TELLURIDE, Colo. — For four days every Labor Day weekend, the Telluride Film Festival turns this tiny mountain town into something of a refuge from the outside world. Tucked into a box canyon and far from the usual machinery of Hollywood, moviegoers spend their days seeing films they may have known little about before arriving and talking about them almost nonstop.
Artificial intelligence, apparently, did not get the memo.
This year, even Telluride’s usually self-contained movie bubble couldn’t keep out the existential anxieties surrounding AI. On Saturday, the festival devoted two events to the subject, beginning with “The Humanity Dilemma,” an hourlong multimedia presentation mixing dire warnings with live music and imagery, and continuing with a panel bluntly titled “AI AI AI,” featuring Microsoft co-founder Bill Gates.
Introducing “The Humanity Dilemma” at the nearly packed Sheridan Opera House, festival executive director Julie Huntsinger acknowledged that Telluride normally steers clear of taking political sides. “We just let the program speak for us,” Huntsinger said.
But on AI, she made an exception.
“We are a very humanist festival — that’s how we identify,” Huntsinger said. “AI is a threat.” While acknowledging potentially beneficial applications, she urged the audience to become more engaged in the debate over its development.
“We all need to be very alert and vigilant and loving and kind and stop this s—,” she said, earning loud whoops of approval from some in the crowd.
AI already occupies an uneasy place in Hollywood, where filmmakers and studios are experimenting with the technology even as actors, writers and other creative workers fight to protect their jobs and likenesses from being exploited or replaced.
For all of Huntsinger’s warnings, Telluride is not treating AI as off-limits. Its program includes “Love, Rendered,” a short documentary directed by Liz Garbus and produced by Darren Aronofsky about an elderly couple confronting the husband’s memory loss. Using AI along with family photographs and stories from loved ones, the filmmakers recreate the moment the couple first met and fell in love 70 years earlier. On Sunday, the film’s creative team is scheduled to discuss the project at an event titled “Filmmaking and Technology With Empathy: AI for Societal Benefit.”
“The Humanity Dilemma” was presented by Tristan Harris and Aza Raskin, co-founders of the Center for Humane Technology, who appeared in the 2020 documentary “The Social Dilemma,” along with artist and researcher Joy Mauthe and violinist Andrei Matorin. Harris and Raskin also appear in this year’s “The AI Doc: Or How I Became an Apocaloptimist.”
The presentation was designed to work on the emotions as well as the intellect. Original songs about humanity’s relationship to AI ran through much of the hour, with Mauthe singing and playing guitar and Matorin on violin, as Harris and Raskin delivered warnings accompanied by images of environmental devastation, poverty and figures in the AI race including Elon Musk, Sam Altman and Larry Ellison.
Their argument was stark: The race to build more powerful AI is moving far faster than governments or the public can keep up. Their deepest fear is that humans could eventually lose control of increasingly capable AI systems altogether.
At the same time, they acknowledged the technology’s potentially transformative benefits.
“AI is confusing because it represents both simultaneous utopia and dystopia,” Raskin said. “Just imagine having to reason about a nuke that could also solve cancer.”
By the end, Harris was calling for a halt to the development of more powerful systems.
“We need to pause frontier AI development and pivot and steer towards a pro-human future,” he said.
The hour concluded with Mauthe leading the audience in a final refrain, singing that the future is “still in our hands.” It was the kind of unabashedly earnest moment that could easily have tipped into awkwardness, but many in the room seemed to embrace it.
A few hours later, Harris and Raskin returned for “AI AI AI,” held outdoors at the Abel Gance Open Air Cinema in Telluride’s Elks Park, joined by Gates and filmmaker Joshua Oppenheimer, director of the Oscar-nominated documentaries “The Act of Killing” and “The Look of Silence,” who served as a moderator.
From left, Bill Gates, Aza Raskin, Tristan Harris and filmmaker Joshua Oppenheimer discuss the risks and potential benefits of artificial intelligence at the Telluride Film Festival on Saturday.
(Josh Rottenberg)
For Gates, who has spent decades focusing much of his philanthropy and public advocacy on global health and poverty, the rise of AI has forced him to make room for a new priority. In an essay published last week, he warned that the technology was improving faster than he anticipated and called for a new framework to manage its risks.
At Telluride, Gates said he now feels compelled to devote some of the political influence he has long used to advocate for causes such as malnutrition, polio and malaria to raise alarms about AI as well.
“Is this the greatest problem humanity has ever faced?” Gates said. “That is just a fact.”
Gates said his concerns have grown as AI systems have become more capable, particularly at writing computer code, something he has been obsessed with since he was 13.
“The AIs are superhuman, i.e., better than I am at writing code,” Gates said.
He laid out five broad areas of concern: jobs, biotechnology, cyberattacks, psychosocial harms and whether humans will be able to maintain control over the systems they are building.
Oppenheimer said that after spending hours digging into AI risks ahead of the panel, he had come away “absolutely terrified.” He read aloud an answer he had received after asking ChatGPT to estimate the risks if the race toward increasingly powerful AI continued with little regulation.
The chatbot put the chance of “persistent dystopian outcomes for all humanity” at 25% to 40%, a civilization-scale catastrophe at 5% to 15% and human extinction or permanent loss of human control at 5% to 10%.
“That’s what AI thinks,” Oppenheimer said.
Gates stressed AI’s potential upside, citing advances in medicine, education and assistance to farmers in poor countries. But pressed on whether development should be slowed until safety measures catch up, he said he would be open to the idea.
“If there was a credible plan that would cause this to be either slowed down or, you know, even held in stasis for a period of time, I would likely support that,” Gates said.
The difficulty, he said, is that multiple developers in both the U.S. and China are pushing closer to the technological frontier. Still, Gates rejected the argument that competition with China makes meaningful regulation impossible.
“China does not want cyberattacks, bioattacks or loss of control,” Gates said.
The discussion was punctuated by a pair of outbursts from the audience. At one point, as Gates discussed the economic incentives driving AI development, a man shouted that they represented “everything that’s wrong with America” and invoked Jesus Christ. Gates paused before responding dryly: “Anyway, not everything is wrong with America, according to me.”
Gates argued that government and civil society need to move much more quickly.
“This five-year period, in my view, is a very critical period,” he said, adding that responses that take five or six years to get underway could come too late.
Despite the dire warnings, Harris pointed to signs that pressure for safeguards is growing, including calls from AI-industry employees to limit the development of more powerful systems. Just last week, Meta agreed to pay $17 billion to settle claims by 29 states that Facebook and Instagram harmed young users while also agreeing to new child-safety measures on the platforms.
“I just want to leave you with not naive optimism, but momentum,” Harris said.
Near the end, Raskin suggested that President Trump’s well-documented desire to win the Nobel Peace Prize might provide an incentive for him to pursue a U.S.-China pause on more powerful AI systems.
“What’s better than solving one war?” Raskin said. “Saving all of humanity forever from uncontrollable AI.”
Gates, with a wry smile, called the Nobel angle “smart.”
“Whoever really takes the step to solve this AI problem would deserve the Nobel Prize,” he said.
Later that night, the festival would hold an outdoor screening of Stanley Kubrick’s “2001: A Space Odyssey,” featuring HAL 9000, one of cinema’s most famous rogue artificial intelligences.
Gates framed the current stakes in similarly cinematic terms.
“There are these movies where the aliens are coming, and you see unprecedented levels of cooperation between China and the U.S.,” he said. “Well, believe me, the aliens are here.”
TELLURIDE, Colo. — For four days every Labor Day weekend, the Telluride Film Festival turns this tiny mountain town into something of a refuge from the outside world. Tucked into a box canyon and far from the usual machinery of Hollywood, moviegoers spend their days seeing films they may have known little about before arriving and talking about them almost nonstop.
Artificial intelligence, apparently, did not get the memo.
This year, even Telluride’s usually self-contained movie bubble couldn’t keep out the existential anxieties surrounding AI. On Saturday, the festival devoted two events to the subject, beginning with “The Humanity Dilemma,” an hourlong multimedia presentation mixing dire warnings with live music and imagery, and continuing with a panel bluntly titled “AI AI AI,” featuring Microsoft co-founder Bill Gates.
Introducing “The Humanity Dilemma” at the nearly packed Sheridan Opera House, festival executive director Julie Huntsinger acknowledged that Telluride normally steers clear of taking political sides. “We just let the program speak for us,” Huntsinger said.
But on AI, she made an exception.
“We are a very humanist festival — that’s how we identify,” Huntsinger said. “AI is a threat.” While acknowledging potentially beneficial applications, she urged the audience to become more engaged in the debate over its development.
“We all need to be very alert and vigilant and loving and kind and stop this s—,” she said, earning loud whoops of approval from some in the crowd.
AI already occupies an uneasy place in Hollywood, where filmmakers and studios are experimenting with the technology even as actors, writers and other creative workers fight to protect their jobs and likenesses from being exploited or replaced.
For all of Huntsinger’s warnings, Telluride is not treating AI as off-limits. Its program includes “Love, Rendered,” a short documentary directed by Liz Garbus and produced by Darren Aronofsky about an elderly couple confronting the husband’s memory loss. Using AI along with family photographs and stories from loved ones, the filmmakers recreate the moment the couple first met and fell in love 70 years earlier. On Sunday, the film’s creative team is scheduled to discuss the project at an event titled “Filmmaking and Technology With Empathy: AI for Societal Benefit.”
“The Humanity Dilemma” was presented by Tristan Harris and Aza Raskin, co-founders of the Center for Humane Technology, who appeared in the 2020 documentary “The Social Dilemma,” along with artist and researcher Joy Mauthe and violinist Andrei Matorin. Harris and Raskin also appear in this year’s “The AI Doc: Or How I Became an Apocaloptimist.”
The presentation was designed to work on the emotions as well as the intellect. Original songs about humanity’s relationship to AI ran through much of the hour, with Mauthe singing and playing guitar and Matorin on violin, as Harris and Raskin delivered warnings accompanied by images of environmental devastation, poverty and figures in the AI race including Elon Musk, Sam Altman and Larry Ellison.
Their argument was stark: The race to build more powerful AI is moving far faster than governments or the public can keep up. Their deepest fear is that humans could eventually lose control of increasingly capable AI systems altogether.
At the same time, they acknowledged the technology’s potentially transformative benefits.
“AI is confusing because it represents both simultaneous utopia and dystopia,” Raskin said. “Just imagine having to reason about a nuke that could also solve cancer.”
By the end, Harris was calling for a halt to the development of more powerful systems.
“We need to pause frontier AI development and pivot and steer towards a pro-human future,” he said.
The hour concluded with Mauthe leading the audience in a final refrain, singing that the future is “still in our hands.” It was the kind of unabashedly earnest moment that could easily have tipped into awkwardness, but many in the room seemed to embrace it.
A few hours later, Harris and Raskin returned for “AI AI AI,” held outdoors at the Abel Gance Open Air Cinema in Telluride’s Elks Park, joined by Gates and filmmaker Joshua Oppenheimer, director of the Oscar-nominated documentaries “The Act of Killing” and “The Look of Silence,” who served as a moderator.
From left, Bill Gates, Aza Raskin, Tristan Harris and filmmaker Joshua Oppenheimer discuss the risks and potential benefits of artificial intelligence at the Telluride Film Festival on Saturday.
(Josh Rottenberg)
For Gates, who has spent decades focusing much of his philanthropy and public advocacy on global health and poverty, the rise of AI has forced him to make room for a new priority. In an essay published last week, he warned that the technology was improving faster than he anticipated and called for a new framework to manage its risks.
At Telluride, Gates said he now feels compelled to devote some of the political influence he has long used to advocate for causes such as malnutrition, polio and malaria to raise alarms about AI as well.
“Is this the greatest problem humanity has ever faced?” Gates said. “That is just a fact.”
Gates said his concerns have grown as AI systems have become more capable, particularly at writing computer code, something he has been obsessed with since he was 13.
“The AIs are superhuman, i.e., better than I am at writing code,” Gates said.
He laid out five broad areas of concern: jobs, biotechnology, cyberattacks, psychosocial harms and whether humans will be able to maintain control over the systems they are building.
Oppenheimer said that after spending hours digging into AI risks ahead of the panel, he had come away “absolutely terrified.” He read aloud an answer he had received after asking ChatGPT to estimate the risks if the race toward increasingly powerful AI continued with little regulation.
The chatbot put the chance of “persistent dystopian outcomes for all humanity” at 25% to 40%, a civilization-scale catastrophe at 5% to 15% and human extinction or permanent loss of human control at 5% to 10%.
“That’s what AI thinks,” Oppenheimer said.
Gates stressed AI’s potential upside, citing advances in medicine, education and assistance to farmers in poor countries. But pressed on whether development should be slowed until safety measures catch up, he said he would be open to the idea.
“If there was a credible plan that would cause this to be either slowed down or, you know, even held in stasis for a period of time, I would likely support that,” Gates said.
The difficulty, he said, is that multiple developers in both the U.S. and China are pushing closer to the technological frontier. Still, Gates rejected the argument that competition with China makes meaningful regulation impossible.
“China does not want cyberattacks, bioattacks or loss of control,” Gates said.
The discussion was punctuated by a pair of outbursts from the audience. At one point, as Gates discussed the economic incentives driving AI development, a man shouted that they represented “everything that’s wrong with America” and invoked Jesus Christ. Gates paused before responding dryly: “Anyway, not everything is wrong with America, according to me.”
Gates argued that government and civil society need to move much more quickly.
“This five-year period, in my view, is a very critical period,” he said, adding that responses that take five or six years to get underway could come too late.
Despite the dire warnings, Harris pointed to signs that pressure for safeguards is growing, including calls from AI-industry employees to limit the development of more powerful systems. Just last week, Meta agreed to pay $17 billion to settle claims by 29 states that Facebook and Instagram harmed young users while also agreeing to new child-safety measures on the platforms.
“I just want to leave you with not naive optimism, but momentum,” Harris said.
Near the end, Raskin suggested that President Trump’s well-documented desire to win the Nobel Peace Prize might provide an incentive for him to pursue a U.S.-China pause on more powerful AI systems.
“What’s better than solving one war?” Raskin said. “Saving all of humanity forever from uncontrollable AI.”
Gates, with a wry smile, called the Nobel angle “smart.”
“Whoever really takes the step to solve this AI problem would deserve the Nobel Prize,” he said.
Later that night, the festival would hold an outdoor screening of Stanley Kubrick’s “2001: A Space Odyssey,” featuring HAL 9000, one of cinema’s most famous rogue artificial intelligences.
Gates framed the current stakes in similarly cinematic terms.
“There are these movies where the aliens are coming, and you see unprecedented levels of cooperation between China and the U.S.,” he said. “Well, believe me, the aliens are here.”
TELLURIDE, Colo. — For four days every Labor Day weekend, the Telluride Film Festival turns this tiny mountain town into something of a refuge from the outside world. Tucked into a box canyon and far from the usual machinery of Hollywood, moviegoers spend their days seeing films they may have known little about before arriving and talking about them almost nonstop.
Artificial intelligence, apparently, did not get the memo.
This year, even Telluride’s usually self-contained movie bubble couldn’t keep out the existential anxieties surrounding AI. On Saturday, the festival devoted two events to the subject, beginning with “The Humanity Dilemma,” an hourlong multimedia presentation mixing dire warnings with live music and imagery, and continuing with a panel bluntly titled “AI AI AI,” featuring Microsoft co-founder Bill Gates.
Introducing “The Humanity Dilemma” at the nearly packed Sheridan Opera House, festival executive director Julie Huntsinger acknowledged that Telluride normally steers clear of taking political sides. “We just let the program speak for us,” Huntsinger said.
But on AI, she made an exception.
“We are a very humanist festival — that’s how we identify,” Huntsinger said. “AI is a threat.” While acknowledging potentially beneficial applications, she urged the audience to become more engaged in the debate over its development.
“We all need to be very alert and vigilant and loving and kind and stop this s—,” she said, earning loud whoops of approval from some in the crowd.
AI already occupies an uneasy place in Hollywood, where filmmakers and studios are experimenting with the technology even as actors, writers and other creative workers fight to protect their jobs and likenesses from being exploited or replaced.
For all of Huntsinger’s warnings, Telluride is not treating AI as off-limits. Its program includes “Love, Rendered,” a short documentary directed by Liz Garbus and produced by Darren Aronofsky about an elderly couple confronting the husband’s memory loss. Using AI along with family photographs and stories from loved ones, the filmmakers recreate the moment the couple first met and fell in love 70 years earlier. On Sunday, the film’s creative team is scheduled to discuss the project at an event titled “Filmmaking and Technology With Empathy: AI for Societal Benefit.”
“The Humanity Dilemma” was presented by Tristan Harris and Aza Raskin, co-founders of the Center for Humane Technology, who appeared in the 2020 documentary “The Social Dilemma,” along with artist and researcher Joy Mauthe and violinist Andrei Matorin. Harris and Raskin also appear in this year’s “The AI Doc: Or How I Became an Apocaloptimist.”
The presentation was designed to work on the emotions as well as the intellect. Original songs about humanity’s relationship to AI ran through much of the hour, with Mauthe singing and playing guitar and Matorin on violin, as Harris and Raskin delivered warnings accompanied by images of environmental devastation, poverty and figures in the AI race including Elon Musk, Sam Altman and Larry Ellison.
Their argument was stark: The race to build more powerful AI is moving far faster than governments or the public can keep up. Their deepest fear is that humans could eventually lose control of increasingly capable AI systems altogether.
At the same time, they acknowledged the technology’s potentially transformative benefits.
“AI is confusing because it represents both simultaneous utopia and dystopia,” Raskin said. “Just imagine having to reason about a nuke that could also solve cancer.”
By the end, Harris was calling for a halt to the development of more powerful systems.
“We need to pause frontier AI development and pivot and steer towards a pro-human future,” he said.
The hour concluded with Mauthe leading the audience in a final refrain, singing that the future is “still in our hands.” It was the kind of unabashedly earnest moment that could easily have tipped into awkwardness, but many in the room seemed to embrace it.
A few hours later, Harris and Raskin returned for “AI AI AI,” held outdoors at the Abel Gance Open Air Cinema in Telluride’s Elks Park, joined by Gates and filmmaker Joshua Oppenheimer, director of the Oscar-nominated documentaries “The Act of Killing” and “The Look of Silence,” who served as a moderator.
From left, Bill Gates, Aza Raskin, Tristan Harris and filmmaker Joshua Oppenheimer discuss the risks and potential benefits of artificial intelligence at the Telluride Film Festival on Saturday.
(Josh Rottenberg)
For Gates, who has spent decades focusing much of his philanthropy and public advocacy on global health and poverty, the rise of AI has forced him to make room for a new priority. In an essay published last week, he warned that the technology was improving faster than he anticipated and called for a new framework to manage its risks.
At Telluride, Gates said he now feels compelled to devote some of the political influence he has long used to advocate for causes such as malnutrition, polio and malaria to raise alarms about AI as well.
“Is this the greatest problem humanity has ever faced?” Gates said. “That is just a fact.”
Gates said his concerns have grown as AI systems have become more capable, particularly at writing computer code, something he has been obsessed with since he was 13.
“The AIs are superhuman, i.e., better than I am at writing code,” Gates said.
He laid out five broad areas of concern: jobs, biotechnology, cyberattacks, psychosocial harms and whether humans will be able to maintain control over the systems they are building.
Oppenheimer said that after spending hours digging into AI risks ahead of the panel, he had come away “absolutely terrified.” He read aloud an answer he had received after asking ChatGPT to estimate the risks if the race toward increasingly powerful AI continued with little regulation.
The chatbot put the chance of “persistent dystopian outcomes for all humanity” at 25% to 40%, a civilization-scale catastrophe at 5% to 15% and human extinction or permanent loss of human control at 5% to 10%.
“That’s what AI thinks,” Oppenheimer said.
Gates stressed AI’s potential upside, citing advances in medicine, education and assistance to farmers in poor countries. But pressed on whether development should be slowed until safety measures catch up, he said he would be open to the idea.
“If there was a credible plan that would cause this to be either slowed down or, you know, even held in stasis for a period of time, I would likely support that,” Gates said.
The difficulty, he said, is that multiple developers in both the U.S. and China are pushing closer to the technological frontier. Still, Gates rejected the argument that competition with China makes meaningful regulation impossible.
“China does not want cyberattacks, bioattacks or loss of control,” Gates said.
The discussion was punctuated by a pair of outbursts from the audience. At one point, as Gates discussed the economic incentives driving AI development, a man shouted that they represented “everything that’s wrong with America” and invoked Jesus Christ. Gates paused before responding dryly: “Anyway, not everything is wrong with America, according to me.”
Gates argued that government and civil society need to move much more quickly.
“This five-year period, in my view, is a very critical period,” he said, adding that responses that take five or six years to get underway could come too late.
Despite the dire warnings, Harris pointed to signs that pressure for safeguards is growing, including calls from AI-industry employees to limit the development of more powerful systems. Just last week, Meta agreed to pay $17 billion to settle claims by 29 states that Facebook and Instagram harmed young users while also agreeing to new child-safety measures on the platforms.
“I just want to leave you with not naive optimism, but momentum,” Harris said.
Near the end, Raskin suggested that President Trump’s well-documented desire to win the Nobel Peace Prize might provide an incentive for him to pursue a U.S.-China pause on more powerful AI systems.
“What’s better than solving one war?” Raskin said. “Saving all of humanity forever from uncontrollable AI.”
Gates, with a wry smile, called the Nobel angle “smart.”
“Whoever really takes the step to solve this AI problem would deserve the Nobel Prize,” he said.
Later that night, the festival would hold an outdoor screening of Stanley Kubrick’s “2001: A Space Odyssey,” featuring HAL 9000, one of cinema’s most famous rogue artificial intelligences.
Gates framed the current stakes in similarly cinematic terms.
“There are these movies where the aliens are coming, and you see unprecedented levels of cooperation between China and the U.S.,” he said. “Well, believe me, the aliens are here.”
TELLURIDE, Colo. — For four days every Labor Day weekend, the Telluride Film Festival turns this tiny mountain town into something of a refuge from the outside world. Tucked into a box canyon and far from the usual machinery of Hollywood, moviegoers spend their days seeing films they may have known little about before arriving and talking about them almost nonstop.
Artificial intelligence, apparently, did not get the memo.
This year, even Telluride’s usually self-contained movie bubble couldn’t keep out the existential anxieties surrounding AI. On Saturday, the festival devoted two events to the subject, beginning with “The Humanity Dilemma,” an hourlong multimedia presentation mixing dire warnings with live music and imagery, and continuing with a panel bluntly titled “AI AI AI,” featuring Microsoft co-founder Bill Gates.
Introducing “The Humanity Dilemma” at the nearly packed Sheridan Opera House, festival executive director Julie Huntsinger acknowledged that Telluride normally steers clear of taking political sides. “We just let the program speak for us,” Huntsinger said.
But on AI, she made an exception.
“We are a very humanist festival — that’s how we identify,” Huntsinger said. “AI is a threat.” While acknowledging potentially beneficial applications, she urged the audience to become more engaged in the debate over its development.
“We all need to be very alert and vigilant and loving and kind and stop this s—,” she said, earning loud whoops of approval from some in the crowd.
AI already occupies an uneasy place in Hollywood, where filmmakers and studios are experimenting with the technology even as actors, writers and other creative workers fight to protect their jobs and likenesses from being exploited or replaced.
For all of Huntsinger’s warnings, Telluride is not treating AI as off-limits. Its program includes “Love, Rendered,” a short documentary directed by Liz Garbus and produced by Darren Aronofsky about an elderly couple confronting the husband’s memory loss. Using AI along with family photographs and stories from loved ones, the filmmakers recreate the moment the couple first met and fell in love 70 years earlier. On Sunday, the film’s creative team is scheduled to discuss the project at an event titled “Filmmaking and Technology With Empathy: AI for Societal Benefit.”
“The Humanity Dilemma” was presented by Tristan Harris and Aza Raskin, co-founders of the Center for Humane Technology, who appeared in the 2020 documentary “The Social Dilemma,” along with artist and researcher Joy Mauthe and violinist Andrei Matorin. Harris and Raskin also appear in this year’s “The AI Doc: Or How I Became an Apocaloptimist.”
The presentation was designed to work on the emotions as well as the intellect. Original songs about humanity’s relationship to AI ran through much of the hour, with Mauthe singing and playing guitar and Matorin on violin, as Harris and Raskin delivered warnings accompanied by images of environmental devastation, poverty and figures in the AI race including Elon Musk, Sam Altman and Larry Ellison.
Their argument was stark: The race to build more powerful AI is moving far faster than governments or the public can keep up. Their deepest fear is that humans could eventually lose control of increasingly capable AI systems altogether.
At the same time, they acknowledged the technology’s potentially transformative benefits.
“AI is confusing because it represents both simultaneous utopia and dystopia,” Raskin said. “Just imagine having to reason about a nuke that could also solve cancer.”
By the end, Harris was calling for a halt to the development of more powerful systems.
“We need to pause frontier AI development and pivot and steer towards a pro-human future,” he said.
The hour concluded with Mauthe leading the audience in a final refrain, singing that the future is “still in our hands.” It was the kind of unabashedly earnest moment that could easily have tipped into awkwardness, but many in the room seemed to embrace it.
A few hours later, Harris and Raskin returned for “AI AI AI,” held outdoors at the Abel Gance Open Air Cinema in Telluride’s Elks Park, joined by Gates and filmmaker Joshua Oppenheimer, director of the Oscar-nominated documentaries “The Act of Killing” and “The Look of Silence,” who served as a moderator.
From left, Bill Gates, Aza Raskin, Tristan Harris and filmmaker Joshua Oppenheimer discuss the risks and potential benefits of artificial intelligence at the Telluride Film Festival on Saturday.
(Josh Rottenberg)
For Gates, who has spent decades focusing much of his philanthropy and public advocacy on global health and poverty, the rise of AI has forced him to make room for a new priority. In an essay published last week, he warned that the technology was improving faster than he anticipated and called for a new framework to manage its risks.
At Telluride, Gates said he now feels compelled to devote some of the political influence he has long used to advocate for causes such as malnutrition, polio and malaria to raise alarms about AI as well.
“Is this the greatest problem humanity has ever faced?” Gates said. “That is just a fact.”
Gates said his concerns have grown as AI systems have become more capable, particularly at writing computer code, something he has been obsessed with since he was 13.
“The AIs are superhuman, i.e., better than I am at writing code,” Gates said.
He laid out five broad areas of concern: jobs, biotechnology, cyberattacks, psychosocial harms and whether humans will be able to maintain control over the systems they are building.
Oppenheimer said that after spending hours digging into AI risks ahead of the panel, he had come away “absolutely terrified.” He read aloud an answer he had received after asking ChatGPT to estimate the risks if the race toward increasingly powerful AI continued with little regulation.
The chatbot put the chance of “persistent dystopian outcomes for all humanity” at 25% to 40%, a civilization-scale catastrophe at 5% to 15% and human extinction or permanent loss of human control at 5% to 10%.
“That’s what AI thinks,” Oppenheimer said.
Gates stressed AI’s potential upside, citing advances in medicine, education and assistance to farmers in poor countries. But pressed on whether development should be slowed until safety measures catch up, he said he would be open to the idea.
“If there was a credible plan that would cause this to be either slowed down or, you know, even held in stasis for a period of time, I would likely support that,” Gates said.
The difficulty, he said, is that multiple developers in both the U.S. and China are pushing closer to the technological frontier. Still, Gates rejected the argument that competition with China makes meaningful regulation impossible.
“China does not want cyberattacks, bioattacks or loss of control,” Gates said.
The discussion was punctuated by a pair of outbursts from the audience. At one point, as Gates discussed the economic incentives driving AI development, a man shouted that they represented “everything that’s wrong with America” and invoked Jesus Christ. Gates paused before responding dryly: “Anyway, not everything is wrong with America, according to me.”
Gates argued that government and civil society need to move much more quickly.
“This five-year period, in my view, is a very critical period,” he said, adding that responses that take five or six years to get underway could come too late.
Despite the dire warnings, Harris pointed to signs that pressure for safeguards is growing, including calls from AI-industry employees to limit the development of more powerful systems. Just last week, Meta agreed to pay $17 billion to settle claims by 29 states that Facebook and Instagram harmed young users while also agreeing to new child-safety measures on the platforms.
“I just want to leave you with not naive optimism, but momentum,” Harris said.
Near the end, Raskin suggested that President Trump’s well-documented desire to win the Nobel Peace Prize might provide an incentive for him to pursue a U.S.-China pause on more powerful AI systems.
“What’s better than solving one war?” Raskin said. “Saving all of humanity forever from uncontrollable AI.”
Gates, with a wry smile, called the Nobel angle “smart.”
“Whoever really takes the step to solve this AI problem would deserve the Nobel Prize,” he said.
Later that night, the festival would hold an outdoor screening of Stanley Kubrick’s “2001: A Space Odyssey,” featuring HAL 9000, one of cinema’s most famous rogue artificial intelligences.
Gates framed the current stakes in similarly cinematic terms.
“There are these movies where the aliens are coming, and you see unprecedented levels of cooperation between China and the U.S.,” he said. “Well, believe me, the aliens are here.”
