Odd, I've found that Gemini will completely fabricate the content of specific DOIs despite being corrected and even it providing a link to a paper which shows it is off about the title and subject of a paper it will cite. This obviously concerns me about its effectiveness as a research aide.
You're under the belief that private actors can't influence state actors to use violence on their behalf, completely isolating them from responsibility? If a private business calls the police on a suspected trespasser and the police shoot that person, is the business held liable? Ever? Seems like they have the better end of the bargain than the state.
>and the police shoot that person, is the business held liable? Ever? Seems like they have the better end of the bargain than the state.
Are you insane? When if ever are the agents of the state held responsible. If anything the civil suit against the business is more likely to go somewhere.
The fact that the state may "pay out" does not mean it has any serious incentive not to shoot the person dead so long as such payouts don't become too regular.
I owe Comcast $200, according to them. I've "owed" it for years. Can you imagine if I owed any government agency the same sum for the same time. I'd be arrested and thrown in jail for non-payment and/or some sort of quasi-contempt charge if I refused.
I deal regularly with landlords that owe government agencies hundreds of thousands to millions of dollars for years or even decades. None of them are in jail or even at risk of it.
This is a class of device usually called an "auto dialer" and have been around for quite some time. This one appears to be fairly low-featured. Newer devices will automatically send a text when a safe has been opened and the better ones are programmed with knowledge of the internal plate schematics so they can shorten the brute force process.
VistaVision is being used more often as a cheaper way to get to IMAX or 70mm projections sizes. The gear and filmstock is less expensive for production and you can laser out to the other formats at roughly the same resolution you capture at.
Not just building GPU infrastructure to serve inference at scale, but also constraining GPU infrastructure and pricing it out of reach for individuals who wish to pursue open models.
I'm not sure what entirely to do about it, but the fact that some providers still release open models is hopeful to me.