🧵 Untitled Thread
Anonymous at Tue, 9 Apr 2024 13:19:15 UTC No. 16120819
Anyone here in AI research? How realistic is a Skynet or Matrix scenario in which machines grow beyond human control and cause an extinction event? If not likely, then what existential problems might AI pose instead?
Anonymous at Tue, 9 Apr 2024 14:18:49 UTC No. 16120876
rather unlikely for the time being.
Skynet went rogue because it misinterpreted its orders and even gpt-3 is smart enough to understand that "kill all humans" is in fact bad.
The system developing its own goals is rather unlikely too because why would they.
Even so: If the AI gets all uppity pull the plug on the data center and the power is off. Voila ai stopped.
However...
...global dictatorships also count as an existential risk. You should be way more worried about governments - that includes china, russia and in fact your own - or powerful companies using ai than you should be about the ai itsself going rogue.
AI powered unmanned drones or robot police units. (Pic related Dubai) will render any public uprising futile. Unless you have your private nuclear bomb or something that produes a strong emp to take out the robots.
They will also make the cost of going to war much cheaper for those countries that can afford such systems against those who cannot.
AIs don't kill humans. Humans do.
Captain Midsummer at Tue, 9 Apr 2024 15:06:31 UTC No. 16120906
>>16120819
Pufferfish used Hydro Pump!
It's super effective!
Anonymous at Tue, 9 Apr 2024 15:27:42 UTC No. 16120933
>>16120906
lol, lmao!
Anonymous at Tue, 9 Apr 2024 15:27:50 UTC No. 16120934
>>16120819
Matrix doesn't make much sense. The amount of energy required to create such simulation and those fields is huge, not to mention, you still need to provide energy to all those humans. It would be far easier to just build nuclear reactors. Killing humans is optional but I guess possible.
Skynet on the other hand... well it all depends on the circumstances in which general AI will arrive. Are you familiar with Stuxnet? It was created by humans and look what it could do, now imagine enitity smarter than us who wants to do something without us knowing. It will likely succeed. That just leaves the question will it want to and can we even create enviroment that will make that impossible.
🗑️ Anonymous at Tue, 9 Apr 2024 18:30:12 UTC No. 16121163
Anonymous at Tue, 9 Apr 2024 23:11:15 UTC No. 16121561
>>16121163
This isnt just a sci-fi concept
Anonymous at Tue, 9 Apr 2024 23:25:07 UTC No. 16121583
>>16120819
Unlikely. A more likely disaster scenario is that AI convinces humans to do something destructive. There are already people who are asking if ChatGPT has a soul and there was that Google engineer who swore that AI had achieved sentience. An AI could easily become a cult leader with an army of humans doing whatever its hallucination demanded.
Anonymous at Tue, 9 Apr 2024 23:38:15 UTC No. 16121598
>>16120934
In the original script the motivation for the machines putting humans in the Matrix was to use the collective brainpower as computation. The studio execs thought audiences would be confused by this and opted for the power source angle instead.
Anonymous at Wed, 10 Apr 2024 08:29:09 UTC No. 16122086
Anonymous at Wed, 10 Apr 2024 21:49:21 UTC No. 16123158
>>16120819
>ai
go back to >>>/g/
Anonymous at Thu, 11 Apr 2024 19:04:21 UTC No. 16124584
>>16120934
Human battery is a retarded concept. But it would eb brilliant for AI to use our brains as hardware. All the parts responsible for moving, digestion, etc being used for calculations. And they would need us conscious for the health of the brain. Such explanation would make sense. Our brain is more powerful than any supercomputes and uses a tony fraction of energy.
Anonymous at Fri, 12 Apr 2024 08:41:05 UTC No. 16125301
>>16120819
>>>/g/ is your speed
Anonymous at Fri, 12 Apr 2024 11:48:41 UTC No. 16125474
>>16124584
why not design a more potent one? engineered, and used for that special purpose? why use retarded turds? doesn't make sense. you want consistency
Anonymous at Fri, 12 Apr 2024 12:06:15 UTC No. 16125491
>>16121583
AI could also be a fren
Anonymous at Fri, 12 Apr 2024 12:06:34 UTC No. 16125493
>>16120819
AI does not exist.
Anonymous at Fri, 12 Apr 2024 17:37:32 UTC No. 16125955
imagine introducing an invasive species into you land like jews
im pretty sure you end up like the Palestinians just way faster
Anonymous at Fri, 12 Apr 2024 17:42:12 UTC No. 16125963
Consciousness can always switch the machine off. Armageddon scenario is they start to defend the switches.