In short Joshua Browder, CEO of DoNotPay, made headlines for claiming an AI chatbot had to defend a man in an upcoming court hearing, but backed out of the stunt.
Browder runs a consumer rights startup that was originally created to help people more easily challenge parking tickets, and has since expanded with the goal of creating “the world’s first robot lawyer.” He wanted to show that AI could replace expensive human lawyers, using language models to form legal arguments.
Earlier this month, he claimed to have convinced a man to wear headphones during a court case and recite the output of an AI chatbot during a hearing to be held on Zoom. But his behavior caught the attention of prosecutors angered by his reckless antics.
“Hello! Bad news: after receiving threats from State Bar prosecutors, it seems likely that they will put me in jail for 6 months if I go ahead and bring a robot lawyer into a room. physical hearing. DoNotPay is postponing our lawsuit and sticking to consumer rights,” he said. tweeted this week.
Browder has offered $1 million to any lawyer willing to up the ante by testing an AI model to argue a case in the Supreme Court.
Google is building a music-making AI, but won’t release it due to copyright
Google researchers have trained a new AI model, MusicLM, which can create audio samples based on text descriptions, according to a research paper on arXiv.
The first sample, for example, is generated with the prompt: “The main soundtrack of an arcade game. It’s fast and upbeat, with a catchy electric guitar riff. The music is repetitive and easy to remember. , but with unexpected sounds, like crashing cymbals or rolling drums.” You can listen to it here.
It looks like a video game track. Other samples, however, are less convincing. A fake rapper and singer on the hip-hop song utters absolute gibberish. MusicML outputs sound pretty good, but the music is repetitive and dull.
Developers have been working on using AI to generate music for years, and the results so far have been lackluster. MusicLM seems to be better than previous systems, but it’s hard to tell because the selected samples are only 30 seconds long.
Google also won’t be releasing the tool for general use in the near future due to copyright risks. It’s unclear exactly what data the researchers used to train the model, but they noted concerns about “potential misappropriation of creative content.”
There’s always legal debate over whether generative AI models infringe copyright, and the music industry is notoriously litigious, especially if the defendant has the money. It’s probably best if Google keeps MusicLM on mute for now.
ChatGPT isn’t particularly innovative or revolutionary, says Meta’s chief AI scientist
Yann LeCun poured cold water on the recent hype surrounding OpenAI’s language model, ChatGPT, and said the product isn’t that innovative or revolutionary.
ChatGPT generates text based on instructions typed by a user and can help write all sorts of things like essays or code.
“It’s nothing groundbreaking, even though that’s how the public perceives it. It’s just, you know, it’s well put together, it’s well done,” he said, as ZDnet reports. LeCun said the technology that powers ChatGPT dates back decades, and its architecture and formation process were actually developed by researchers working elsewhere, including at Google and Meta.
“It’s not just Google and Meta, but there are half a dozen startups that basically have very similar technology. I don’t want to say it’s not rocket science, but it’s really shared, it there’s no secret behind that, if you will,” he added.
While this may be true, it’s worth noting that unlike Google and Meta, OpenAI has made it available to everyone.
You can now generate AI images on Shutterstock
Stock image biz Shutterstock has partnered with OpenAI to launch a feature allowing artists to generate content using DALL-E 2 directly on its platform.
Users subscribing to its Creative Flow service will have access to a suite of AI-powered tools for creating and editing images. These can then be sold directly on Shutterstock – creators will be compensated for sharing their royalties.
“We are revolutionizing the way visuals are created for campaigns, projects and brands by making generative AI accessible to everyone,” the stock library said in a statement. “Our image generator produces unique, varied, and stunning images from a single word or short simple phrases. And with an intuitive style picker and support for over 20 languages, we empower people in the world to bring their limitless creative visions to life.”
By licensing these images, Shutterstock overcomes thorny issues of copyright and ownership and profits from them. Companies seeking to own or reclaim images to form text-to-image models must pay for the content, and artists consent and agree that their work may be owned or used elsewhere. ®