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 talk about. The chatbot ChatGPT, which answers questions like a human, has been the talk of the nation for 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 declined to speak about current business transactions of the company, but he did take a little warning shot when we asked her a related question during the sit-down.
However, he did share 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 might take some time.
Altman clarified that OpenAI’s ongoing partnership with Microsoft — who first invested in OpenA — was in the midst of an evolving partnership in 2019 And earlier today 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 can make a counter move here. They’re smart and competent.”
Altman will only confirm that ChatGPT’s fourth version of GPT will be released by OpenAI when it is ready. [release] You can do it safely and responsibly. He also GPT-4 expectations were tempered when OpenAI stated that they don’t have an “actual AGI” meaning artificial general intelligence or a technology with its very own emergent intelligence. OpenAI’s deep learning models are currently solving problems and identifying patterns by trial and error.
“I think [AGI] It is kind of what is expected from us”, he stated, and GPT-4 “going to disappoint” those with this expectation.
Întreg, inquire about When Altman believes that artificial general intelligence will be realized. He said that it is closer than anyone might think, but that the transition to “AGI”, may not be as sudden as many believe. “The closer we are, the better.” [to AGI]He said that it was difficult for him to answer because he believes that the transition will be more gradual and blurrier than people think.
Naturally, before wrapping up, we had to talk about safety. This included whether society has enough safeguards for OpenAI’s technology. (Plenty are the critics) we do not. Google is rumored to be reluctant to release LaMDA, its AI chatbot, due to concerns about its “ability”reputational risk.”)
Altman stated that OpenAI has an internal process in which they try to find out the causes and effects of problems. External auditors are used. External auditors are used. red teamers. We collaborate with other labs, and safety organizations review stuff.”
He also suggested that the tech is coming from OpenAI and other sources, and people should start to learn how to deal with it. ChatGPT is causing or facilitating a number of social 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] Now [makes sense]It is better to focus on the areas where the stakes remain low than just putting out what the entire industry will have in a few decades with no time for society’s updating.”
However, educators and parents need to understand that the genie cannot be put back in the bottle. Altman indicated that OpenAI, along with other AI outfits, “will experiment” using watermarking technology and other verification methods to determine whether students are trying misappropriate AI-generated copy. However he cautioned that it is not wise to focus too much on this specific scenario. 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 added that it won’t be the only time people have succeeded in adapting to major shifts. Altman pointed out calculators that “changed how we test for math in math classes” as well as Google, which made the need to remember facts much less important. Altman noted that deep-learning models were “an extreme version” of both. He argued that the “benefits” are also more extreme. 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 discuss the best- and worst-case scenarios in relation to AI’s promise and dangers.
What’s the quick version? He’d stated, “The positive case is so unbelievable good that it makes you sound insane to talk about it.” “And the bad case, and I believe this is important to mention — is like lights out for us all.”