Sam Altman, CEO of Openai, was present at the Morgan Stanley Technology Conference in San Francisco a few days ago. At the conference, he talked about a variety of topics such as the impact of artificial intelligence on reducing inflation.
According to Business Insider, Sam Altman spoke at the Morgan Stanley Technology Conference on various topics of artificial intelligence. Of course, this was not a public event and the media was not there. Morgan Stanley, of course, has published some of his talk.
Sam Altman is managing one of the most valuable technology startups in the world. Even Altman may be considered the second most famous technology manager after Ilan Musk; That is why it is important to know Altman’s talk about various topics, especially when it comes to a semi -private session such as Morgan Stanley’s Morgan Technology Conference.
Sam Altman’s talk about negative inflation of artificial intelligence
According to Morgan Stanley, Sam Altman spoke at a recent technology conference on how to reduce inflation by artificial intelligence. He said this is one of the effects of the technology that investors have not paid much attention to or misunderstood. He said, given the potential of artificial intelligence to increase efficiency and productivity, the technology will help compensate for inflation.
Similar to the Altman poison says artificial intelligence itself also had negative inflation. According to him, the cost of access and use of productive artificial intelligence models has been reduced. This is partly due to new techniques that make it easier to create more advanced models of artificial intelligence. As expected, the development of thousands of excellent models and supply rules will make them cheaper.
According to him, this is probably good news for developers and jobs that need access to these models of artificial intelligence; Because they pay much less than last year.
Sam Altman also mentioned the lack of graphics processors. According to him, the Openai GPU units are completely saturated. He says that this restriction is true both in the training phase, where the models of artificial intelligence are created, and in the inference stage that these models are implemented.
On the one hand, Openai is not concerned about providing data to teach its models. Altman points to the company’s ability to use graphic processors and existing artificial intelligence models to create more data, known as “artificial data”.
RCO NEWS