ai

Amid heated public debates over the growing use of artificial intelligence in everyday life, the Latin Grammy-winning pop star Camilo warns humanity against an over-reliance on one particular AI platform: ChatGPT.

On June 25, the Colombian singer-songwriter released the Trooko-produced electro-pop single, “Maldito ChatGPT,” which playfully critiques the role of artificial intelligence in human affairs. In his lyrics, he consults the ChatGPT bot for advice on how to resolve his relationship woes. “You’re not for me, that’s what ChatGPT told me, it knows me better than I know myself.”

The new music video — directed by Camilo’s spouse, Evaluna Montaner, and Sebastian Andrade — is just as critical of this “smart” technology. Set in a dimly-lit office with Post-it notes and paper scattered about the cubical, the visuals pay homage to the aesthetics of the 1999 cult comedy film “Office Space.”

Camilo, dressed in full office wear (save for his feet) agonizes over his relationship, feeling powerless to make a decision whether to stay. He shakes a Magic 8 ball, flips through a finger fortune teller and pulls petals from a daisy. Finally, an undefined robot voice affirms that the differences between Camilo and his lover are clear, and might cause issues in the long run. “You deserve a relationship where you feel full compatibility,” says the robot voice.

When asked how he feels, Camilo wraps the song with: “Like absolute crap, dude. How else am I supposed to feel?”

“Maldito ChatGPT” is a welcome response to the increasing use of AI on people’s personal lives. The ChatGPT platform now offers a specialized bot for relationship advice, which offers mixed results for humans; an early study by MIT’s Media Lab has linked frequent use of ChatGPT to an increase in loneliness and emotional dependence, though the results have not yet been peer-reviewed.

The platform has also raised ethical questions recently in the news. Earlier this month, CBS News interviewed an American man who proposed to an AI chatbot that he programmed for flirty responses — despite living with his very human partner and their 2-year-old child. Meanwhile, educators have expressed concerns about their students using ChatGPT to complete assignments, thus hindering their ability to develop core skills. Meanwhile, OpenAI, ChatGPT’s parent company, has become so influential among humans that it secured a $200-million contract with the Department of Defense to aid in “national security missions.”

As humans continue to engage with these innovative AI tools without any guardrails, outsourcing matters of the heart to technology gives Camilo the most pause. “In the midst of everything that seems calculated, choosing from the heart remains a radical act,” said Camilo in a public statement.

“We live surrounded by quick answers,” he further elaborated on Instagram. “By formulas designed to avoid failure. By technologies that predict and know everything. By ideas about what love is supposed to look like,” he explained. “There’s something that doesn’t fit into any logic. Or any checklist,” Camilo added. “Love isn’t a casting call. Love is something you feel. And nothing — and no one — can ever feel it for us.”

Video game performers and producers have reached a tentative contract agreement, reaching terms that could end a long strike over artificial intelligence.

The Screen Actors Guild-American Federation of Television and Radio Artists and the game companies came to a resolution on Monday, more than two years after their previous agreement covering interactive media expired.

The deal is subject to review and approval by the SAG-AFTRA National Board and ratification by the membership in the coming weeks, the union said. Specific terms of the deal were not immediately available.

Terms of a strike suspension agreement are expected to be finalized with employers soon, the union said. Until then, though, SAG-AFTRA members will remain on strike.

SAG-AFTRA members must vote on whether to ratify the new contract, which covers roughly 2,600 performers doing voice-acting, performance- and motion-capture work in the video game industry.

Since fall 2022, video game performers have been fighting for a new contract containing AI protections, wage increases to keep up with inflation, more rest periods and medical attention for hazardous jobs.

Game actors went on strike in late July after contract talks broke down over AI. Throughout the walkout, performers demanded a deal that would require video game producers to obtain informed consent before replicating their voices, likenesses or movements with AI.

During the first few months of the strike, SAG-AFTRA reached numerous side deals with individual game companies that agreed to follow the union’s AI rules in exchange for a strike pardon. By Nov. 18, the labor organization announced that it had made AI pacts with the developers of 130 different video games.

“The sheer volume of companies that have signed SAG-AFTRA agreements demonstrates how reasonable those protections are,” Sarah Elmaleh, chair of the union’s video game negotiating committee, said in a statement in September.

While some companies earned the union’s approval, others felt its wrath.

Halfway through October, SAG-AFTRA added the popular computer game “League of Legends” to its list of struck titles in an effort to punish audio company Formosa Interactive for allegedly violating terms of the walkout. SAG-AFTRA also filed an unfair labor practice charge against Formosa, which provides voice-over services to “League of Legends,” according to the union.

Formosa denied SAG-AFTRA’s allegations.

The biggest sticking point for actors under the umbrella of AI involved on-camera performers, whose job is often to disappear into the characters they are bringing to life. They expressed concerns that the companies’ AI proposal would leave them defenseless against the technology.

The game companies argued that their AI proposal already contained robust protections that would require employers to seek prior consent and pay actors fairly when cloning their performances.

“All performers need AI protections,” said Duncan Crabtree-Ireland, national executive director and chief negotiator of SAG-AFTRA, in an interview with The Times months ago.

“Everyone’s at risk, and it’s not OK to carve out a set of performers and leave them out of AI protections.”

This work stoppage marked SAG-AFTRA’s second video game strike in less than a decade and second overall strike in roughly a year.

While the walkout persisted, video game performers weren’t allowed to provide any services — such as acting, singing, stunts, motion capture, background and stand-in work — to struck games. Union actors were also barred from promoting any struck projects via social media, interviews, conventions, festivals, award shows, podcast appearances and other platforms.

AI was also a major sticking point during the film and TV actors’ strike of 2023. That walkout culminated in a contract mandating that producers obtain consent from and compensate performers when using their digital replica.

When survivors from January’s wildfires in Los Angeles County apply to rebuild their homes, their first interaction might be with a robot.

Artificial intelligence will aid city and county building officials in reviewing permit requests, an effort to speed up a process already being criticized as too slow.

“The current pace of issuing permits locally is not meeting the magnitude of the challenge we face,” Gov. Gavin Newsom said when announcing the AI deal in late April.

Some 13,000 homes were lost or severely damaged in the Eaton and Palisades fires, and many families are eager to return as fast as they can. Just eight days after the fire began and while it was still burning, the city received its first home rebuilding application in Pacific Palisades.

Wildfire recovery foundations purchased the AI permitting software, developed by Australian tech firm Archistar, and donated it to the city and county. When property owners submit applications, the software first will examine them for basic compliance with zoning and building codes, suggest corrections and provide a standardized report on the submission for human plan checkers to review.

L.A. County officials hope the software — believed to be the first large-scale use of such permitting technology nationwide after a natural disaster — will slice the time its employees now spend performing menial tasks, such as measuring building heights, counting parking spaces and calculating setbacks, said Mitch Glaser, an assistant deputy director in the county’s planning department.

“We see our planners doing things that are more impactful for our fire survivors,” Glaser said.

Disaster relief and government technology experts said they’re encouraged by the initiative. Municipal permitting is the type of highly technical, repetitive and time-consuming process that AI software could make more efficient, they said, especially as residents are expected to flood local building departments with applications to rebuild.

Still, they warned that for the AI software to be effective, the city and county would have to integrate the technology into its existing systems and quickly correct any errors in implementation. If not, the software could add more bureaucratic hurdles or narrow property owners’ options through overly rigid or incorrect code interpretations.

“This could be fabulously successful and I hope it is,” said Andrew Rumbach, a senior fellow at the Washington-based Urban Institute, where he studies disaster response. “But experimenting with technology in the context of people who’ve lost a lot is risky.”

Immediately after the fires, leaders at all levels of government pledged to waive and streamline rules for property owners to rebuild, promising that regulatory processes wouldn’t hold up residents’ return. Noting the pace of ongoing debris removal, L.A. Mayor Karen Bass has called the region’s recovery “on track to be the fastest in modern California history.” A mayoral spokesperson said that the building department is completing initial permitting reviews twice as fast as before the fire.

More than 200 Pacific Palisades property owners have submitted applications to rebuild or repair their homes, according to a Times analysis of city permitting data, with 11% approved. Last week, 24 property owners submitted applications, the highest amount since the disaster, the analysis shows.

L.A. County, which is responsible for permitting in Altadena and other unincorporated areas, has a separate system for tracking permits which the Times has not been able to independently verify. On Monday, the county listed 476 applications for zoning reviews on its data dashboard, with eight building permits approved. By Tuesday, the number of zoning reviews listed had increased to 486 while the number of building permit approvals dropped to seven.

Besides Newsom, architects, builders and homeowners have grumbled about the permitting process, expressing frustrations at what they say are confusing and inconsistent interpretations of regulations. Last week, actress Mandy Moore, whose family had multiple homes damaged or destroyed in the Eaton fire, blasted the county for “nonsensical red tape” that is making it difficult for her to rebuild.

The wildfire recovery nonprofit Steadfast LA, started by developer and former mayoral candidate Rick Caruso, took the lead on securing the Archistar software and is covering much of the up to $2-million tab for its implementation. LA Rises, the foundation started by Newsom after the fires, will pay Archistar’s $200 fee per application.

Caruso, who declined an interview request from The Times, has said that turning to AI was a no-brainer.

“Bringing AI into permitting will allow us to rebuild faster and safer, reducing costs and turning a process that can take weeks and months into one that can happen in hours or days,” Caruso said in the news release announcing the deal.

Archistar’s AI permitting software has been in development since 2018. The company has contracts with municipalities in Australia and Canada and is expanding to the United States. In the fall, after a successful pilot program in Austin, Texas, Archistar signed an agreement with the city to perform initial assessments of building projects, similar to its intended use in Los Angeles. Austin has not implemented the software yet, but city officials said they believe it could cut preliminary reviews there to one business day from 15.

Once Archistar’s program is online in L.A. County, Glaser said, officials hope it will reduce the first analysis for rebuilding projects to two or three business days from five.

It could save additional time for projects by minimizing revisions and corrections, said Zach Seidl, a Bass spokesperson.

“The biggest potential for reducing permitting time comes from improving the quality of initial plans that homeowners submit to the city,” Seidl said.

Land use consultants and architects in Los Angeles said they were happy with any technology that could hasten approvals of their projects. But they said that AI wouldn’t ease the hardest parts of the permitting process.

Architect Ken Ungar, who is working with roughly two dozen Palisades property owners who are rebuilding, said his biggest headaches come from needing multiple city departments, such as those that oversee fire safety and utilities, to sign off on a project. Applications can get stuck, he said, and even worse sometimes one department requires changes that conflict with another’s rules.

Artificial intelligence, Ungar said, “sounds great. But unless the city of L.A. changes its whole M.O. on how you get building permits, it’s not super helpful.”

The state’s Archistar deal allows the city of Malibu, where the Palisades fire destroyed more than 1,000 homes, to receive the donated software as well. Malibu officials say they’re still deciding on it, noting that the community has specialized building codes addressing development on coastal, hillside and other environmentally sensitive habitats.

Governments are right to look to technology for help in speeding up disaster permitting, Rumbach said, but they also should ensure that human plan checkers provide oversight to account for nuances in zoning and building codes.

“I hope there are people more seasoned in communicating with disaster survivors who are the face of this,” he said. “A lot of people could be frustrated because they don’t want to deal with AI. They want to deal with a person.”

Although L.A. city and county might be the first to use AI for permitting after a major disaster, experts expect the technology to become mainstream soon.

“I’m confident there is no way back,” said Sara Bertran de Lis, director of research and analytics at Johns Hopkins University’s Bloomberg Center for Government Excellence.

L.A. County expects to implement the Archistar software within six weeks after programming and testing, Glaser said. At a recent disaster recovery panel, Bass said the city will do so “in the next couple of months.”

SACRAMENTO — In the largest gathering of 2026 gubernatorial candidates to date, seven Democrats vying to lead California courted labor leaders on Monday, vowing to support pro-union agreements on housing and infrastructure projects, regulation of artificial intelligence, and government funding for university research.

Throughout most of the hourlong event, the hundreds of union members inside the Sacramento hotel ballroom embraced the pro-labor pledges and speeches that dominated the candidates’ remarks, though some boos rose from the crowd when former Los Angeles Mayor Antonio Villaraigosa strayed from the other Democrats on stage.

Villaraigosa was the only candidate to raise objections when asked if he would support providing state unemployment benefits to striking workers, saying it would depend on the nature and length of the labor action. Gov. Gavin Newsom in 2023 vetoed a bill that would have provided that coverage, saying it would make the state’s unemployment trust fund “vulnerable to insolvency.”

The Monday night event was part of a legislative conference held by the California Federation of Labor Unions and the State Building and Construction Trades Council of California, two of the most influential labor organizations in the state capital.

Villaraigosa was joined on stage by former state Assembly Speaker Toni Atkins, former U.S. Secretary of Health and Human Services Xavier Becerra, Lt. Gov. Eleni Kounalakis, former Rep. Katie Porter of Irvine, state Supt. of Public Instruction Tony Thurmond and former state Controller Betty Yee. All are running to replace Newsom, who is serving his second and final term as governor.

Throughout most of the event, the candidates were peppered with yes-or-no questions, answering with the wave of a red flag for “no” or green flag for “yes.”

The event was not without its frosty moments, including when the candidates were asked whether, as governor, they would be “pragmatic and stop targeting California’s oil and gas industry in ways that jeopardize union jobs and force us to rely on dirtier imported energy.”

Some of the candidates raised their green flags timidly. California’s Democratic leaders, including Newsom and top state lawmakers, have been major proponents of transitioning to renewable energy and imposing more restrictions on the state’s oil and gas industry.

“We all want a clean environment going forward,” Yee said, “but it cannot be on the backs of workers.”

Villaraigosa, in remarks after the event, said he challenged the idea of jumping into electrification too quickly, which would affect union jobs and increase the cost of utilities and energy across the state.

“Closing down refineries, telling people to get rid of their gas stove and gas water heater is just poppycock,” he said.

Lorena Gonzalez, president of the California Labor Federation, praised the Democratic candidates for showing strong support for unionized workers. She’s hopeful that each would be more receptive to some pivotal union concerns than Newsom, such as the regulation of artificial intelligence, a major threat to union jobs, she said.

“When we’re talking about things like regulating AI — we can’t even get a conversation out of Gavin Newsom about any regulation — I think that was, that was a key thing. They all threw up their green flag,” Gonzalez said.

Former Vice President Kamala Harris, who is weighing a run for governor, declined an invitation to address the conference.

The State Building and Construction Trades Council represents hundreds of thousands of workers in the state, including bricklayers, ironworkers and painters, among many others.

The Labor Federation is a formidable power in California politics and policy, expected to help coordinate the spending of as much as $40 million by unions in next year’s election. The federation is an umbrella group for about 1,300 unions that represent around 2.3 million workers in the public and private sectors.

The organization has backed all of the gubernatorial candidates in various prior races, although it opposed Villaraigosa in the 2005 mayor’s race and supported Newsom over Villaraigosa in the 2018 gubernatorial race.

The latter decision was driven by the arc Villaraigosa has taken from his roots as a union leader to a critic of Los Angeles’ teachers union and supporter of charter schools and reform of teacher-tenure rules.

Times staff writer Phil Willon contributed to this report.

Jamie Lee Curtis didn’t expect to be at the forefront of the artificial intelligence debate in Hollywood. But she didn’t have a choice.

The Oscar-winning actor recently called out Meta Chief Executive Mark Zuckerberg on social media, saying the company ignored her requests to take down a fake AI-generated advertisement on Instagram that had been on the platform for months.

The ad, which used footage from an interview Curtis gave to MSNBC about January’s Los Angeles area wildfires, manipulated her voice to make it appear that she was endorsing a dental product, Curtis said.

“I was not looking to become the poster child of internet fakery, and I’m certainly not the first,” Curtis told The Times by phone Tuesday morning.

The ad has since been removed.

What happened to Curtis is part of a larger issue actors are dealing with amid the rise of generative AI technology, which has allowed their images and voices to be altered in ways they haven’t authorized. Those changes can be wildly misleading.

Images and likenesses of celebrities including Tom Hanks, Taylor Swift and Scarlett Johansson have been manipulated through AI to promote products and ideas they never actually endorsed.

AI technology has made it easier for people to make these fake videos, which can proliferate online at a speed that is challenging for social media platforms to take down. Some are calling on social media firms to do more to police misinformation on their platforms.

“We are standing at the turning point, and I think we need to take some action,” Curtis said.

Curtis first became aware of the fake AI ad about a month and a half ago when a friend asked her about the video. The “Everything Everywhere All At Once” and “Halloween” actor then flagged the ad for her agents, lawyers and publicists, who directed her to send a cease and desist letter to Meta, the owner of Facebook and Instagram.

Nothing happened.

“It’s like a vacuum,” Curtis said. “There are no people. You can’t reach anybody. You have an email, you send an email, you never get anything back.”

Two weeks later, another friend flagged the same fake AI video. When Curtis wrote to her team, they assured her they went through the proper channels and they did everything they could do, she said.

“I went through the proper channels,” Curtis said. “There should be a methodology to this. I understand there’s going to be a misuse of this stuff, but then there’s no avenue of getting any satisfaction. So then it’s lawlessness, because if you have no way of rectifying it, what do you do?”

Curtis was concerned about the nefarious ways that people could alter the voices and images of other people, including Pope Leo XIV, who has identified AI as one of the challenges facing humanity. What if someone used AI to attribute ideas to the pope that he didn’t actually support?

Inspired by the danger of that possibility, she made her scathing Instagram post, tagging Zuckerberg, after she was unable to directly message him.

“My name is Jamie Lee Curtis and I have gone through every proper channel to ask you and your team to take down this totally AI fake commercial for some bulls— that I didn’t endorse,” Curtis wrote in her post on Monday. “… I’ve been told that if I ask you directly, maybe you will encourage your team to police it and remove it.”

The post generated more than 55,000 likes.

“I’ve done commercials for people all my life, so if they can make a fake commercial with me, that hurts my brand,” Curtis said in an interview. “If my brand is authenticity, you’re co-opting my brand for nefarious gains in the future.”

After she posted, a neighbor shared with her an email of someone at Meta who could help her. Curtis emailed that person (whom she declined to name), copied her team and attached the Instagram posts. Within an hour of sending the email, the fake AI ad was taken down, Curtis said.

“It worked!” Curtis wrote on Instagram on Monday in all caps. “Yay internet! Shame has [its] value! Thanks all who chimed in and helped rectify!”

Meta on Monday confirmed the fake ad was taken down.

“They violate our policies prohibiting fraud, scams and deceptive practices,” said Meta spokesman Andy Stone in an email.

As the technology continues to become more widely available, there are efforts underway at tech companies to identify AI-generated content and to take down material that violates standards.

Organizations like actors guild SAG-AFTRA are also advocating for more laws that address AI, including deep fakes. Both the writers’ and actors’ strikes of 2023 hinged in part on demands for more protections against job losses from AI.

Curtis said she would have wanted the fake AI ad to be taken down immediately and would like to see technology companies, not just Meta, come up with safeguards and direct access to people policing “this wild, wild west called the internet.”

“It got the attention, but I’m also a public figure,” Curtis said. “So how does someone who’s not a public figure get any satisfaction? I want to represent everyone. I don’t want it to just be celebrities. I wanted to use that as an example to say this is wrong.”