From legal challenges against OpenAI, the creators of ChatGPT, to rethinking the types of jobs that computer programmers will (or won’t) be asked to do in the next five years, to AI-generated movie trailers, AI technology and issues continues to be filed. Key questions about the brave new world centered around conversational AI.
Stability AI founder Emad Mostak makes wild predictions about computer programmers, but this week he made a comment at a journalism roundtable hosted by the International Center for Journalists in mid-June: It’s about ‘how to use AI technology and tools rather than . ”
“We have a 521-million-year-old technology called the human brain, and to optimize care, compassion, deep listening, fully embodied information gathering, co-creation, dissemination, etc. investment,” said Jennifer. Blundell of Hearken, a company that advises startups.
“We humans still have a competitive advantage when it comes to one aspect of caring for AI. And people doing journalism need to make up for what’s been lost.
Other notable efforts in AI include:
Goodbye Programmer?
The buzz around AI technology around the world is due to a group of people, including many computer programmers, coming together to advance genAI technology. Well, the founder of one of his most notable AI companies, the developer of the popular Stable Diffusion text-to-image generator, Stability AI’s Emad Mostaque, said, “In five years there will be no programmers.” I predict. Mostak, who was summoned by Forbes in June for making misleading claims, said in an interview with tech luminary Peter Diamandis that much of a programmer’s job could already be done in an AI engine. Told. Below is a one-minute exchange from a 31-minute interview Diamandis posted earlier this week.
Mostake: No programmers in 5 years.
Diamandis: oh.
Mostake: I think we should always look at what is constant and what is inevitable. So inevitably, 41% of all code on GitHub today is AI-generated. ChatGPT can pass the Google Level 3 Programmer Exam and can almost run on MacBooks and mobile phones.
Diamandis: And is it this year?
Mostake: this year. right now.
Diamandis said, “For those of you who have children who are taking Python lessons and such, perhaps the important thing is helping them understand how to ask great questions and give great direction and prompts.” It may be.”
When it comes to children and the next generation,
While the potential use and misuse of AI tools by students could lead to a flurry of similar-sounding social studies reports on constitutional law, teachers and educators using chatbots like ChatGPT are We are exploring positive ways in which this technology can transform the classroom. That’s why this CNET analysis, “AI’s Teachable Moment: How ChatGPT Is Transforming The Classroom,” is worth reading.
So does the US Department of Education’s 71-page report on AI and the Future of Teaching and Learning, released May 24 and released here.
Why the focus on AI and education now? The ministry cited three reasons:
First, AI has the potential to help us achieve our educational priorities in a better way, at scale and at a lower cost.. Addressing student learning gaps due to the pandemic is a policy priority, and AI has the potential to improve the adaptability of learning resources to student strengths and needs. Improving the teaching profession is a top priority, and through automated assistants and other tools, AI has the potential to provide better support for teachers.
Second, urgency and importance arise through awareness of system-level risks and anxiety about potential future risks. For example, students may be subject to stricter scrutiny. Some teachers fear they will be replaced, but instead, the ministry categorically rejects the idea that AI could replace teachers.
Third, urgency arises because unintended or unexpected consequences are likely to occur. As AI automates instructional decisions at scale, educators may discover undesirable consequences. In a simple example, AI adapted by speeding up the curriculum for some students and slowing others down (based on incomplete data, poor theory, or biased assumptions about learning). In that case, the gap in achievement may widen.
Read the full report here (PDF).
Isn’t OpenAI so novel?
OpenAI, the company behind ChatGPT, has been hit with a class action lawsuit in California, alleging it violated copyright and the privacy of an unspecified number of people when it “used data collected from the internet to train its technology.” reported the Washington Post. report.
“This lawsuit seeks to test a new legal theory that OpenAI violated the rights of millions of internet users when it used social media comments, blog posts, Wikipedia articles and family recipes. ‘ said the paper. (For those who haven’t seen it yet, the Post reported in April that he’s scraping a “secret list of websites” that AI engines are scraping to build large-scale language models. posted an article calling
This is not the only legal challenge facing OpenAI. Two creators are also suing OpenAI for infringing on their copyrights to “train” ChatGPT, Reuters reported. Authors Paul Tremblay and Mona Awad, authors of books such as The Hut at the End of the World, said in the lawsuit that OpenAI’s training data contains a “shadow library” that shares copyrighted material without permission. It states that more than 300,000 books were incorporated, including the use of the website.
In non-legal news, monthly traffic to ChatGPT’s website dropped 9.7% in June compared to May, according to data compiled by analytics firm Similarweb, and unique visitors also fell, according to a blog post that said: The “novelty” is fading and users are looking to other tools such as Google’s Bard and Microsoft’s Bing.
Still, ChatGPT is “the fastest growing consumer application ever, currently boasting over 1.5 billion monthly visits,” said Reuters, and one of the top 20 websites in the world. and And Similarweb said, “While Google is in no danger of being overshadowed by the cultural phenomenon of its OpenAI tech demo site, ChatGPT still outstrips Microsoft’s search engine Bing.com and second-most-popular Character.AI. It attracts many visitors from all over the world,” he added. A popular standalone AI chatbot site. ”
Google powers up with Gemini
Seven years ago, Google’s DeepMind AI lab captured the world’s attention by creating an AI program called AlphaGo that defeated a human champion of the strategy board game Go. Wired reports that DeepMind is working on a new large-scale language model called Gemini, similar to the GPT-4 that underpins ChatGPT, but claims it has significantly better features than its rivals. . (Yes, I know. Competitors always say they do more than they do.) What’s more than that?Demis Hassabis, CEO of DeepMind, on Planning and Problem Solving In his Wired interview, Mr.
“Broadly speaking, Gemini can be thought of as combining some of the strengths of AlphaGo-type systems with the amazing linguistic capabilities of large models,” Hassabis said.
AlphaGo was “based on a technique pioneered by DeepMind called reinforcement learning,” Wired wrote. The technology learns by repeatedly trying difficult problems, such as Go or a video game, where you have to choose which action to take, and receiving feedback on your performance. A method called tree search was also used to explore and remember possible moves on the board. ”
Is it really better than ChatGPT, which cost over $100 million to develop OpenAI? But what we do know is that the Gemini will also be expensive, with Hassabis saying it would cost tens or hundreds of millions of dollars to create.
see who is talking
CNET Audio Producer Stephen Beacham has put together a 5-minute tutorial showing you how to try out an AI voice generator and clone your own voice. He used his AI text-to-speech tool from Eleven Labs. We highly recommend that you read it through to the end. As Beecham discovered, it is “creepy and cool.” And he suggests a surprising development.
Summer means movie trailers, fake trailers
Back in May, the internet was abuzz with a minute-long AI-generated trailer for “Star Wars,” if the iconic sci-fi movie was directed by Wes Anderson. As Anderson’s fans attest, Star Wars: Galactic Menagerie is a hit and hilarious piece of work that has racked up nearly 3 million views for him in the past two months.
But this is an AI promoting several notable movies, including Anderson-inspired trailers for The Lord of the Rings, Harry Potter, The Matrix, and The Godfather. It’s not just a generating trailer. Yes, they’re all short, clever, and worth watching.
want more? There are also fake movie trailers such as the WWII movie narrated by a Michael Caine clone and the apocalyptic sci-fi story “The Great Catspy” (not a typo).
These trailers may be fun to watch, but artists, filmmakers, and others concerned that AI will take away new opportunities for humans and actors (faces and voices) to work on creative projects. It also highlights the very real concerns of the creators. ) being recruited without permission is a major point of contention during the current SAG-AFTRA labor negotiations with Hollywood. According to Vanity Fair, some actors are selling the rights to their voices, and 92-year-old James Earl Jones has said he’s allowed an AI to clone the voice he gave Darth Vader last year. , agreed to allow the character of Darth Vader to continue to live in the Star Wars series, but other actors, including those who voice the characters in the game, had their voices altered by AI. He describes the work as being “on the cutting board.”
Editor’s Note: CNET uses an AI engine to create some stories. See this post for more details.
