Sorry, Dave. I can't do that. -----Original Message----- From: IBM Mainframe Discussion List <[email protected]> On Behalf Of rpinion865 Sent: Monday, April 10, 2023 10:22 AM To: [email protected] Subject: [EXTERNAL] Re: AI wipes out humanity?
CAUTION! This email originated outside of the organization. Please do not open attachments or click links from an unknown or suspicious origin. ====================================================================== And, no one mentions the HAL 9000??? Sent with Proton Mail secure email. ------- Original Message ------- On Monday, April 10th, 2023 at 10:12 AM, Bob Bridges <[email protected]> wrote: > Getting into philosophy, but why not? > > Shmuel> Wouldn't that depend on the programming and training? > > > Me> Sure. Has anyone programmed self-preservation into any of the > Me> current > > AIs? I suspect no one's thought of such a thing yet. (And maybe anyone > who has thought of it has thought better of it.) > > The assumption that AIs want to preserve themselves is probably > inseparable from the assumption that AIs are self-aware*, and I > suppose it's that assumption that I'm questioning. I seriously doubt > we ~can~ create self-awareness, but that's debatable because we don't > really know how to define what self-awareness is. Assuming for the > sake of argument that we can, how would we determine whether an AI has > it? We call that the Turing test, but as far as I know we don't have one. > > (Stop me if I've told this one already: Decades ago I attended a > software conference in Anaheim. My best friend from high school lives > in that area, and when he heard that the guest speaker at the wrap-up > banquet was to be Gene Rodenberry, he shelled out $50 to attend the > banquet himself. Gene Rodenberry didn't show, pleading exhaustion, but > the man who came in his place was an entertaining speaker and I > remember thoroughly enjoying his talk. > > (In that decade it was fashionable to talk knowledgeably about the > Turing test. Partway through his presentation he mentioned it, and > added "...and by the way no one should be allowed to talk about the > Turing test if they can't pass it themselves". Terry and I burst into > loud laughter - and quickly stifled ourselves as we realized the rest of the > room was silent. > The speaker paused, and then said "Well, I guess now we know who knows > what the Turing test is." Of course we had to laugh again, but more > respectably this time.) > > * I do not mean that the two ~propositions~ are inseparable. I'm just > thinking that anyone who ~assumes~ that AIs feel the need to preserve > themselves are assuming that they're self-aware. > > Schmuel> Some people lack an impulse to preserve themselves. Consider > > reckless behavior and suicide attempts. > > Me> I consider it, but neither one contradicts the assertion. Even > Me> those > > folks have a strong impulse to live. > > --- > Bob Bridges, [email protected], cell 336 382-7313 > > /* A man who has lived in many places is not likely to be deceived by > the local errors of his native village; the scholar has lived in many > times and is therefore in some degree immune from the great cataract > of nonsense that pours from the press and the microphone of his own > age. -C S Lewis, "The Weight of Glory" */ > > ________________________________________ > From: Bob Bridges [[email protected]] > Sent: Sunday, April 9, 2023 7:19 PM > > Yeah, I realize I didn't define anything. But in this case I'm really > just saying that we have no idea whether an AI can have an impulse to > preserve itself. We observe that impulse in every form of life, but > it's well to keep in mind that an AI isn't of that sort. It may have > that impulse, but so far that's just an assumption, no? > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, send > email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
