������Ƶ

I went to an alumni reunion of the JSK Journalism Fellowships at Stanford University this past week and returned home humbled by my ignorance about the sweeping impacts of artificial intelligence.

Discussions about AI — computer systems that can do tasks and solve problems like a human — were popular at the gathering.

Computer-driven helpers have been with us for years including mobile banking systems, voice assistants like Alexa and Siri, and clumsy new self-driving cars. But now AI output is smarter, more human-like.

A new wave called generative AI is a set of algorithms working with a large data set to generate its own original content including stories, texts, paintings and music compositions. Its output is often indistinguishable from a human’s writing or a human’s speaking voice. It also can make realistic human images.

A darker capability is generative AIʻs ability to spread disinformation.

ChatGPT launched by OpenAI in November is a generative AI system that has sparked more enthusiasm from the public than any artificial intelligence that’s come before.

“It hit us with the power of an atomic bomb of interest. People are writing articles about it everywhere. For the first time with ChatGPT a robot can talk back to people in a way that is mind blowing. It can do things only a human being could do in the past such as writing poems, summarizing text, creating information for different age levels and (painting) its own beautiful original artworks. Computer scientists didn’t expect this to happen for decades,” says Jason Leigh.

Leigh is the co-founder of and a professor of computer science at the University of Hawaii Manoa.

I agree with him about the bewitching power of ChatGPT. I began playing with it in April on a trip to Australia, asking it to write haikus and heroic poems about my friends and relatives. After being enchanted by the web-based tool, I was eager to find out more during the JSK reunion about AI’s benefits and perils.

Wall Street Journal about a similar awakening her friends were experiencing this summer: “At almost every gathering artificial intelligence came up. I’d say people are approaching AI with a free floating dread leavened by a pragmatic commitment to make the best of it, see what it can do to make life better. It canʻt be stopped any more than you can stop the tide. There’s a sense of, ‘It may break cancerʻs deepest codes,’ combined with ‘It may turn on us and get us nuked.’”

The keynote presentation at the Stanford reunion by Jeff Hancock and Natasha Zouves was titled “Generative AI. Trust and Deception.” Hancock offered a generally optimistic view about the ability of people to utilize, control and make AI systems operate ethically.

Yet, a key concern is how to retain trust in fact-driven, accurate information from humans when today’s more sophisticated AI systems — capable of spinning disinformation — can generate content that is indistinguishable from work done by people.

Hancock says mistrust is common whenever new technology comes along, yet over time people have been able to adapt and manage it.

“AI is a tool created by humans. AI is not a person. We have to remind ourselves we are the agents. We are creating the agency. We have to stay in the agent role,” says Hancock.

But all the needed guardrails to prevent abuses are not created yet. Maybe they never will be.

Hancock is the and the co-director of the university’s Cyber Policy Center as well as a professor in Stanford’s Department of Communications.

Andrew Finlayson, who led a roundtable discussion after the keynote presentation, had a more apocalyptical vision.

“Generative AI will destroy elections,” he said. He forsees AI generated images of Joe Biden and Donald Trump saying things we have never heard either of them utter before. Yet they will seem convincing.

“You will be hard-pressed to tell the difference between a human being and a non-human being. AI is moving faster than we can keep up with it,” he said.

Finlayson is a and executive VP for digital media strategy at SmithGeiger Group.

The actors and screenwriters — — are worried about the impact of generative AI’s ability to destroy their careers as it harvests their writing and personal images without their permission to feed content into data sets to make supposedly “original” content. The actors don’t want their images showing up in AI created films they never worked in.

I suspect the public’s recent fascination with computer-generated intelligence has been heightened by the Hollywood strikes. When you have a union to which Meryl Streep and Jennifer Lawrence belong and support, with celebrities marching on picket lines over grievances including the threat of AI to derail their livelihoods, people want to know more about it.

At the San Francisco International Airport on the way home to Honolulu, I bought a copy of “The Age of AI and Our Human Future” by Henry Kissinger, Eric Schmidt and Daniel Huttenlocher. When I paid for the book, the clerk told me, “Everyone is buying this.”

The book was copyrighted in 2021, which makes it already old as AI innovations move exponentially fast, but it offers an understanding of the language of AI and a history of how we got here.

“AI will lead human beings to realms that we cannot reach solely by human reason, now or perhaps ever.  Its technical achievements in health and ecomonics promise to make the age of AI an age of abundance. While we celebrate that potential, we recognize that a new reality is emerging. As the stakes rise, our response must meet them,” the authors write.

When I went to my fellowship at Stanford in 1969, I brought along my manual portable typewriter. It would be decades before the proliferation of personal computers, the internet and mobile phones. Now, a half century later, it is both thrilling and frightening to be on the edge of a technological innovation more powerful than any before it that could eventually surpass the capabilities of the human brain.