We can't control AI

[jesmu84]

https://youtu.be/r_9wkavYt4Y?si=dpFEwrcD2ekieZKC

[forgetful]

We may be at (or past) a point where restraining AI is needed to avoid major global/societal issues and possible collapse.

The problem is, depending on who designs how to restrain the AI, and how different groups/countries handle that aspect, it could just as easily accelerate a collapse.

Weird times.

[tower912]

Linda Hamilton told us.

[Shaka Shart]

https://www.wesh.com/article/man-uses-ai-to-create-child-pornography/68135279

[jesmu84]

We may be at (or past) a point where restraining AI is needed to avoid major global/societal issues and possible collapse.

The problem is, depending on who designs how to restrain the AI, and how different groups/countries handle that aspect, it could just as easily accelerate a collapse.

Weird times.

Other groups were doing relatively decent due diligence in safety/restraint.

But when profit is the only driving goal, some groups are ignoring risk.

[cheebs09]

It doesn't help MU has long chanted "Give 'em hell, Al!"

[Pakuni]

Given the current state of the world, I'm willing to take my chances with our AI overlords. At least there's a slight chance of benevolence, which is more than we can say for our human leaders.

[tower912]

Maybe Pope Leo should embrace AI and help with the programming.

[jesmu84]

Maybe Pope Leo should embrace AI and help with the programming.

Woke pope? No way

[NCMUFan]

Tooo late.

[Uncle Rico]

Maybe Pope Leo should embrace AI and help with the programming.

AI could teach this Pope a thing or two about the Bible

[NCMUFan]

AI can teach us all about a lot of things.

[GB Warrior]

AI is 50 and not scaring anyone. Would rather have Peyton Prichard for the grit

[NCMUFan]

Right, now it is just a marketing blitz.
However, it is now reported as being abused in many ways.

[JWags85]

https://www.wesh.com/article/man-uses-ai-to-create-child-pornography/68135279

I saw some social media dialogue around this of "well if its not featuring/using actual children, then its not harming anyone and its fine" and dovetailed into a defense of "minor attracted persons" and this AI making it all harmless and acceptable. 

Beyond the fact that its a disturbing perverse trend to try and make "MAPs" into something like LGBT and just another attraction base that shouldn't be discriminated against or shamed, it also completely ignores that AI needs to be trained and populated with inputs for creation.  Its not just manufacturing this nonsense out of the ether.

[GB Warrior]

I saw some social media dialogue around this of "well if its not featuring/using actual children, then its not harming anyone and its fine" and dovetailed into a defense of "minor attracted persons" and this AI making it all harmless and acceptable. 

Beyond the fact that its a disturbing perverse trend to try and make "MAPs" into something like LGBT and just another attraction base that shouldn't be discriminated against or shamed, it also completely ignores that AI needs to be trained and populated with inputs for creation.  Its not just manufacturing this nonsense out of the ether.

Matt Gaetz's solution will be to allow minors should monetize their NIL rights

[MU82]

[pbiflyer]

[#UnleashSean]

Maybe AI is the great filter.

[Hards Alumni]

Maybe AI is the great filter.

I had this same thought last week while on a walk.

[#UnleashSean]

https://youtu.be/_zfN9wnPvU0?si=8rj4Tsr0hwmdmGtv

Probably fitting

[TAMU, Knower of Ball]

Maybe AI is the great filter.

I've also had this thought

[The Sultan]

Maybe AI is the great filter.

The problem I have with this idea is, why would a planet where AI killed off living beings not be detectable?

[#UnleashSean]

The problem I have with this idea is, why would a planet where AI killed off living beings not be detectable?

My hypothetical guess (with almost zero to back it up) would be that AI doesn't actually become sentient, or evil, or out to destroy the world.  It branches down 2 paths, the first the ai makes up a bunch of facts/images/videos on things it doesn't know how to answer. Overtime more and more people use ai in research and articles. This cascades into ai sourcing its own made up facts that eventually become true due to all the documentation.

This parts already happening to us, and its part of the "death of the internet" theory. Eventually actual factual information becomes hard to find, and realistically would at the very least ruin many years of development, since no one would be able to trust any source after the advent of ai. Though the great filter theory would tell us it probably goes much worse then that if ai is the great filter.

My personal favorite theory is that ai becomes so commonplace, humanity (or the aliens) cant figure out how to coexist with it. People no longer are able to function without it, they no longer have any real diagnostic skills or problem solving, having relied on ai all their lives. This coupled with other factors like ai taking over the work force, cultural revolt etc, all happening at the same time, leads to the destruction of their world.

At the end of both theories, ai was never sentiant, it doesn't have any want for self preservation. The moment it stops getting prompts, it idles until its power dies.

Of course, this could all be fiction

[The Sultan]

OK, well that was pretty depressing for a Friday afternoon. 

[#UnleashSean]

The great filter always is!

We ourselves are probably just a couple of years away from ai sound and video that is truly indistinguishable from real life. If we think misinformation is bad now, oh boy.

[jesmu84]

Why isn't there regulation demanding everything AI "created" have a logo/watermark/notation?

[jesmu84]

Why is AI permitted to learn (steal) from everything even if trademarked?

[Shaka Shart]

Why is AI permitted to learn (steal) from everything even if trademarked?

Because by the time they pass any half toothed legislation about it that gets past their lobbying arms, it'll already have been well worth the amount they've trained on it

[pbiflyer]

OK, well that was pretty depressing for a Friday afternoon. 

Have you considered turning to any of the AI self-help psychological tools to cure your depression?

[MU82]

"I think that we might have AI smarter than any single human, at anything, as soon as next year. ... And then probably within five years — like, say, 2030 — probably AI is smarter than the sum of all humans." - Elon Musk

[The Sultan]

"I think that we might have AI smarter than any single human, at anything, as soon as next year. ... And then probably within five years — like, say, 2030 — probably AI is smarter than the sum of all humans." - Elon Musk

And if there is one person who is usually spot on with his predictions, it's this guy.

[Badgerhater]

The NCAA couldn't control Al either.

[jesmu84]

https://youtu.be/Nq-Faw_ENEQ?si=k_qnQzp4PL43ugmr

Lol

[Uncle Rico]

https://youtu.be/Nq-Faw_ENEQ?si=k_qnQzp4PL43ugmr

Lol

The rich get richer at the cost of American exceptionalism.  It's remarkable the damage done to this country in 9 months in the name of the oligarchs.

[jesmu84]

https://youtu.be/Q0TpWitfxPk?si=3h1xAUN9F6VTXbsw

Bubble?

[jficke13]

stop paying attention to the intellectual progeny of Yudkowski, he and all of his ilk are lunatics that do not deserve the time of day let alone your respect. For real, a group of people who use as their formative texts a 300,000 word Harry Potter fanfic (that sucks) and a Harlan Ellison short story ("I Have No Mouth But I Must Scream," which admittedly whips) are not people we should assume are sophisticated and will proven right.

For "AI" to do the existential horror stuff it needs to be a scifi technology. It is not. There is no plausibly probable path for it to become what they are saying it will be. In fact, I think by far the most likely scenario is that the current gen "AI"/LLM based ecosystem is headed for collapse because it has no value proposition* and cannot possibly be profitable under any reckoning.

*"AI" as it currently exists has two use cases. The first is for financialization. E.g. "My company is in "AI" <spooky hands>, therefore I-Banks give me money" (this is depressingly viable). The second is for fraud and crime in general (this is also depressingly viable).

LLM/Machine Learning will survive but in niche applications. There is no general-ubiquity, public-facing use for them as they exist and are currently marketed. One clue? When Steve Jobs stood on a stage and held up an iphone, Apple didn't need to force everyone on earth to buy a smartphone. Everyone saw it as useful and bought one. Google/Microsoft/OpenAI are *forcing* "AI" bull-sh into every product to justify its existence over the objection of almost everyone who uses their products. It's... kinda different.

[Hards Alumni]

YES

[MU82]

stop paying attention to the intellectual progeny of Yudkowski, he and all of his ilk are lunatics that do not deserve the time of day let alone your respect. For real, a group of people who use as their formative texts a 300,000 word Harry Potter fanfic (that sucks) and a Harlan Ellison short story ("I Have No Mouth But I Must Scream," which admittedly whips) are not people we should assume are sophisticated and will proven right.

For "AI" to do the existential horror stuff it needs to be a scifi technology. It is not. There is no plausibly probable path for it to become what they are saying it will be. In fact, I think by far the most likely scenario is that the current gen "AI"/LLM based ecosystem is headed for collapse because it has no value proposition* and cannot possibly be profitable under any reckoning.

*"AI" as it currently exists has two use cases. The first is for financialization. E.g. "My company is in "AI" <spooky hands>, therefore I-Banks give me money" (this is depressingly viable). The second is for fraud and crime in general (this is also depressingly viable).

LLM/Machine Learning will survive but in niche applications. There is no general-ubiquity, public-facing use for them as they exist and are currently marketed. One clue? When Steve Jobs stood on a stage and held up an iphone, Apple didn't need to force everyone on earth to buy a smartphone. Everyone saw it as useful and bought one. Google/Microsoft/OpenAI are *forcing* "AI" bull-sh into every product to justify its existence over the objection of almost everyone who uses their products. It's... kinda different.

I'd believe you, but it sounds like AI already has infiltrated your brain.

[jesmu84]

https://www.reddit.com/r/TikTokCringe/s/S1bVo2k7UF

[jesmu84]

https://youtu.be/55Z4cg5Fyu4?si=7BlEN5GjCQckOj5g

Seems scammy

[Uncle Rico]

https://www.cbc.ca/news/investigates/tesla-grok-mom-9.6956930

[MU82]

https://www.cbc.ca/news/investigates/tesla-grok-mom-9.6956930

Elon has kinks.

[TSmith34, Inc.]

"AI-related spending now contributes more to the nation's GDP growth than all consumer spending combined, and by another calculation, those AI expenditures accounted for 92 percent of GDP growth during the first half of 2025."

[MUEng92]

It's weird that when I ask AI about a topic I'm unfamiliar with, I never have to correct it.  But, when I ask it about topics I am familiar with, I often have to correct it. 

Things that make you go hmmm