AI and human extinction: Just another conspiracy theory?

AI elimination of humans will save the earth. Who needs carbon credits.

Huey, Dewey, and Louie, (silent running) will tend to the garden.

 

I dont think its a conspiracy that the people who develop AI want you to think its potentially super powerful and dangerous vs the cheap parlor trick it actually is. Why would you invest in the latter? Its just good old fashion American hucksterism, patent medicine, AI, same thing.

 

The current AI wave is just predictive in nature. It only works if there is stuff already out there where it can sort of "fill in the blanks" of what comes next after some query or prompt. Its like a glorified google search. If AI just starts generating enough stuff in place of people (like college essays or whatever), its just going to be a big feedback loop of garbage, because its not actually "intelligent."

 

It would likely do that, honestly (assuming you gave it enough reason to). They are sort of driven to say what you want, not tell you you're an idiot. It's predictive of what you want too! So if you lead it with questions, it will follow.

 

Yeap. It seems we are getting one of these statements/letters every week now. I also think huge egos play a role. These tech bros really believe they are creating life and consciousness.

LLMs will certainly have a big impact on jobs and how we perform them. But let’s be honest. These LLMs are really good at predicting the next word based on prompts and context. It’s impressive as hell. But that’s it, they are predicting the next word.

Musk has been promising fully autonomous driving for a decade. They have some of the best ML minds in the world working on it, pretty much unlimited budget. And while it’s pretty impressive, it still fails at aspects of driving that a below average human can handle with ease, while having a conversation on Bluetooth and flipping birds to other drivers. So spare me the ‘we are on the verge of skynet’ silliness.

 

One of the best shows in years to me, though not renewed, so not many agreed. But that was a big plot point - AI gives you what you think you need

 

Hey Robot, bring me a "shrubbery".
Robot sticks a Red-Wood down your throat.

 

"Musk has been promising fully autonomous driving for a decade."

What he didn't tell you is that, it works, but only to one destination, it drives the passenger(s) to the morgue.

 

I think it will send a few pedestrians there too

 

Like all gifts to mankind, it can definitely be exploited by the unethical.

 

This is true, but not a complete picture of what's currently being discussed.

There are three levels of AI

ANI (Narrow): This is what we have now. Siri, Alexa, etc.
AGI (General): When the machine is as smart as a person. There are several tests for this - none of which have been passed yet.
ASI (Super): Machines are smarter than us. Terminator time, baby!

It is largely believed that ANI will take a long time to create (and it has). As it get's strong, it will help create AGI and once at AGI it won't take long to get to ASI. If we get to AGI without proper guardrails, ASI may come to exist in ways we'd rather it not.

Or as Tim Urban once illustrated:

 

I think there is a bit of Zeno's paradox here though too, to reach a destination, you must travel halfway first, and so its just and endless set of half-ways.

 

Perhaps. Should we dismiss it as motionless and just hope it doesn't come to exist?

 

No I believe all things are dynamic in nature, but there is a reason technology should be democratized rather than controlled by tech billionaires who's main clients will be military contractors and nation states. A democratized economy would mean tech that benefits ordinary people, in our current set up, tech will most likely benefit people looking to deskill labor or for social control. That being said, hyping up AI as dangerous gets those governments interested in investing. So I think there is an element of Star Wars missile defense to this. Something that you can milk for decades with minimal results and breakthroughs that are just around the corner with more investment.

 

Ouch

 

Nearly every major technological jump has come at a cost to labor - so I would expect no difference here. The printing press, cotton gin, nor iphone threatened to actually kill humanity (well, I guess there is an argument for smartphones, depending on how you define "kill").

I don't dismiss your hypothesis of this being the big guns pushing for legislation to create an AI oligarchy to keep the ruffian start-ups out, but I also have a hard time not agreeing that this has the type of potential we've never experienced. That could be good or bad.

Is this a great filter of Fermi's paradox? If so, is this a filter of destruction, or a filter that creates so much bounty that it can change the brutal course of our species? Knowing what I know if humanity, I'm leaning towards the first one - thus my concern here.

 

I think the main threat of AI is just further flattening culture, at the expense of labor. That's why the WGA strike is really interesting. I think studios would love to crank out garbage with AI written scripts and eliminate the cost of "creatives" as much as possible. And that isnt even a business that was broke, everyone was making a living. It just the people that control the money want more, that's all. Their product is just slop as far as they see it, and they think people will eat whatever they are given.

 

I don't see that as simply a threat - I see it as an inevitabilty.