OpenAI CEO Sam Altman and COO Brad Lightcap were the guests on a recent live episode of the New York Times' technology podcast, Hard Fork, and entered the San Francisco venue like a pair of WWE wrestlers.
I'm not kidding: hosts Kevin Roose and Casey Newton began their introduction before Altman and Lightcap strode out early to take their seats. Roose told the crowd they'd been planning to list some recent NYT headlines about OpenAI to set the scene, and after Altman awkwardly insisted "this is more fun that we're out here for this" the chest-beating began.
"Are you going to talk about where you sue us because you don't like user privacy" Altman demanded, as if he were delivering some great zinger, which had the hosts and audience laughing and saying "woo!" like Ric Flair. Altman had clearly arrived in a bit of a mood about the NYT's ongoing lawsuit against OpenAI and Microsoft (the company's largest investor), and the fact that neither of the hosts is actually involved with that suit was immaterial.
The NYT's lawsuit against OpenAI alleges that the AI company used the traditional media company's articles in the training of its large language models. Altman's spikiness relates to a new development in which the NYT's lawyers asked that OpenAI be compelled to retain consumer ChatGPT data and API customer data.
"The New York Times, one of the great institutions for a long time, is taking a position that we should have to preserve our users' logs even if they’re chatting in private mode, even if they’ve asked us to delete them," said Altman. "And the lawsuit we're having to fight out, but that thing, we think privacy and AI is an extremely important concept to get right for the future, still love you guys, still love The New York Times, but that one we feel strongly about."
"Well thank you for your views and I'll just say it must be really hard when someone does something with your data you don't want them to," responds Roose to laughs from the entire room. "I don't know what that's like personally but maybe someone else does."
"I was recently told by another guest on this stage that the singularity would be 'gentle'," adds co-host Newton to further laughs.
And they're right of course… the dude whose technology is built on scraping as much content as possible is now an advocate for privacy. Altman tried to get the hosts involved in commenting on the lawsuit but they weren't having any of it. "I think people should read the relevant filings and make up their own minds," said Roose as Altman continually asked about the suit.
Things somewhat simmered down after this, with the hosts asking about "that rascal Mark Zuckerberg" poaching OpenAI's employees in recent months, as Meta amps up its own investment in AI and the Zuck's belief in superintelligence. Asked if they think Zuckerberg's belief in such technology is sincere, Lightcap said simply:
"I think [Zuckerberg] believes he is superintelligent."
The rest of the interview sees the OpenAI execs go over some familiar ground, with Altman making his usual overblown claims about everyone now having a "PhD level intelligence" in their pocket. The reported tensions with Microsoft are hand-waved away by Altman as "points of tension" in a relationship that brings "deep value" to both.
Towards the end of the chat, Newton asks the pair about recent stories of mentally unwell people using ChatGPT, whether that's users who think they've connected with god, people going into the weeds on conspiracy theories, or individuals with suicidal thoughts.
"If conversations are going down a rabbit hole in this way we try to cut them off or suggest something different to the user," begins Altman, adding that ChatGPT will suggest professional services as an alternative. “We don’t want to slide into the mistakes that I think the previous generation of tech companies made by not reacting quickly enough to a new thing, a psychological reaction."
Asked whether ChatGPT should carry a warning to the effect that it is not in fact god, Altman says "the model will tell you things like that, then users will write us and say 'you changed this.'" But after the hosts follow-up, he does admit that "to users that are in a fragile enough mental place, that are on the edge of a psychotic break, we haven’t yet figured out how a warning gets through there."
Lightcap then tries to spin towards the positives, claiming ChatGPT has "rehabilitated marriages" and for some people "it's the first time in their life where they've had something they can confide in… and it doesn't cost $1000 an hour."
Then, a very relatable anecdote. "I was surfing in Costa Rica the other day," says Lightcap, when he got chatting to a local and mentioned he worked at OpenAI. "And he started crying… [he said] 'ChaptGPT saved my marriage. I didn't know how to talk to my wife, it gave me tips to talk to my wife and I've learned that and we're on a much better path.' It sounds like a dumb and stupid story but it's not, I was there."
Press X to doubt. "That's great, we're back to even," cracks Roose, "because a chatbot tried to break up my marriage."
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
