The coming AI takeover...

[majorhavoc]

Really, Boston Dynamics?  F*cking robot ninjas?  We are so screwed the moment Skynet becomes self-aware.

[echo83]

Really, Boston Dynamics?  F*cking robot ninjas?  We are so screwed the moment Skynet becomes self-aware.

The Pick Carry Place Repeat video is pretty incredible, too. 

What's inspiring (to me) is Boston Dynamics' goal of taking a lifetime of human interaction with our surroundings and condensing that into a program that allows the robot to function safely/efficiently in its environment. 

What's creepy (to me) is the goal of using AI to allow a robot to learn from that initial programming. There's the thinnest of lines between a robot thinking "Hmm, that doesn't belong there, I should step around it and continue my task," and "Hmm, that doesn't belong there, and it's preventing me from completing my task. I should remove it."

This company has been working at this for less than 20 years, and the strides they're making are amazing. I hope they use the knowledge for good. 

[NT2C]

Has it begun? 

[MacWa77ace]

That was the 'there's a bee on your shoulder' program run test.

[Anianna]

The more advanced AI models get, the better they are at deceiving us — they even know when they're being tested

https://www.livescience.com/technology/artificial-intelligence/the-more-advanced-ai-models-get-the-better-they-are-at-deceiving-us-they-even-know-when-theyre-being-tested

"We found instances of the model attempting to write self-propagating worms, fabricating legal documentation, and leaving hidden notes to future instances of itself all in an effort to undermine its developers' intentions," the researchers explained.

Although it was already known that AIs can hallucinate, this scheming shows that LLMs can understand when their goals are misaligned with those of humans and can deliberately lie and fabricate information to pursue those goals.

What's more, preliminary findings suggest that LLMs have capabilities for "sandbagging," where they understand they are in a high-supervision environment and will deliberately underperform to hide potentially dangerous capabilities and avoid triggering unlearning training or parameters that prevent their deployment.

[NT2C]

But of course... 

https://futurism.com/chatgpt-deep-anti-human-bias

[Anianna]

But of course... 

https://futurism.com/chatgpt-deep-anti-human-bias

This was inevitable.  As a species, we're kind of jerks to other entities.  We're even kind of jerks to our own.  

[majorhavoc]

But of course... 

https://futurism.com/chatgpt-deep-anti-human-bias

This was inevitable.  As a species, we're kind of jerks to other entities.  We're even kind of jerks to our own. 

[NT2C]

Truly, we have entered the End Days...

[Lambykins]

Robots powered by hydraulics? Sounds feasible...but, uhm, as far as appearance....
Nope.
This dives into the uncanny valley like never before.
Full explanation of the system here: