realitista

joined 1 year ago
[–] realitista@lemmy.world 3 points 2 months ago (1 children)

If they are using GPL code, shouldn't they also release their source code?

[–] realitista@lemmy.world 1 points 6 months ago

See my post above in the thread where I show the laws I am talking about and cite source.

[–] realitista@lemmy.world 1 points 6 months ago

See my post above with citation.

[–] realitista@lemmy.world 2 points 6 months ago

This article summarizes the subsidies I'm talking about. Here's an excerpt:

For now, the important point is that trucks generally are more profitable than cars thanks to two big government incentives, both of them historical footnotes.

The first is the so-called chicken tax, a 25 percent tariff imposed by Lyndon B. Johnson in 1964 on foreign-built work vehicles as part of a chicken-related trade war with Europe. If you’re making a pickup or cargo van in the United States, profits should be higher, because foreign factories can’t come close to undercutting you on price.

The second incentive lies in the fine print of Corporate Average Fuel Economy (CAFE) standards adopted in 1975, Gerald Ford’s reluctant response to a crippling Middle East oil embargo that sent gas prices soaring. To protect American commerce, work trucks and light trucks were subject to less-strict CAFE standards than family sedans. Trucks are also exempt from the 1978 gas guzzler tax, which adds $1,000 to $7,700 to the price of sedans that get 22.5 or fewer miles to the gallon.

[–] realitista@lemmy.world 33 points 6 months ago (14 children)

That's because the USA subsidizes bigger trucks as "work vehicles". This practice needs to stop and they need to be taxed more than smaller vehicles.

[–] realitista@lemmy.world 15 points 6 months ago

Until his profile gets high enough that they find some permit he doesn't have and he gets shut down.

[–] realitista@lemmy.world 1 points 6 months ago

I was working tech in the Bay Area in the '90s, I remember it well.

[–] realitista@lemmy.world 8 points 6 months ago (2 children)

Probably you are right, but so far no one has demonstrated any LLM that can be controlled within these tight types of adjustments and it feels like it might be something that the technology just never is able to do. We might have to wait for a whole new generation of technology for this.

[–] realitista@lemmy.world 8 points 6 months ago (5 children)

It will be really interesting to see how long it actually takes before this can be done accurately enough to execute a directors vision and high quality enough to actually make a film from. It could be anything from a few months to decades, it's so hard to know how much we are actually able to control these models to get them to do what we really want accurately enough.

[–] realitista@lemmy.world 16 points 6 months ago* (last edited 6 months ago)

This would really be the ultimate fuck you to everyone if they did this. I sort of wish they would so we'd all stop giving them all our personal data.

[–] realitista@lemmy.world 5 points 7 months ago

Not clear from this what the difference is between a TPU and an LPU.

[–] realitista@lemmy.world 11 points 7 months ago

Yes, however recruiters see someone like this and see a great opportunity to fill their quota, so it all balances out.

 

Google’s new video generation AI model Lumiere uses a new diffusion model called Space-Time-U-Net, or STUNet, that figures out where things are in a video (space) and how they simultaneously move and change (time). Ars Technica reports this method lets Lumiere create the video in one process instead of putting smaller still frames together.

Lumiere starts with creating a base frame from the prompt. Then, it uses the STUNet framework to begin approximating where objects within that frame will move to create more frames that flow into each other, creating the appearance of seamless motion. Lumiere also generates 80 frames compared to 25 frames from Stable Video Diffusion.

Beyond text-to-video generation, Lumiere will also allow for image-to-video generation, stylized generation, which lets users make videos in a specific style, cinemagraphs that animate only a portion of a video, and inpainting to mask out an area of the video to change the color or pattern.

Google’s Lumiere paper, though, noted that “there is a risk of misuse for creating fake or harmful content with our technology, and we believe that it is crucial to develop and apply tools for detecting biases and malicious use cases to ensure a safe and fair use.” The paper’s authors didn’t explain how this can be achieved.

Synopsis excerpted from The Verge article.

 
 
 
 
view more: next ›