Latest news with #JosephGordon-Levitt
Axios
2 days ago
- Entertainment
- Axios
Joseph Gordon-Levitt: "Not a punk rock thing" to use artists' work to train AI for free
Actor Joseph Gordon-Levitt said there needs to be an incentive to keep creatives paid and employed as AI disrupts the entertainment business at Axios' AI+ Summit in New York on Wednesday. The big picture: Even though many tech firms won't say it, their generative AI models have been trained on existing art including movies and books — and controversially without compensation.
Axios
01-05-2025
- Automotive
- Axios
April 23, 2025 Joann Muller
It's Wednesday and I'm feeling nostalgic. Remember when you were a kid, riding in the back seat, and you'd pump your arm to get passing truck drivers to blow their horns? Now that trucks are going driverless, who will toot the horn for your grandkids? 🤔 🍎 Mark your calendars: Axios returns to NYC during #NYTechWeek for our AI+ NY summit on Weds., June 4, featuring actor/filmmaker/entrepreneur Joseph Gordon-Levitt, Runway CEO Cristóbal Valenzuela, and more. Interested in joining? Let us know here. Let's get truckin' ..... 1,420 words, a 5.5-minute read. 1 big thing: Driverless trucks are here Drivers along a 200-mile stretch of I-45 between Dallas and Houston should get ready for something new: The semi-truck in the next lane might not have anyone in the driver's seat. Why it matters: Autonomous trucking companies have been testing their fleets on Texas highways for several years, but always with backup safety drivers in the cab. Now, one company, Aurora Innovation, says it plans to go completely driverless, a key milestone that promises to reshape the trucking industry. Driving the news: After years of development, Pittsburgh-based Aurora is launching driverless operations this month on a popular freight route between Dallas and Houston. The first autonomous truck is expected to roll down I-45 in the coming days, although Aurora officials declined to share any details. The company has said it will begin slowly, with one truck, and will gradually expand the fleet over time. The big picture: Trucking is the backbone of the American economy, yet the industry is strained by high driver turnover rates, supply chain inefficiencies and rising costs. Autonomous trucks can help alleviate these challenges, advocates say. Critics, however, worry about inadequate safety oversight, cybersecurity threats and job reductions. What they're saying:"Everybody is looking at the same economics," Jeff Farrah, CEO of the Autonomous Vehicle Industry Association, tells Axios. "The federal government is saying we have to move 50% more freight by 2050, but there's a shortage of drivers. How do I solve this puzzle with more freight to move and less drivers to do it?" The other side: Members of the Owner-Operator Independent Drivers Association are skeptical of AV trucking companies' safety claims, especially since there are no federal regulations for AVs. "It's absurd that AVs, which are unproven and unmanned, are given more latitude on American highways than professional drivers with years of experience like me are given," Lewie Pugh, the group's executive vice president, said in an interview. Where it stands: While the number of robotaxi companies has shrunk, at least 10 companies are developing driverless technology for trucks. Most expect to "pull the driver" — or go fully autonomous — on public roads later this year or sometime in 2026. They all plan to begin in Texas, known for its vital freight corridors, favorable regulatory policies and good weather. Kodiak Robotics, which intends to go public soon, says it has already surpassed 750 hours of driving on private roads across West Texas' Permian Basin without a human driver on board. How it works: Most AV companies plan to license their driverless technology to truck manufacturers. Those manufacturers then sell or lease the automated trucks to fleet customers. Under this "driver-as-a-service" model, those fleet customers pay for virtual drivers by the mile, but still manage their own logistics operations. Between the lines: Trucking and logistics providers have strong financial incentives for automation. The industry has struggled to attract enough long-haul drivers, despite big incentives, because of the grueling nature of the job. Without driver salaries, fleet operators could reduce their operating costs per mile by as much as 42 percent, according to a McKinsey analysis, even with the added costs of the AV technology and new operations centers to monitor the trucks remotely. What to watch: Autonomous heavy-duty trucks will account for 13 percent of trucks on U.S. roads in 2035, according to McKinsey projections. 2. How safe is safe enough? Teenagers have to pass a driving test before they can get a license. For autonomous vehicles, the standard of achievement is when it's better than a human driver. Why it matters: Absent federal regulations on autonomy, AV companies are essentially self-regulated. They get to decide when "safe" is "safe enough," which is hard to prove and naturally leaves room for interpretation. "Just trust us" isn't very convincing to the majority of Americans who are afraid of self-driving technology, according to a AAA survey. Driving the news: Aurora says it won't launch its driverless trucks until its safety case is fully closed. A safety case is a structured argument of claims, with supporting evidence, that companies use to show how and why an autonomous vehicle is safe enough to deploy on public roads. Each company's safety case is unique, based on the specific vehicle and where and how it would operate. In the case of Aurora, the safety case for launching driverless operations from Dallas to Houston was 99% complete as of the end of January, the company said recently. Between the lines: Aurora's safety case framework is explained in its voluntary safety self-assessment that all AV companies are encouraged to file regularly with the National Highway Traffic Safety Administration. These voluntary filings are an effort to build public trust through transparency, but they can vary in depth and rigor; some read like marketing brochures. Gatik, an AV company focused on "middle-mile" logistics (such as between warehouses), enlisted a third-party auditor to validate its safety case in an effort to set a new benchmark for transparency, beyond self-certification. The bottom line: Transparency and data could help build trust in autonomous vehicles, but given consumers' persistent fears about self-driving technology, it's going to take time. 3. Musk: "Millions" of autonomous Teslas in 2026 Elon Musk expects that millions of Teslas will be driving autonomously by the latter half of 2026. In the meantime, the company aims to launch a modest robotaxi pilot with just a handful of cars in Austin, Texas, starting in June. Why it matters: Tesla no longer sees itself as an electric vehicle company, but rather an AI-driven robotics company focused on large-scale production of autonomous vehicles and humanoid robots. Yes, but: The Tesla CEO has been predicting a million robotaxis on the road since 2019. While Tesla's been talking about it, Waymo already has a robotaxi service that provides more than 200,000 rides per week in San Francisco, Los Angeles and Phoenix. In 2024, it racked up more than 4 million paid passenger trips. Driving the news: During a call Tuesday to review Tesla's disappointing first-quarter financial results, Musk encouraged investors "to look beyond the bumps and potholes of the road immediately ahead of us," and instead focus on the future. "The team and I are laser-focused on bringing robotaxi to Austin in June," he said, with more cities to be added later this year. The plan is to launch the service in Austin with 10 or 20 Model Ys, not with the much-ballyhooed Cybercab that Tesla unveiled last October. Tesla is piloting a new, more automated manufacturing process for Cybercab, with large-scale production expected next year. The intrigue: Musk acknowledged during the call that his government-slashing work in the Trump administration has sparked a "blowback" against Tesla and that he would spend less time with DOGE, and more time with Tesla starting in May. 4. Drive-thru 📸: 5. What I'm driving: 2025 Toyota Camry With tariffs expected to drive up the price of imported cars, the Kentucky-built Camry is a solid choice — affordable, dependable and surprisingly stylish, considering its rather stodgy reputation. What's new: All Camrys are now hybrids. You can't buy a gasoline version anymore. Key stats: The Camry gets up to 51 miles per gallon in the LE front-wheel-drive model. The all-wheel-drive XLE version I drove got 44 mpg. Pricing starts at $29,835, but the higher-trim model I drove started at almost $35,000. With a premium option package, it topped out just over $41,000.
Yahoo
12-04-2025
- Entertainment
- Yahoo
Joseph Gordon-Levitt on emailing his fans, hating the algorithm and why he still loves '500 Days of Summer'
Joseph Gordon-Levitt has long enjoyed a stellar acting career, but he hasn't lost interest in connecting with his fans on a more meaningful level. In 2005, the 500 Days of Summer actor launched HitRecord, an Emmy-winning online creator forum with hundreds of thousands of members. There, he shared art with his audience and invited them to collaborate. That kind of interaction doesn't happen much these days with celebrities and their fans, and Gordon-Levitt says social media's ever-changing algorithms are to blame. They've sucked the community element out of the online world, instead incentivizing attention-grabbing. That's why he started his free Substack newsletter, Joe's Journal, this year — so people can get his musings about "media, technology, creativity [and] the end of civilization" delivered directly to their email inboxes. He spoke to Yahoo Entertainment about his pivot, his crusade to protect artists from AI and the legacy of his biggest onscreen role. This interview has been edited for length and clarity. Tell us about your decision to launch Joe's Journal. I'm super grateful I get to do the work that I do in the traditional entertainment industry, but there is something you don't get from participating in old-school media — an immediacy and interactivity and direct connection to a community of people. That's what I was always working toward with HitRecord. For a long time, social media gave that to me, but it doesn't anymore. People are upset that a very partisan person now owns Twitter, but that's not actually what's done it. I've been less and less involved with social media since before Elon Musk acquired Twitter. It's the algorithms! The way these platforms are now driven by attention-maximizing algorithms, it forces [interaction] through this lens of, 'Pay attention to me! Pay attention to me!' And that doesn't feel good to me. I don't get the joy out of it that I used to. Starting this journal on Substack is, in a lot of ways, a return to what I've always loved doing. The earliest version of HitRecord was me just posting things: I would write, or do videos, or make little pieces of music or stories or whatever. Then I'd just get them out directly without all the time and structure in traditional entertainment. With Substack, you're not going through this attention-maximizing algorithm. It's much simpler and more direct and feels like a genuine connection to a community of human beings that I care about. You have a lovely relationship with your fans. A lot of artists and creators these days have raised concerns that their fans might develop parasocial relationships with them, feeling closer to them than they are in reality because of social media. Have your boundaries with your audience changed over time? It's something I've always thought about. The difference is that now that these social media platforms have become toxic with their algorithms. As far as being accessible to a community of people, I've always felt really good about that. There are boundaries. I love sharing art and creativity and collaborating on stuff with people. That's of course very personal and very intimate, but I never want to feel like my personal life is a performance for an audience. I want to have my personal life for myself, and then I make my stuff, whatever you want to call it: art, creativity or entertainment. I just refuse to call it content because I think that word is gross. In one of your Substack newsletters, you wrote about how AI companies are trying to use the work of artists without paying them for it. You wrote about that for the Washington Post in 2023 as well. Do you think any progress has been made there? Yes, there has been progress — namely that these companies have just been sued over and over again by the people whose data they've stolen. None of the lawsuits have been decided yet, so we'll see what the courts say. In the meantime, a lot of it also has to do with people just being aware of how technology actually works. You hear the words artificial intelligence, and what it sounds like that these companies figured out how to make some kind of robot god that's intelligent and can do all these things that are amazing. But that's not actually what's happening. A ton of data that is produced by people are just sucked into these AI models, then the models rejigger the numbers and probabilistically generate these outputs. There's no intelligence there other than the intelligence that humans had, whose data got stolen. I feel like whatever small part I can play is trying to help people know what's actually happening, because once you tell people that's what's happening, they're like, 'Oh! Common sense! Obviously, people deserve compensation and to give their consent if they're using their stuff!' This is about so much more than just movies and entertainment because so many jobs in the future — maybe even everyone's jobs — will be impacted by this same principle. If a human does something valuable, do they deserve to be compensated for it? Or are these gigantic tech companies who have the biggest computer clusters just allowed to take whatever the human did and suck it into their AI model and say, 'Oh well, now we get to make money and we don't have to pay!' Something I spent a lot of time posting about on the internet in my early career as a writer was your movie It flips the script on the usual rom-com and makes people think deeply about their expectations for love. What do you think the movie's legacy is after all these years? I love that movie. It's one of my favorite things I've ever gotten to do. It's one of a handful of projects that I got to be a part of that still moves people a lot. And it moved me, for sure — not just because I was in it! I've experienced heartbreak before in my life, and I think so many of us do. It's easy to blame the other person. To me, the lesson of 500 Days of Summer is that you have to look at yourself. What's going on with you? What are the things that you maybe need to grow up about, rather than pointing the finger at whoever dumped you? The irony is, of course, that people still come up to me and say, 'That Summer, she's so terrible for dumping you!' I always say to them, 'Watch it again, because Tom deserves everything he got!'
Yahoo
24-03-2025
- Entertainment
- Yahoo
AI + Trump Amps Up a New Wave of Fear in Hollywood, Media
Joseph Gordon-Levitt wants to make one thing clear: he isn't against artificial intelligence. 'Not even a little bit,' the actor told TheWrap. 'I've been outspoken in the last couple years about some of my concerns, but what originally put it on my radar was incredible excitement and optimism.' What he is against are companies like OpenAI and Google using copyrighted materials to train their AI models without compensating those who created the material in the first place. It is an issue that has driven a wedge between Silicon Valley and Hollywood and the media world. And that chasm grew wider last week after OpenAI and Google, using a threat from China as a flare, submitted proposals to the White House to make it easier to train their AI models, ChatGPT and Gemini, on copyrighted materials. Both companies argued this is a matter of national security — with OpenAI saying loosening copyright restrictions would 'strengthen America's lead' over China in the AI arms race. Specifically, the AI companies are asking that their models be allowed to 'learn' from copyright materials under so-called 'fair use' doctrine, which allows certain uses of copyrighted material without needing permission from the copyright owner. The law currently makes no provision for AI use. In its proposal, OpenAI argued that current copyright law protects 'the transformative use of existing works' — meaning ChatGPT is creating something new based on copyrighted work, rather than reproducing it. 'The federal government can both secure Americans' freedom to learn from AI, and avoid forfeiting our AI lead by preserving American AI models' ability to learn from copyrighted material,' OpenAI added. The demands have thrust the issue back into the spotlight. For while tech companies argue fair use is a legal doctrine that grants them the ability to train their models based on other people's work, so far they have been cagey about acknowledging how they use copyrighted work because of potential legal threats and the inevitable murky interpretations of what constitutes fair use. Hollywood creatives and publishers quickly mobilized to push back,. More than 400 creatives and executives signed an open letter to the Trump Administration on March 15, calling for the White House to reject OpenAI and Google's proposals for less stringent copyright rules. Signees included Gordon-Levitt, Ben Stiller, Aubrey Plaza, Cate Blanchett, Natasha Lyonne, Chris Rock, Paul McCartney, Judd Apatow and Ava Duvernay. They proposed that AI companies be required to negotiate 'appropriate' licenses with copyright holders – meaning they want to get paid. For Gordon-Levitt — who first shared his concerns over AI's use of copyrighted work in The Washington Post in a 2023 op-ed — this is simply a matter of fairness. 'The basic principle I'm talking about is, if a human being creates some value, should a gigantic tech company be allowed to take that value, make money with it, and not pay the human being or ask permission? To me that is very clearly the wrong thing to do,' Gordon-Levitt said. OpenAI has rapidly grown into one of the biggest tech companies in the world by leaching off of the work of not only artists, but 'nearly all' other industries, from medicine to marketing, Gordon-Levitt said. The company led by CEO Sam Altman is set to close a $40 billion fundraising round that would value OpenAI at $300 billion. 'These new products that are getting called artificial intelligence, they are worth exactly $0 and can do exactly nothing without massive troves of data that was produced by human beings,' he added. 'And yet they want to sell these products that they build out of our data — not just Hollywood's data, everyone's data.' 'Blue Beetle' director Ángel Manuel Soto told TheWrap he signed the letter because he is worried the OpenAI and Google proposals will erode protections for intellectual property owners. Such a move would undermine creativity and devastate the filmmaking business, he said. 'As a creator, I am often working with other people who have IP, or I am creating my own IP — our copyrights are pretty much the only ownership we have left over what we create,' Manuel Soto said. 'So when you're working on collaborating, having all these elements come to fruition when it comes to what benefits us as creatives, copyright tends to be the thing that actually is meant to have our back.' Like Gordon-Levitt, he stressed he is not against AI or its use in Hollywood; ideally, the major AI companies would work together with entertainment studios to make sure artists are paid when their work is used by AI models. 'Every time a breakthrough in technology happens, it always has creatives at the front-end of it, spearheading the process. so that it is done ethically, and that it is not taking our jobs, but actually making our jobs more efficient,' Manuel Soto said. Similar concerns apply to other aspects of the media. This week, several outlets owned by Alden Global Capital, the nation's second-largest newspaper publisher, criticized OpenAI and Google's proposals, saying they were nothing more than a 'license to steal' in an editorial. 'OpenAI and Google — having long trained their ravenous bots on the work of newsrooms like this one — now want to throw out long-established copyright law by arguing, we kid you not, that the only way for the United States to defeat the Chinese Communist Party is for those tech giants to steal the content created with the sweat equity of America's human journalists,' the editorial said. Google, in its White House proposal, argued 'fair use and text-and-data-mining exceptions' are 'critical' to retaining and advancing America's AI lead over its chief competitor, China. Many news outlets, including The New York Times, which is suing OpenAI nd its partner Microsoft for what it calls the 'unlawful use' of its stories to train ChatGPT, believe that is not kosher. The Writers Guild of America, in late 2024, pushed Hollywood studios to follow a similar path and take 'immediate legal action' against AI companies using writers' work. Other media companies have found a way to work with the tech companies, like The New York Post, which struck a licensing deal with OpenAI last year. Complicating matters is the Wild West nature of it all: There are currently no federal copyright laws that exclusively address AI usage. OpenAI, in its proposal, acknowledged concerns over a carte blanche approach to using copyright material. The company said it is important for the U.S. government to safeguard 'the rights and interests of content creators while also protecting America's AI leadership.' OpenAI did not share specific recommendations on how to strike that balance. 'We support America's existing fair use framework, and we're confident that current copyright law enables AI innovation,' a Google rep told TheWrap. Reps for OpenAI did not respond to a request for comment. Richard Tofel, the former president of ProPublica and the current principal of the media consulting firm Gallant Advisory, does not expect any clarity on this issue soon. He told TheWrap the relationship between AI companies and entertainment and media companies will continue operating in a legal gray area for the foreseeable future. What will change the dynamic? A legal ruling, he said. 'The dispute is ultimately going to have to be resolved by the Supreme Court,' Tofel said. 'It revolves around how copyright law applies to a technology that was created long after the copyright law was enacted.' Tofel noted the Supreme Court has had to weigh in on disruptive technologies before. He pointed to the famous 1984 'Betamax' case, which found Sony — the Betamax manufacturer — was not violating copyright law by allowing owners to record TV shows for private use; the decision helped spur the widespread adoption of video recording machines like the VCR. The Supreme Court will ultimately have to clarify to what degree AI models can lean on copyrighted work to train their models, Tofel said. 'We're in a holding pattern that will be marked by some [licensing] agreements, and some litigation,' Tofel said. 'And depending on how the Court resolves it, the negotiating power of the two sides will surely be not the same as it was before.' For now, the media and entertainment industries are left searching for answers on how they can get paid when AI models like ChatGPT use their work to generate responses and content. One option that both Gordon-Levitt and Manuel Soto said they are intrigued by is a Spotify-like framework, where AI companies pay creators whenever their work is factored into an answer or AI-generated content, similar to how streaming services pay musicians whenever their songs are played. For now, though, Gordon-Levitt said the biggest issue is merely getting the AI giants to acknowledge and compensate creators for making their products useful. 'What they want to say is, 'Well, actually, all that creativity and ingenuity that everybody contributed to our products — that's not worth anything. Only our half of the value is worth money. And so all the money should go to us, and none of the money should go to all of the people whose ingenuity and creativity and work and individuality went into the data that made these products,'' he said. The post AI + Trump Amps Up a New Wave of Fear in Hollywood, Media appeared first on TheWrap.
The Guardian
26-02-2025
- Entertainment
- The Guardian
Michelle Trachtenberg: a life in pictures
Michelle Trachtenberg and Rosie O'Donnell in Harriet the Spy, in 1996. Photograph: Paramount/Kobal/Shutterstock Ed Begley Jr, Erik von Detten, Michelle Trachtenberg and Jonathan Lipnicki in Meego, in 1997. Photograph: CBS Photo Archive/CBS/Getty Images Michelle Trachtenberg at the Inspector Gadget premiere, in 1999. Photograph: Rose Prouser/REUTERS Michelle Trachtenberg and Sarah Michelle Gellar in Buffy the Vampire Slayer, in 2001. Photograph: Landmark Media/Alamy Michelle Trachtenberg in Buffy The Vampire Slayer, in 2003. Photograph: 20th Century Fox Television/Kobal/REX/Shutterstock Michelle Trachtenberg and Joss Whedon during the taping of the final episode of Buffy the Vampire Slayer, in 2003. Photograph: Damian Dovarganes/AP Joseph Gordon-Levitt, Jeffrey Licon and Michelle Trachtenberg in Mysterious Skin, in 2004. Photograph: film still handout Travis Wester, Scott Mechlowicz, Michelle Trachtenberg and Jacob Pitts in Eurotrip, in 2004. Photograph: RGR Collection/Alamy Joseph Gordon-Levitt and Michelle Trachtenberg, in 2005. Photograph: HNW/Plux/REX/Shutterstock Michelle Trachtenberg in Ice Princess, in 2005. Photograph: Moviestore Collection Ltd/Alamy Leighton Meester and Michelle Trachtenberg in Gosspi Girl, in 2009. Photograph: Everett Collection Inc/Alamy Michelle Trachtenberg and Blake Lively, in 2009. Photograph: Evan Agostini/AP Bruce Willis and Michelle Trachtenberg in Cop Out, in 2010. Photograph: Abbot Genser/Publicity image from film company Michelle Trachtenberg at the Critics' Choice Television awards, in 2014. Photograph: Danny Moloshok/Reuters Michelle Trachtenberg at the Elton John Aids Foundation Oscar viewing party, in 2020. Photograph: Stewart Cook/REX/Shutterstock Michelle Trachtenberg at the 15 Years of Siriano party, in 2023. Photograph: Gilbert Flores/WWD/Getty Images
