Toolbox

After a year of basking in global fame, the San Francisco company OpenAI is now confronting a multitude of challenges that could threaten its position at the vanguard of artificial intelligence research.

Some of its conflicts stem from decisions made well before the debut of ChatGPT, particularly its unusual shift from an idealistic nonprofit to a big business backed by billions of dollars in investments.

It's too early to tell if OpenAI and its attorneys will beat back a barrage of lawsuits from Elon Musk, The New York Times and bestselling novelists such as John Grisham, not to mention escalating scrutiny from government regulators, or if any of it will stick.

Feud with Elon Musk
OpenAI isn't waiting for the court process to unfold before publicly defending itself against legal claims made by billionaire Elon Musk, an early funder of OpenAI who now alleges it has betrayed its founding nonprofit mission to benefit humanity as it pursued profits instead.

In its first response since the Tesla CEO sued last week, OpenAI vowed to get the claim thrown out and released emails from Musk that purport to show he supported making OpenAI a for-profit company and even suggested merging it with the electric vehicle maker.

Legal experts have expressed doubt about whether Musk's arguments, centered around an alleged breach of contract, will hold up in court. But it has already forced open the company's internal conflicts about its unusual governance structure, how "open" it should be about its research and how to pursue what's known as artificial general intelligence, or AI systems that can perform just as well as — or even better than — humans in a wide variety of tasks.

Its own internal investigation
There's still a lot of mystery about what led OpenAI to abruptly fire its co-founder and CEO Sam Altman in November, only to have him return days later with a new board that replaced the one that ousted him. OpenAI tapped the law firm WilmerHale to investigate what happened, but it's unclear how broad its scope will be and to what extent OpenAI will publicly release its findings.

Among the big questions is what OpenAI — under its previous board of directors — meant in November when it said Altman was "not consistently candid in his communications" in a way that hindered the board's ability to exercise its responsibilities. While now primarily a for-profit business, OpenAI is still governed by a nonprofit board of directors whose duty is to advance its mission.

The investigators are probably looking more closely at that structure as well as the internal conflicts that led to communication breakdowns, said Diane Rulke, a professor of organizational behavior and theory at Carnegie Mellon University.

Rulke said it would be "useful and very good practice" for OpenAI to publicly release at least part of the findings, especially given the underlying concerns about how future AI technology will affect society.

"Not only because it was a major event, but because OpenAI works with a lot of businesses, a lot of companies and their impact is widespread," Rulke said. "Even though they're a privately held company, it's very much in the public interest to know what happened at OpenAI."

Government scrutiny
OpenAI's close business ties to Microsoft have invited scrutiny from antitrust regulators in the U.S. and Europe. Microsoft has invested billions of dollars into OpenAI and switched on its vast computing power to help build the smaller company's AI models. The software giant has also secured exclusive rights to infuse much of the technology into Microsoft products.

Unlike a big business merger, such partnerships don't automatically trigger a government review. But the Federal Trade Commission wants to know if such arrangements "enable dominant firms to exert undue influence or gain privileged access in ways that could undermine fair competition," FTC Chair Lina Khan said in January.

FTC is awaiting responses to "compulsory orders" it sent to both companies — as well as OpenAI rival Anthropic and its own cloud computing backers, Amazon and Google — requiring them to provide information about the partnerships and the decision-making around them. The companies' responses are due as soon as next week. Similar scrutiny is happening in the European Union and the United Kingdom.

Copyright lawsuits
Bestselling novelists, nonfiction authors, The New York Times and other media outlets have sued OpenAI over allegations that the company violated copyright laws in building the AI large language models that power ChatGPT. Several of the lawsuits also target Microsoft. (The Associated Press took a different approach in securing a deal last year that gives OpenAI access to the AP's text archive for an undisclosed fee).

OpenAI has argued that its practice of training AI models on huge troves of writings found on the internet is protected by the "fair use" doctrine of copyright law. Federal judges in New York and San Francisco must now sort through evidence of harm brought by numerous plaintiffs, including Grisham, comedian Sarah Silverman and "Game of Thrones" author George R. R. Martin.

The stakes are high. The Times, for instance, is asking a judge to order the "destruction" of all of OpenAI's GPT large language models — the foundation of ChatGPT and most of OpenAI's business — if they were trained on its news articles.

Matt O'Brien is an AP technologyh writer. AP business writer Kelvin Chan contributed to this report.

Whenever Madonna sings the 1980s hit "La Isla Bonita" on her concert tour, moving images of swirling, sunset-tinted clouds play on the giant arena screens behind her.

To get that ethereal look, the pop legend embraced a still-uncharted branch of generative artificial intelligence – the text-to-video tool. Type some words — say, "surreal cloud sunset" or "waterfall in the jungle at dawn" — and an instant video is made.

Following in the footsteps of AI chatbots and still image-generators, some AI video enthusiasts say the emerging technology could one day upend entertainment, enabling you to choose your own movie with customizable story lines and endings. But there's a long way to go before they can do that, and plenty of ethical pitfalls on the way.

For early adopters like Madonna, who's long pushed art's boundaries, it was more of an experiment. She nixed an earlier version of "La Isla Bonita" concert visuals that used more conventional computer graphics to evoke a tropical mood.

"We tried CGI. It looked pretty bland and cheesy and she didn't like it," said Sasha Kasiuha, content director for Madonna's Celebration Tour that continues through late April. "And then we decided to try AI."

ChatGPT-maker OpenAI gave a glimpse of what sophisticated text-to-video technology might look like when the company recently showed off Sora, a new tool that's not yet publicly available. Madonna's team tried a different product from New York-based startup Runway, which helped pioneer the technology by releasing its first public text-to-video model last March. The company released a more advanced "Gen-2" version in June.

Runway CEO Cristóbal Valenzuela said while some see these tools as a "magical device that you type a word and somehow it conjures exactly what you had in your head," the most effective approaches are by creative professionals looking for an upgrade to the decades-old digital editing software they're already using.

He said Runway can't yet make a full-length documentary. But it could help fill in some background video, or b-roll — the supporting shots and scenes that help tell the story.

"That saves you perhaps like a week of work," Valenzuela said. "The common thread of a lot of use cases is people use it as a way of augmenting or speeding up something they could have done before."

Runway's target customers are "large streaming companies, production companies, post-production companies, visual effects companies, marketing teams, advertising companies. A lot of folks that make content for a living," Valenzuela said.

Dangers await. Without effective safeguards, AI video-generators could threaten democracies with convincing "deepfake" videos of things that never happened, or — as is already the case with AI image generators — flood the internet with fake pornographic scenes depicting what appear to be real people with recognizable faces. Under pressure from regulators, major tech companies have promised to watermark AI-generated outputs to help identify what's real.

There also are copyright disputes brewing about the video and image collections the AI systems are being trained upon (neither Runway nor OpenAI discloses its data sources) and to what extent they are unfairly replicating trademarked works. And there are fears that, at some point, video-making machines could replace human jobs and artistry.

For now, the longest AI-generated video clips are still measured in seconds, and can feature jerky movements and telltale glitches such as distorted hands and fingers. Fixing that is "just a question of more data and more training," and the computing power on which that training depends, said Alexander Waibel, a computer science professor at Carnegie Mellon University who's been researching AI since the 1970s.

"Now I can say, 'Make me a video of a rabbit dressed as Napoleon walking through New York City,'" Waibel said. "It knows what New York City looks like, what a rabbit looks like, what Napoleon looks like."

Which is impressive, he said, but still far from crafting a compelling storyline.

Before it released its first-generation model last year, Runway's claim to AI fame was as a co-developer of the image-generator Stable Diffusion. Another company, London-based Stability AI, has since taken over Stable Diffusion's development.

The underlying "diffusion model" technology behind most leading AI generators of images and video works by mapping noise, or random data, onto images, effectively destroying an original image and then predicting what a new one should look like. It borrows an idea from physics that can be used to describe, for instance, how gas diffuses outward.

"What diffusion models do is they reverse that process," said Phillip Isola, an associate professor of computer science at the Massachusetts Institute of Technology. "They kind of take the randomness and they congeal it back into the volume. That's the way of going from randomness to content. And that's how you can make random videos."

Generating video is more complicated than still images because it needs to take into account temporal dynamics, or how elements within the video change over time and across sequences of frames, said Daniela Rus, another MIT professor who directs its Computer Science and Artificial Intelligence Laboratory.

Rus said the computing resources required are "significantly higher than for still image generation" because "it involves processing and generating multiple frames for each second of video."

That's not stopping some well-heeled tech companies from trying to keep outdoing each other in showing off higher-quality AI video generation at longer durations. Requiring written descriptions to make an image was just the start. Google recently demonstrated a new project called Genie that can be prompted to transform a photograph or even a sketch into "an endless variety" of explorable video game worlds.

In the near term, AI-generated videos will likely show up in marketing and educational content, providing a cheaper alternative to producing original footage or obtaining stock videos, said Aditi Singh, a researcher at Cleveland State University who has surveyed the text-to-video market.

When Madonna first talked to her team about AI, the "main intention wasn't, 'Oh, look, it's an AI video,'" said Kasiuha, the creative director.

"She asked me, 'Can you just use one of those AI tools to make the picture more crisp, to make sure it looks current and looks high resolution?'" Kasiuha said. "She loves when you bring in new technology and new kinds of visual elements."

Longer AI-generated movies are already being made. Runway hosts an annual AI film festival to showcase such works. But whether that's what human audiences will choose to watch remains to be seen.

"I still believe in humans," said Waibel, the CMU professor. "I still believe that it will end up being a symbiosis where you get some AI proposing something and a human improves or guides it. Or the humans will do it and the AI will fix it up."

Matt O'Brien is an AP technology writer. AP journalists Joseph B. Frederick and Rodrique Ngowi contributed to this report.

Google apologized Friday for its faulty rollout of a new artificial intelligence image-generator, acknowledging that in some cases the tool would "overcompensate" in seeking a diverse range of people even when such a range didn't make sense.

The partial explanation for why its images put people of color in historical settings where they wouldn't normally be found came a day after Google said it was temporarily stopping its Gemini chatbot from generating any images with people in them. That was in response to a social media outcry from some users claiming the tool had an anti-white bias in the way it generated a racially diverse set of images in response to written prompts.

"It's clear that this feature missed the mark," said a blog post Friday from Prabhakar Raghavan, a senior vice president who runs Google's search engine and other businesses. "Some of the images generated are inaccurate or even offensive. We're grateful for users' feedback and are sorry the feature didn't work well."

Raghavan didn't mention specific examples but among those that drew attention on social media this week were images that depicted a Black woman as a U.S. founding father and showed Black and Asian people as Nazi-era German soldiers. The Associated Press was not able to independently verify what prompts were used to generate those images.

Google added the new image-generating feature to its Gemini chatbot, formerly known as Bard, about three weeks ago. It was built atop an earlier Google research experiment called Imagen 2.

Google has known for a while that such tools can be unwieldly. In a 2022 technical paper, the researchers who developed Imagen warned that generative AI tools can be used for harassment or spreading misinformation "and raise many concerns regarding social and cultural exclusion and bias." Those considerations informed Google's decision not to release "a public demo" of Imagen or its underlying code, the researchers added at the time.

Since then, the pressure to publicly release generative AI products has grown because of a competitive race between tech companies trying to capitalize on interest in the emerging technology sparked by the advent of OpenAI's chatbot ChatGPT.

The problems with Gemini are not the first to recently affect an image-generator. Microsoft had to adjust its own Designer tool several weeks ago after some were using it to create deepfake pornographic images of Taylor Swift and other celebrities. Studies have also shown AI image-generators can amplify racial and gender stereotypes found in their training data, and without filters they are more likely to show lighter-skinned men when asked to generate a person in various contexts.

"When we built this feature in Gemini, we tuned it to ensure it doesn't fall into some of the traps we've seen in the past with image generation technology — such as creating violent or sexually explicit images, or depictions of real people," Raghavan said Friday. "And because our users come from all over the world, we want it to work well for everyone."

He said many people might "want to receive a range of people" when asking for a picture of football players or someone walking a dog. But users looking for someone of a specific race or ethnicity or in particular cultural contexts "should absolutely get a response that accurately reflects what you ask for."

While it overcompensated in response to some prompts, in others it was "more cautious than we intended and refused to answer certain prompts entirely — wrongly interpreting some very anodyne prompts as sensitive."

He didn't explain what prompts he meant but Gemini routinely rejects requests for certain subjects such as protest movements, according to tests of the tool by the AP on Friday, in which it declined to generate images about the Arab Spring, the George Floyd protests or Tiananmen Square. In one instance, the chatbot said it didn't want to contribute to the spread of misinformation or "trivialization of sensitive topics."

Much of this week's outrage about Gemini's outputs originated on X, formerly Twitter, and was amplified by the social media platform's owner Elon Musk who decried Google for what he described as its "insane racist, anti-civilizational programming." Musk, who has his own AI startup, has frequently criticized rival AI developers as well as Hollywood for alleged liberal bias.

Raghavan said Google will do "extensive testing" before turning on the chatbot's ability to show people again.

University of Washington researcher Sourojit Ghosh, who has studied bias in AI image-generators, said Friday he was disappointed that Raghavan's message ended with a disclaimer that the Google executive "can't promise that Gemini won't occasionally generate embarrassing, inaccurate or offensive results."

For a company that has perfected search algorithms and has "one of the biggest troves of data in the world, generating accurate results or unoffensive results should be a fairly low bar we can hold them accountable to," Ghosh said.

Matt O'Brien is an AP technology writer

Two tech CEOs scrambling to produce more of the sophisticated chips needed for artificial intelligence met for a brainstorming session Wednesday while the booming market's early leader reported another quarter of eye-popping growth.

The on-stage conversation between Intel CEO Pat Gelsinger and OpenAI CEO Sam Altman unfolded in a San Jose, California, convention center a few hours after Nvidia disclosed its revenue for the November-January period nearly quadrupled from the previous year.

Intel, a Silicon Valley pioneer that has been struggling in recent years, laid out its plans for catching up to Nvidia during a daylong conference. Gelsinger kicked things off with a opening speech outlining how he envisions the feverish demand for AI-equipped chips revitalizing his company in a surge he dubbed the "Siliconomy."

"It's just magic the way these tiny chips are enabling the modern economic cycle we are in today," Gelsinger said.

OpenAI, a San Francisco startup backed by Microsoft, has become one of technology's brightest stars since unleashing its most popular AI innovation, ChatGPT, in late 2022. Altman is now eager to push the envelope even further while competing against Google and other companies such as Anthropic and Inflection AI. But the next leaps he wants to make will take far more processing power than what's currently available.

The imbalance between supply and the voracious appetite for AI chips explains why Altman is keenly interested in securing more money to help expand the industry's manufacturing capacity. During his talk with Gelsinger, he dodged a question about whether he is trying to raise as much as $7 trillion — more the combined market value of Microsoft and Apple — as was recently reported by The Wall Street Journal.

"The kernel of truth is we think the world is going to need a lot more (chips for) AI compute," Altman said. "That is going to require a global investment in a lot of stuff beyond what we are thinking of. We are not in a place where we have numbers yet."

Altman emphasized the importance of accelerating the AI momentum of the past year to advance a technology that he maintains will lead to a better future for humanity, although he acknowledged there will be downsides along the way.

"We are heading to a world where more content is going to be generated by AI than content generated by humans," Altman said. "This is not going to be only a good story, but it's going to be a net good story."

Perhaps no company is benefiting more from the AI gold rush now than Nvidia. The 31-year-old chipmaker has catapulted to the technological forefront because of its head start in making the graphics processing units, or GPUs, required to fuel popular AI products such as ChatGPT and Google's Gemini chatbot.

Over the past year, Nvidia has been a stunning streak of growth that has created more than $1.3 trillion in shareholder wealth in less than 14 months. That has turned it into the fifth most valuable U.S. publicly traded company behind only Microsoft, Apple, Amazon and Google's corporate parent, Alphabet Inc.

Intel, in contrast, has been trying to convince investors that Gelsinger has the Santa Clara, California, company on a comeback trail three years after he was hired as CEO.

Since his arrival, Gelsinger already has pushed the company into the business of making chips for other firms and has committed $20 billion to building new factories in Ohio as part of its expansion into running so-called "foundries" for third parties.

During Wednesday's conference, Gelsinger predicted that by 2030 Intel would be overseeing the world's second largest foundry business, presumably behind the current leader, Taiwan Semiconductor Manufacturing Co., or TMSC, largely by meeting the demand for AI chips.

"There's sort of a space race going on," Gelsinger told reporters Wednesday after delivering the conference's keynote speech. "The overall demand (for AI chips) appears to be insatiable for several years into the future."

Gelsinger's turnaround efforts haven't impressed investors so far. Intel's stock price has fallen by 30% under his reign while Nvidia's shares have increased by roughly fivefold during the same span.

Intel also is angling for a chunk of the $52 billion that the U.S. Commerce Department plans to spread around in an effort to increase the country's manufacturing capacity in the $527 billion market for processors, based on last year's worldwide sales.

Less than $2 billion of the funds available under the 2022 CHIPs and Science Act has been awarded so far, but Commerce Secretary Gina Raimondo, in a video appearance at Wednesday's conference, promised "a steady drumbeat" of announcements about more money being distributed.

Raimondo also told Gelsinger that she emerged from recent discussions with Altman and other executives leading the AI movement having a difficult time processing how big the market could become.

"The volume of chips they say they need is mind-boggling," she said.