The Microsoft deal is not exclusive. Video is coming soon, and more from OpenAI CEO Sam Altman

OpenAI Sam Altman, co-founder and CEO, sat down with this editor for a broad interview last week. He answered questions about his most important accomplishments. ambitious personal investmentsYou can also find out more about OpenAI’s future.

There was much to be discussed. The chatbot ChatGPT, which answers questions like a human, has become a national talking point in the past two months. It is now an eight-year-old company. OpenAI’s products don’t only amaze users, the company is apparently in talks to manage the sale existing shares to new investors. $29 billion Valuation despite its relatively low value nominal revenue. Meanwhile, worried educators These are becoming more prevalent blocking ChatGPT is blocked to students because of fears that they will cheat.

Altman refused to discuss the company’s business affairs, and he fired a warning shot when we asked him a similar question during our sit down.

Altman did however reveal some details about the company’s future plans. Altman confirmed that ChatGPT will be adding a video model to its popular digital art generator DALL–E. However, he stated that he doesn’t know when it would happen and that it could be quite soon. It’s a legitimate research endeavor. It may take some time.

Altman clarified that OpenAI’s ongoing partnership with Microsoft — who first invested in OpenA — was not a new one. in 2019 Today, and earlier confirmed it plans ChatGPT is an AI tool that can be integrated into all products of the company. This is not an exclusive arrangement.

Altman further confirmed that OpenAI has the ability to create its own software products and services as well as licensing its technology to companies. This is a significant development for industry watchers, who have wondered whether OpenAI will ever be able to compete directly with Google using its own search engine. Altman responded to the question by saying that when someone speaks of a technology as the end for a giant company it’s usually wrong. People forget that they have the ability to make a counter-move here.

Altman will only confirm that ChatGPT’s fourth version of GPT will be released by OpenAI when it is ready. [release] It safely and responsibly. He also GPT-4 was tempered by OpenAI’s current deep learning models, which solve problems and find patterns through trial-and-error.

“I think” [AGI] He said that GPT-4 was “sort of what is expected” of them and that it would “deceive” people who have this expectation.

In the interim, you can ask about When Altman anticipates artificial general intelligence. He believes it will come closer than we might imagine, and that the transition to “AGI”, however, won’t be as abrupt or as rapid as some believe. “The closer it gets [to AGI]”The harder it is to answer, because I think that it will be much blurrier than people think and much more gradual,” he stated.

Naturally, before wrapping up, we had to talk about safety. This included whether society has enough safeguards for OpenAI’s technology. (There are many critics. we do not. Google has been reported to be reluctant about releasing its AI chatbot LaMDA due to concerns over its “ability to communicate with humans.”reputational risk.”)

Altman stated that OpenAI has an internal process in which they try to find out the causes and effects of problems. We have external auditors. We use external auditors. red teamers. We cooperate with other labs to have safety agencies look at things.”

He said that tech is coming, from OpenAI to other places, and people must start to figure out how they can live with it. ChatGPT is causing or facilitating societal changes. ChatGPT’s impact on education, academic integrity and all that is related to it is a big issue right now. He argued that “starting these” is the best way to do it. [product releases] Get it now [makes sense]”, as the stakes are still low. Rather than just putting out what the industry will have within a few years with no chance for society to update,”

Teachers and parents should realize that there is no way to put the genie back in a bottle. Altman stated that OpenAI and other AI companies “will experiment with watermarking technologies, and other verification techniques, to assess whether students are trying pass off AI-generated content as their own. However, he warned that too much attention to this particular scenario could prove futile. While there may be ways to make teachers more likely to detect the output of a GPT-like system than they are, I believe that a determined individual will get around them and it won’t be something society should or can rely on for the long-term.

He said that this is not the first time people have been able to adapt to major changes. Altman cited calculators that “changed the content of math classes” and Google, which made it less important to remember facts. Altman said that deep learning models are “an extreme version” both of these developments. Altman argued that deep learning models have “more extreme” benefits. Teachers are often very worried about the effects of this on homework. Teachers who feel like this are a great personal tutor for their students are also common.

Watch the video below to hear the whole conversation about OpenAI and Altman’s evolving views regarding the commodification, regulation, and why AI is moving in the “opposite direction” many people imagined five to seven year ago.

Altman will also address the worst-case scenario and promise of AI.

Here’s the short version: He said, “The good case just so unbelievably great that you sound crazy to start talking about,” “And the worst case, and this is what I feel is important to state — is lights out for everyone.”

Previous post The Microsoft deal is not exclusive. Video is coming soon, and more from OpenAI CEO Sam Altman
Next post China’s NetEase condemns Blizzard as unjust and unequal