Isaac Asimov - Mirror Image

Isaac Asimov - Mirror Image | Audiobook

Рет қаралды 677

Ай бұрын

I'm reading Isaac Asimov's story "Mirror Image" to the music Cyberpunk 2077 OST - Pacific Dreams Radio 88.9 (Cyberpunk 2077 official soundtrack) in the style of the MDS radio show (Model for assembly).
Robot Daneel Olivaw, a former partner of detective Elijah Bailey, turned to him with a request: to help unravel one complicated case. Two outstanding mathematicians dispute the authorship of one brilliant theory. Each claims that he is right, and the other is a blatant liar. Elijah decides to interrogate the robot servants who witnessed this incident...
A very interesting story that predicts the problems that are facing us in the very near future, given the rapid development of artificial intelligence and neural networks. Nowadays, neural networks are still quite limited, both on their own (by their natural capabilities) and additionally artificially - so that you don’t accidentally force the neural network to tell you how to create a bomb, for example. But the same artificial restrictions can sometimes be circumvented. So, as AI develops in the very near future, it is very likely that pirated neural networks will appear - copies of well-known models without paying for the corresponding subscriptions. It is also possible for “black” neural networks to appear and spread without any artificial restrictions on the same darknet. And how to sort all this out is not very clear yet.
Text: Isaac Asimov
Reader: Eugene Gubanov aka Kivlov
Music: Cyberpunk 2077 OST - Pacific Dreams Radio 88.9
Graphics: AI Brain Tunnel Digital Background Loop from videezy.com
Links:
- Official website: eg-art.pro/
- My VK page: eugene_gubanov
- My Telegram channel: t.me/eugene_gubanov
- Patreon: patreon.com/user?u=33080164
- Boosty: boosty.to/eugene_gubanov/purc...
- DonationAlerts: www.donationalerts.com/r/euge...

Пікірлер: 3

@Eugene.Gubanov Ай бұрын

О рассказе: Робот Дэниэл Оливо, в прошлом напарник детектива Элайджа Бейли, обратился к последнему с просьбой: помочь распутать одно запутанное дельце. Два выдающихся математика оспаривают авторство одной гениальной теории. Каждый утверждает, что прав он, а другой - наглый лжец. Элайдж принимает решение допросить роботов-слуг, которые были свидетелями данного происшествия... Весьма любопытный рассказ, прогнозирующий проблемы, которые нам светят уже в самом ближайщем будущем, учитывая бурное развитие искусственного интеллекта и нейросетей. Сейчас пока ещё нейросети довольно ограничены как сами по себе (по своим естественным возможностям), так и дополнительно искусственно - чтобы вы случайно не заставили нейросеть рассказать вам - как создать бомбу, например. Но те же искусственные ограничения иногда можно обойти. Так что по мере развития ИИ в самом ближайщем будущем весьма вероятно появление пиратских нейросетей - копий известных моделей без оплаты соответствующих подписок. Также возможно появление и распространение "чёрных" нейросетей без каких-либо искуственных ограничений в том же даркнете. И что со всем этим делать - пока не очень понятно.

@vasaneff2253 Ай бұрын

Ну по крайней мере пальцем я "проголосовал" :)

@Eugene.Gubanov Ай бұрын

About the story: Robot Daneel Olivaw, a former partner of detective Elijah Bailey, turned to him with a request: to help unravel one complicated case. Two outstanding mathematicians dispute the authorship of one brilliant theory. Each claims that he is right, and the other is a blatant liar. Elijah decides to interrogate the robot servants who witnessed this incident... A very interesting story that predicts the problems that are facing us in the very near future, given the rapid development of artificial intelligence and neural networks. Nowadays, neural networks are still quite limited, both on their own (by their natural capabilities) and additionally artificially - so that you don’t accidentally force the neural network to tell you how to create a bomb, for example. But the same artificial restrictions can sometimes be circumvented. So, as AI develops in the very near future, it is very likely that pirated neural networks will appear - copies of well-known models without paying for the corresponding subscriptions. It is also possible for “black” neural networks to appear and spread without any artificial restrictions on the same darknet. And how to sort all this out is not very clear yet.