Every AI Existential Risk Explained

Join my Discord to discuss this video:

Every AI Risk gets explained in 20 minutes! (please keep in mind that all of this is highly speculative)

Subscribe and activate the bell!

— TIMESTAMPS —
00:00 WMD Democratization
03:33 Jobless Society
05:19 AI Arms Race
06:46 Value Alignment
09:33 Cryptography Destruction
10:19 Connected Brains Theory
11:32 Risky Experiments
14:37 Generative AI
15:16 Behavioral Sink
16:35 Grey Goo
17:03 Superintelligent Destruction
17:53 Other Side of The Coin

— SOURCES —

— DISCLAIMER —
Do not use this video as your only source of information. This video is for entertainment/edutainment purposes, and some information could be too oversimplified or incorrect. This channel's goal is to spark your curiosity and let you do your own research on these topics.

[AdrianSal-ry]

Fun fact: the opening scene in terminator took place in 2029

[RCGC]

‘All it would take for humanity to end would be one single person who wants it to end’ That’s deep

[adzn]

i suppose we draw closer to the “i have no mouth, and i must scream” type of future

[brokenhabits]

Just imagine that if things exponentially accelerate, a lot of these predictions could become reality by 2040. In the year 2040 there will still be many people living that were born in the 1970s...Just think about how different that new world would be for the people that grew up in the 80s and 90s. That's how fast things are moving.

[vardaant]

thanks for the daily dose of existential dread 👍

[trenperse]

"Downloading Social Skills" is the only good part about this video

[jrylandsiii]

we really living in the future they warned us about

[crpineo]

Goddammit, I didn't need even _more_ anxiety about the future...

[SilverLovesJesus]

Loving seeing this guy change up the “every kind of X explained” style and unique things with it.

[mse]

11:20 bro in the 2038 groupchat sending 3 terrabytes of 2020 era Among Us brain rot...

[wk]

General Artificial Intelligence is going to be a lot harder than we think. In the sixties computer scientists thought image recognition was just a few years away. But rather it took over 50 years to achieve. The leap between a program being able to essentially summarize the internet is just a drop in a bucket compared to AGI.

At this point, we don’t even know if you can have AGI without consciousness. And we have no idea what consciousness is.

My only prediction is that in 10 years we will look back at today’s AGI hype as pretty silly.

[YokaK]

This is like how people in the 70s thought we would have flying cars by now

[stewiegriffin]

Thank you for crushing my hopes and dreams.

[Atypical_Orthodoxy]

I'd like to be the first to say "This video didn't age well" so somebody in the future doesn't have to type it out.

[PineSap]

In regards to value alignment. GPT2, they were trying to get it to produce harmful results, but the developer forgot a single “-“ minus sign and rather than producing what they anticipated it become incredibly lustful and pornographic. The fact that a single keystroke can entirely change how it behaves is a cause for concern.

[AndreaLina-rh]

So, in 2030, you’re telling me, that Monika will finally be freed?

Asking for a friend of course….yeah. 🤞🏻

[rickwrites]

The rat/mouse utopia is not counterintuitive at all. They overbred due to good food and no predators, but were not given extra space for the extra population. This created conflict and the development of ways to depopulate to offset the imbalance. This is basic evolution and ecology stuff.

[Bluerbinn]

reminder this is going to happen in my lifetime! how awesome!

[ocdkirby]

This entire video needs the South Park “this is what effective altruists actually believe” text, lol. I love how many of these come from reading too much sci-fi and having a playground sandbox understanding of machine learning and theory of mind, thank you for increasing exposure to how silly it all is.

[sometimes]

we're beyond fucked, gotta live our lives to the fullest as much as we can