The first thing that we humans made were weapons and since then everything we make is considered first for its potential value as a defensive/offensive weapon.
-
Please AI will never experience pleasure or pain so it has no motivation for propagation or domination, it will always only magnify the human who has pushed the button Enter on the prompt.
-
The ultimate prompt "Find a way to eliminate human suffering without eliminating humans?"
-
If someone wants to train a physical task the data is sent to us for training, so we would not allow this.
And even if you somehow did that, the brain itself has knowledge of what it task is and what to do with it, and since it's running on very smart VLMs trained by the best labs, I expect they have protections naturally (on top of ours)
Wile E. Coyote is my hero, he has so much fun building his Roadrunner traps, so it is not the result the most important thing but the process which is more or equally important never less
More like crazy censorship that masquerades as anti piracy measures are incredibly short sighted measures by dumb governments that just hurt the honest users.
Pretty sure the pirates are all laughing their ass cheeks off right now.
It could, but it currently doesn't. Regulating and enforcing censorship is a lot harder when the offending material is not publicly accessible / easy to find. So even if you have offending content on your own private cloud, as opposed to have someone elses offending content on a shared public cloud, the censors would first have to find out about it.
If your connection to your private cloud passes through public internet, all bets are off. I don't think all companies have their private clouds at the basement level, completely owned data centers, connected via edge switches near the water cooler.
In my experience, a specific service in "The cloud" has an uptime similar to a single raspberry pi. Some services are more reliable, maybe upto three-nines.
If you design your systems as such - expecting them to fail, then using "the cloud" is fine. If a VM running in AWS goes down, fine, the ones in hertzner continue to work.
There's levels of availability you need -- what services can go down, in what geographic location, and for how long (do you need to go sub-second from all geographic locations? Because I'm not sure you can do that in any situations. On the other extreme are you happy with a TTL of 60 and failover to another IP, which barely counts as available in my book)
a back door should be the front door where the one who needs the information goes to the office of the information holder and accesses the information inside the provider's space with the proper court warrants.
whoever wins the election, Musk is supplying NASA with spaceships for now and the immediate future and next probably will take over Boeing to make self flying personal airplanes
There is hardly any hope in controlling human expression by whatever means you employ. It is like trying to stop the waves from arriving on the beach by using a toy plastic bucket.
Yeah, I'm not sure what people are really expecting from these "platforms".
At the scale of Instagram you're basically expecting them to act as a censor for billions of people. It's like imagining an agent on every street corner preventing people from discussing what they like in the pub. You might be able to influence the direction, and you might be able to use models to blanket ban certain things, but real moderation? You'd need to employ hundreds of millions.
we created the problem by abusing technology so I doubt more technology will solve it. We need to change our consumerism culture from throw away to keep forever.
