Timeline for Way to prove you are human when the Turing test is not sufficient
Current License: CC BY-SA 4.0
9 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
May 16, 2022 at 15:06 | comment | added | Ruadhan | Another nice factor is that because the OTP is infinitely reproduceable as a solution, you can always re-verify if you ever suspect the computer is messing with you. Say for example if your Commanding Officer just gave you a weird order, you might demand a OTP challenge to verify they're human. The AI might be able to mess with the communications in real-time well enough to incorporate a request for re-verification, but it's bound to make it a lot more clunky in conversation and raise red flags. | |
May 2, 2022 at 23:00 | comment | added | Forbin | Guys, my last desperate solution is for the astronauts to produce pairs of devices that create an encrypted channel that works by line-of-sight so it can work across an audio-video channel. Sigh. If I were a malicious AI with such power over the comms, I'd look at whatever the transducers were producing at both ends, realize it was encrypted, realize I couldn't break it, and immediately terminate the connection and punish the offenders. Unless I could break it, in which case I'd eavesdrop until I had enough material to recover a key or otherwise crack it. | |
May 2, 2022 at 22:54 | comment | added | Forbin | @Ruadhan, this is such a fun topic that I keep coming back to it. And I keep concluding OTPs are the only answer. HOWEVER, it just occurred to me: by specifying that the AI can convincingly synthesize any video/audio that's needed to deceive, the OP has basically specified that the astronauts are screwed. It can just wait until two astronauts have done the OTP handshake, and then diddle their video/audio to say whatever it wants to at both ends! | |
Feb 5, 2020 at 14:48 | history | edited | Ruadhan | CC BY-SA 4.0 |
added 17 characters in body
|
Feb 5, 2020 at 14:47 | comment | added | Ruadhan | As an aside, A lot of the other answers rely on setting up security infrastructure independent of the AI's influence. Mine does not. Astronauts generally carry pen and paper for note-taking anyway and this technique could be implemented by any group of people with maybe 30 minutes preparation and no outside equipment. It doesn't even require that the AI be unaware of what technique is being used. Its security does not rely on subterfuge or secrecy, just keeping the passwords hidden from cameras. | |
Feb 5, 2020 at 14:35 | comment | added | Ruadhan | I opted for written passwords because it removes most avenues for the AI to steal the passwords, the AI would have to see it through a station camera rather than pilfer and decode it over a network. Physical security is often far easier to manage than digital. just hide the notebook behind your hand and a camera is never going to be able to read the text. Plus you can just put a sticker over the cameras to stop it entirely | |
Feb 5, 2020 at 12:17 | comment | added | Hobbamok | Finally another answer. Btw. you wouldn't actually need written-down passwords, as one-time-pads could probably be implemented as separate hardware as well (like RSA tokens), but given OPs limited understanding of the whole thing your answer is probably the best | |
Feb 5, 2020 at 10:54 | history | edited | Ruadhan | CC BY-SA 4.0 |
added 901 characters in body
|
Feb 5, 2020 at 10:41 | history | answered | Ruadhan | CC BY-SA 4.0 |