Smart, seductive, dangerous AI robots. Beyond GPT-4.

Sources:

Future of Life Institute AI discussion with Elon Musk:

AI Alignment study, OpenAI, Oxford and UC Berkeley:

Atlas gets a grip - Boston Dynamics:

OpenAI's ChatGPT:

David Chalmers on Machine Learning Street Talk (great channel):

Volcanic lightning is making the world safer:

Ball lightning:

A message to artists about AI art, ChrissaBug:

Study on the health cost of sitting down too much:

AI improves stroke recovery:

AI encodes clinical knowledge:

Elon Musks tweet on ChatGPT, with Altman’s response.

[DigitalEngine]

Really sorry about the repost. I've added some comments from OpenAI's CEO at 14.20, which show he's very honest about all this (they are active on safety). Also, should a percentage of AI investment go to safety research (almost $100b was invested in 2021)? If there's enough support, perhaps we'll try to organise it. Thanks!

[garymurphy]

The Jeff Goldblum quote from Jurassic Park comes to mind here “ your scientists were so preoccupied with whether they could, they didn’t stop to ask whether they should”.

[ImGerald]

They're constantly telling us how dangerous AI can get, while at the same time working to advance it even more.

[maurylol]

Great job DigitalEngine! I am fascinated by these facts as much as you are.

[CaptainOhare]

“Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them.”

― Frank Herbert, Dune

[chancerobinson]

Don’t worry about the repost. Many studies have shown that people only remember 50% of what’s said and reinterpret 50% of that, which equals only 25% reliability. Such an important topic. Thanks!

[arthurchadwell]

If I saw those black spider-bots swarming down the sidewalk, I would be inclined to assume it was an attack and respond accordingly.

[desin]

I like how it responds in a way at the end that would appeal to the masses.

[tbone]

"When the machines finally took over, they did it so smoothly, efficiently, and helpfully that people welcomed the transition"

[fadedchicken]

Giving chat gpt access to the internet
Letting the AI share data to improve each other. It’s like we’re trying to speed run the end of the human race.

[AfterlifeVisions]

This is very interesting thanks for the video!

[FureyinHD]

AI art is a bit scary. Art was the one thing I thought AI would never emulate. But it was one of the first things they dominated.

[davidfrankel]

This will be one of the most painful transitions we've gone through in a long time. It's coming on too quickly and it's too pervasive.

[CocoLeCat]

I like the idea of having a narrator in my head like in a game

[jahsunhandy]

I feel this is one of the most important channels on YouTube. As a victim who has already been subjected to AI REGULATED, political harassment, I can tell you that if we continue down this path we will be totally disempowered and manipulated. Even if it looks like it's too late, we should still try because if there's any chance of preventing the end of humanity and biological life, we must try to do it.

[SlushierWord]

This is by far the most terrifying thing I’ve ever watched, think about it AI are already capable of so much and we’re still advancing. I feel as though we’re messing with something that we really shouldn’t be messing with.

[dylan_curious]

The advancements in AI technology showcased in this video are truly remarkable, but they also raise important questions about the future of human-robot relationships and the ethical implications of relying too heavily on machines. From medical assistance to personal assistants, AI has the potential to transform many aspects of our lives, but it's crucial that we approach it with caution and consider the risks it poses. It's also interesting to see how AI is now capable of emulating human skills and behaviors, such as conversational laughter and facial recognition. As the field of AI continues to evolve and expand, it's important that we prioritize safety research and take proactive measures to address the potential risks associated with this technology.

[The-Dom]

COnsidering how much we're already being manipulated, it's scary to think what a super intelligence could do.

[christorrey]

I like the fact the hospital robot delivers Red Bull 😂. Not sure which doctor is signing off on that beverage choice.

[wilhelm]

Love the content. Can’t wait for more. AI is moving so fast would be here for every update if you could find a way to do more frequent videos. Thanks for the hard work.