Satya Nadella, the CEO of Microsoft, expressed his views on the dangers of extremely advanced artificial intelligence models and how we should confront these risks in a longer X post. #mesterségesintelligencia #biztonság #Nadella (translated)
Something that cannot be monitored cannot be trusted. A person must always be able to stop or shut down an AI model in the middle of a task, wrote Nadella in a post on the online platform X. Microsoft CEO Satya Nadella has called for a kind of "emergency brake" in artificial intelligence out of concern for a loss of control. The difference from previously used software is that with today's artificial intelligence, one can no longer attribute behavior and results to specific inputs, training data, or configurations, wrote Nadella. #KünstlicheIntelligenz #Microsoft #Notbremse (translated)
Something that cannot be monitored cannot be trusted. A person must always be able to stop or shut down an AI model in the middle of a task, Nadella wrote in a post on the online platform X. #KünstlicheIntelligenz #Microsoft #Technologiekontrolle (translated)
"Something that cannot be monitored cannot be trusted." Microsoft CEO Satya Nadella has called for a kind of "emergency brake" in artificial intelligence (AI) due to concerns about a loss of control. #KünstlicheIntelligenz #Sicherheit #Microsoft (translated)
Something that cannot be monitored cannot be trusted. Microsoft CEO Satya Nadella has called for a sort of "emergency brake" in artificial intelligence out of concern for a loss of control. #KünstlicheIntelligenz #Microsoft #SatyaNadella (translated)
Nadella called his award on X an "incredible honor." #Microsoft #Trump #Visaprogramm (translated)
Microsoft CEO Nadella said that PCs need to become "the safest place" for AI agents. #KünstlicheIntelligenz #Microsoft #Nvidia (translated)
Si hubiera sabido que OpenAI había scrapeado y entrenado en información tras muros de pago, hubiéramos invocado nuestro derecho a pedirle que reentrenaran los modelos
I said AI development cannot be controlled by a handful of entities.
Der Start könnte bereits zwischen Juli und September 2026 erfolgen.
En esencia, pagamos por la inteligencia dos veces: primero con dinero y luego con algo aún más valioso: el conocimiento especializado que hay que revelar para que esa inteligencia sea útil. Cuanto mejor queramos que funcione el modelo, ¡más conocimiento especializado tenemos que proporcionarle!
siamo in una situazione nella quale le imprese concedono, senza pensarci, informazioni fondamentali, rinunciando a guadagnarci.
Indem ich Intelligenz nutze, erschaffe ich Intelligenz. Und das, was ich erschaffe, sollte mir gehören.
Deși este necesar ca furnizorii de modele să beneficieze de dreptul la utilizare echitabilă («fair use») pentru a-și antrena modelele pe date publice, mi se pare ironic că, ulterior, impun condiții restrictive privind distilarea
Spun că clienţii soluţiilor de inteligenţă artificială plătesc de două ori. O dată plătesc bani, iar a doua oară plătesc cu ceva şi mai important: propriile informaţii.
Nu este corect ca noi, producătorii de IA, să câştigăm şi bani şi să dobândim şi datele oamenilor, în vreme ce controlăm complet modelele pe care le oferim.
J'ai présenté « le centre de données IA le plus puissant au monde », permettant d'y connecter « des centaines de milliers » de puces et fonctionnant « sans interruption ».
Non basterà cambiare il racconto. Dobbiamo dimostrare con i fatti che le persone mantengono il controllo e hanno opportunità economiche. Ora dobbiamo fare il lavoro più difficile: conquistare la legittimazione sociale
Claim it. Contact us at [email protected]