Seems as it would be easier to slip in some anti-training, and have the AIs screw systems up so badly that there is a 'recall' of all the current models. The LLMs and their corresponding systems crawl the web constantly. So, poison the well. Good data behind paywalls and credentialing and the poison pill open and free. Seems like it'd be worth a try anyway.

▲

cschep 4 days ago | parent | next [-]

Is this the equivalent to the humans nuking the sky to fight the robots in the Matrix? I don't think that worked.

▲

bloqs 3 days ago | parent | next [-]

Do you keep a copy of the Matrix running on a second monitor at work to aid with decisionmaking

▲

righthand 4 days ago | parent | prev | next [-]

I don’t think our basis for what works and what doesn’t should stem from fiction.

	▲	lazide 3 days ago \| parent [-]
		Especially since even in fiction, that entire backstory had clearly been manipulated/outright made up by the clank.. uh. Machines.

▲

K0balt 4 days ago | parent | prev | next [-]

I wonder about the possibility that AI “clankers” and slop are being weaponised to attack the open internet to push human “data generators” into walled gardens where they can be properly farmed?

I mean, from an incentive and capability matrix, it seems probable if not inevitable.

▲

cschep 3 days ago | parent | prev [-]

to the replies that we shouldn't use fiction to aid decision making -- yes of course! how rational. how reasonable.

.. but perhaps can we access deep wisdom by paying attention the recurring themes of myths?

.. and perhaps does "The Matrix" access any of these themes?

(yes and yes!)

▲

toofy 3 days ago | parent | prev | next [-]

this sounds dangerous in our current situation.

consider how many in our current administration are entirely completely ill-equipped for their positions. many of them almost certainly rely on llms for even basic shit.

considering how many of these people try to make up for their … inexperience by asking a chatbot to make even basic decisions, poisoning the well would almost certainly cause very real very serious national or even international consequences.

i mean if we had people who were actually equipped for their jobs, it could be hilarious to do. they wouldn’t be nearly as likely to fall for entirely wrong absurd answers. but in our current reality it could actually lead to a nightmare.

i mean that genuinely. many many many people in this current government would -in actuality- fall for the wildest simplest dumbest information poisoning and that terrifies me.

“yes, glue on your pizza will stop the cheese from sliding off” only with actual real consequences.

▲

levzzz 4 days ago | parent | prev [-]

[dead]