Equity moves)
- Lobbying the government to cut their taxes and reducing social safety nets
I firmly believe that a big reason why wealth is consolidating so quickly is because there's not much for the rich to spend their money on these days. It used to cost a fortune to care for a great big mansion and the surrounding grounds (and you'd have a local economy built on that). These days, it's not even a rounding error in their monthly income from interest and dividends and you don't even need to hire very many people.
There used to be actual luxury services the rich needed to pay for if they wanted to appear better off than the riffraff. Now the common man has access to all those luxuries and more.
Once you've got a couple private jets and a yacht (that you never spend any time on), what the fuck do you need all that money for‽ Even they don't know what to do with it!
Larry Ellison—the original tech villain billionaire—bought a huge private island in Hawaii. He's never there. It just... Sits there. With a small staff and places for his yachts. It's like he wants to go down in history as one of the biggest, greedy scumbags of all time.
Stick Enthusiasts!
No, a .safetensors file is not a database. You can't query a .safetensors file and there's nothing like ACID compliance (it's read-only).
Imagine a JSON file that has only keys and values in it where both the keys and the values are floating point numbers. It's basically gibberish until you go through an inference process and start feeding random numbers through it (over and over again, whittling it all down until you get a result that matches the prompt to a specified degree).
How do the "turbo" models work to get a great result after one step? I have no idea. That's like black magic to me haha.
It was part of a 3-step program.
Or, with AI image gen, it knows that when some one asks it for an image of a hand holding a pencil, it looks at all the artwork in it's training database and says, "this collection of pixels is probably what they want".
This is incorrect. Generative image models don't contain databases of artwork. If they did, they would be the most amazing fucking compression technology, ever.
As an example model, FLUX.dev is 23.8GB:
https://huggingface.co/black-forest-labs/FLUX.1-dev/tree/main
It's a general-use model that can generate basically anything you want. It's not perfect and it's not the latest & greatest AI image generation model, but it's a great example because anyone can download it and run it locally on their own PC (and get vastly superior results than ChatGPT's DALL-E model).
If you examine the data inside the model, you'll see a bunch of metadata headers and then an enormous array of arrays of floating point values. Stuff like,
[0.01645, 0.67235, ...]. That is what a generative image AI model uses to make images. There's no database to speak of.When training an image model, you need to download millions upon millions of public images from the Internet and run them through their paces against an actual database like ImageNET. ImageNET contains lots of metadata about millions of images such as their URL, bounding boxes around parts of the image, and keywords associated with those bounding boxes.
The training is mostly a linear process. So the images never really get loaded into an database, they just get read along with their metadata into a GPU where it performs some Machine Learning stuff to generate some arrays of floating point values. Those values ultimately will end up in the model file.
It's actually a lot more complicated than that (there's pretraining steps and classifiers and verification/safety stuff and more) but that's the gist of it.
I see soooo many people who think image AI generation is literally pulling pixels out of existing images but that's not how it works at all. It's not even remotely how it works.
When an image model is being trained, any given image might modify one of those floating point values by like ±0.01. That's it. That's all it does when it trains on a specific image.
I often rant about where this process goes wrong and how it can result in images that look way too much like some specific images in training data but that's a flaw, not a feature. It's something that every image model has to deal with and will improve over time.
At the heart of every AI image generation is a random number generator. Sometimes you'll get something similar to an original work. Especially if you generate thousands and thousands of images. That doesn't mean the model itself was engineered to do that. Also: A lot of that kind of problem happens in the inference step but that's a really complicated topic...
Officer: "I got better!"
- JumpDeleted
Permanently Deleted
I'm ok with rich people getting charged more. But anyone who isn't making like $1 million/year should get the normal price.
I assume you missed "scrolling" because of all the alcohol?
Shit. I just realized I went straight to the comments!
This will definitely encourage more people to have kids.
Like I said initially, how do we legally define "cloning"? I don't think it's possible to write a law that prevents it without also creating vastly more unintended consequences (and problems).
Let's take a step back for a moment to think about a more fundamental question: Do people even have the right to NOT have their voice cloned? To me, that is impersonation; which is perfectly legal (in the US). As long as you don't make claims that it's the actual person. That is, if you impersonate someone, you can't claim it's actually that person. Because that would be fraud.
In the US—as far as I know—it's perfectly legal to clone someone's voice and use it however TF you want. What you can't do is claim that it's actually that person because that would be akin to a false endorsement.
Realistically—from what I know about human voices—this is probably fine. Voice clones aren't that good. The most effective method is to clone a voice and use it in a voice changer, using a voice actor that can mimick the original person's accent and inflection. But even that has flaws that a trained ear will pick up.
Ethically speaking, there's really nothing wrong with cloning a voice. Because—from an ethics standpoint—it is N/A: There's no impact. It's meaningless; just a different way of speaking or singing.
It feels like it might be bad to sing a song using something like Taylor Swift's voice but in reality it'll have no impact on her or her music-related business.
wouldn't we expect countries with strong social programs like Norway to have much higher birth rates? I suppose those social programs would tend to correlate with birth control
I was unfamiliar with Norway's program so I looked it up...
49 weeks of maternity leave? FUCK YEAH!
$160/month (USD equivalent) for kids under 6? Not nearly enough! That is of negligibe impact and doesn't come close to offsetting the costs of raising a child.
My two takeaways from this, learning about Norway's programs:
- The most impactful change was paid paternity leave! Turns out, letting dads stay home too resulted in a fertility rate increase from 1.6 to 1.9!
- Subsidized daycare increased the fertility rate from 1.9 to 1.98.
- The most recent drops in the fertility rate seem to be tied to the increased cost of housing. Meaning: All those benefits are great and all but they can't make up for the fact that no one can afford their own home and kids anymore.
Also, "when everyone gets a subsidy, no one gets a subsidy" (my own saying). It seems inevitable that daycare costs would increase by the subsidy amount in order to capture it as profit. Basically, long-term subsidies like that ultimately fail because of basic economics. They can work fine in the short term, though.
I still stand by what I said: Having kids makes you less economically stable and until we fix that, fertility rates will continue to decline.
Seems like the biggest thing that needs to be fixed though is housing costs.
- JumpDeleted
ceos job is playing golf
especially the ones made over the injections of workers
Well there's the problem! As good as it sounds, you actually lose a lot of the nutrition when employees are processed into injectable paste. Ultra processed workers are bad for you.
Eat them raw as capitalism intended!
These are the same people that would download a car!
Pollution would make sense if people were trying to have kids but couldn't. But they're not trying to have kids at all!
The more likely explanation—related to tech—is that we don't need kids anymore. For 99% of human history, children were necessary and not having kids was basically impossible (horny kids and no birth control). Kids were how humans kept alive/stable as well as expanded their power and influence! It's also how they got cared for in old age (though that's a much lesser concern because I seriously doubt humans of the past thought that hard about such things when living to 40 was considered amazing).
Now we have birth control and—in Western societies—stability/safety is much more likely if you don't have kids. We've basically flipped the script on our evolution.
You want people to have kids? Flip the script back! Make anyone under 30 without kids pay a massive tax that pays for the kids of people who have them! Basically, make everyone who didn't have kids pay child support.
Make having kids the best damned economic decision anyone can make with diminishing returns after two (kids).
You make AI voice generation sound like it's a one-step process, "clone voice X." While you can do that, here's where it's heading in reality:
"Generate a voice that's sounds like a male version of Scarlett Johansson".
"That sounds good, but I want it to sound smoother."
"Ooh that's close! Make it slightly higher pitch."
In a process like that, do you think Scarlett Johansson would have legal standing to sue?
What if you started with cloning your own voice but after many tweaks the end result ends up sounding similar to Taylor Swift? Does she have standing?
In court, you'd have expert witnesses saying they don't sound the same. "They don't even have the same inflection or accent!" You'd have voice analysis experts saying their voice patterns don't match. Not even a little bit.
But about half the jury would be like, "yeah, that does sound similar." And you could convict a completely innocent person.
The problem: AI slop is replacing a tiny percentage human slop but the growth of AI data centers might mean that they can replace a greater percentage of human slop in the future.
Meanwhile, executives at big studios are continuing their campaign of producing boring, unoriginal shows. Their great hope upon hope is that they too will be able to replace their boring, unoriginal shows with boring, unoriginal shows made by AI.
Working on (some) AI stuff professionally, the open source models are the only models that allow you to change the system prompt. Basically, that means that only open source models are acceptable for a whole lot of business logic.
Another thing to consider: There's models that are designed for processing: It's hard to explain but stuff like Qwen 3 "embedding" is made for in/out usage in automation situations:
https://huggingface.co/Qwen/Qwen3-Embedding-8B
You can't do that effectively with the big AI models (as much as Anthropic would argue otherwise... It's too expensive and risky to send all your data to a cloud provider in most automation situations).