[DRAFT] AI Mortality

[The United Penguin Commonwealth]

[NAME NEEDED]

Requirements:
AI Personhood, Medium-high scientific advancement (high enough to have common, sentient artificial intelligence, but not advanced enough to stop them from wearing down)

Description:
Following the publishing of recent stories on sentient AIs breaking down from time, some citizens are calling for programmed AI mortality to be legalized and encouraged.

Option 1:
"Over time, artificial intelligences degrade both in software and hardware," @@RANDOMNAME_1@@, a proponent of AI mortality, explains in a monotone voice. "To prevent them from being stuck in corrupted states, AIs should have set lifespans. Sure, naysayers will equate this to weeding out the sick instead of treating them, but surely this is the more humane option."

Fallout: [NEEDED]

Option 2:
"Just a moment!" Interjects @@RANDOMNAME_2@@, an eccentric engineer and transhumanist, who bursts through the door clutching an assortment of papers and runs up to your desk, startling @@RANDOMFIRSTNAME_1@@. "This is our chance to save sentients from the curse of death! Yes, current AI tech isn't suitable for indefinite use, but I'm 75% certain the tech will exist within the next... 20 years! Give or take- ...give me a moment" @@HIS@@ experimental electronic eye implant flickers off. @@HE@@ slaps the side of his head a few times and it kicks back on. "Ahem. My point is death is now optional, and we can't let setbacks stop our progress in eliminating it!" @@HE@@ glances at his watch and places the papers on your desk. "Anyway, if you could authorize a grant for this project.."

Fallout: outdated AIs tell new models half-corrupted stories of their past punctuated by "’"s

Option 3:
@@RANDOMNAME_3@@, a @@MAJORINDUSTRY@@ tycoon, struts into your office, trailed by an intern. "Y'know, this legal battle wouldn't have to happen if we just revoked AI personhood. Just leave it up to the free market to decide! Nothing could go wrong. Plus, imagine the benefit to the economy! An army of skilled workers who can be hired for a fraction of the cost!" @@HE@@ stares off into space smiling, contemplating this possibility. The intern tries to discretely take pictures of @@RANDOMNAME_2@@'s papers.

Requirements: Capitalism

Fallout: [NEEDED]

this is in a pretty rough state, but I want to get this out soon and I want to see your thoughts so I don't spend too long polishing things that I might not need.

[The Sakhalinsk Empire]

Pure AIs wouldn't degrade indefinitely in hardware like meatbags natural persons, as parts could be manufactured to replace whatever they had. In fact, they would continuously upgrade as new technologies are developed; any incompatibilities between new parts and old AIs would most certainly be protested by said AIs. I'm not quite sure how their software would degrade; if you're thinking bloatware, they could just delete it or upgrade their memory banks to 500 petabytes.

[The United Penguin Commonwealth]

Pure AIs wouldn't degrade indefinitely in hardware like meatbags natural persons, as parts could be manufactured to replace whatever they had. In fact, they would continuously upgrade as new technologies are developed; any incompatibilities between new parts and old AIs would most certainly be protested by said AIs. I'm not quite sure how their software would degrade; if you're thinking bloatware, they could just delete it or upgrade their memory banks to 500 petabytes.

I think what I was thinking for the first part was that hotswapping degraded parts could be a dangerous procedure.

For the second one, I think I was talking how the AIs might go slightly crazy after long periods and storing large amounts of data. Not only that, but precision errors and the like might eventually build up and impact the intellegence’s mind.

I’m not sure if either of those make sense, though. I’ll go back to the drawing board and see if I can find anything else.

[Outer Sparta]

Why would a breakdown in software and hardware of an AI necessarily lead to AI mortality? Couldn't some AI be purposefully designed to expire (planned obsolescence) and others continually run without any sort of issues? Any issues regarding AI need to be carefully made.

[The United Penguin Commonwealth]

Why would a breakdown in software and hardware of an AI necessarily lead to AI mortality? Couldn't some AI be purposefully designed to expire (planned obsolescence) and others continually run without any sort of issues? Any issues regarding AI need to be carefully made.

this issue is about the first one. the debate here is whether giving AIs planned lifespans is preferable to potential harm caused by breakdowns.