Mostrando entradas con la etiqueta artificial intelligence. Mostrar todas las entradas
Mostrando entradas con la etiqueta artificial intelligence. Mostrar todas las entradas

jueves, 15 de diciembre de 2022

Omnifarious opportunities or Overlord: OpenAI


Skynet took just 25 days* to become self-aware...

Who knows. Last week Spartacus had a talk with ChatGPT : We've Come a Long Way [LINK]. This is Open AI's Generative Pretrained Transformer (ChatGTP...you'll have to wait now because it got to over a million users in about a day, with almost zero media coverage!). 

Have a read, the chat is very complicated, we're way beyond Delphi and Sophia. 

Despite the reservations and limitations that Spartacus mentions in the conclusions, it sounds mind-blowingly good. 
"Also, with how fast these deep learning neural networks are improving, and their potential for sudden and hugely disruptive innovation across a wide variety of industries, all I can say is, buckle up. Things are about to get very, very weird." 
More at Epoch Times from Hans Mahncke [LINK] who was 'blown away' by it. 

By the way...Artificial Intelligence in fiction has been around for about 150 years! Funnily enough on the Wiki page (yeah, yeah, I know) on the 'Implausibility'..."Engineers and scientists have taken an interest in the way AI is presented in fiction. In films like the 2014 Ex Machina or 2015 Chappie, a single isolated genius becomes the first to successfully build an artificial general intelligence; scientists in the real world deem this to be unlikely."...erm, Open AI is Elon Musk's baby [late edit: Link added]

[Apologies: sign-up or £$€ may be required for some of the links.]

* Very later edit: this was in Terminator 2: Judgment Day