Oct 6, 2026 · 45 min · 10 segments
As Open AI apologises to Australia the CEO of Sovereign AI takes us through the relative merits of open weight, open source and closed AI models. Regular panellists Digital Rights Watch chair…
Simon KrissGuest
Dan StintonPanelistPeter LewisHostLizzie O'SheaPanelist[laughs] Lizzie, beyond the spectacle of the OpenAI mea culpa today, we've had Sam Altman telling Politico this week that we need to accept the bad of AI to get to the good.
Um, the reason why I wanted to raise it is it just seems a prime example of us importing sophisticated technology where it need not be.
So, um, Deep Cut News were reporting on this, that there was a proposal for a, a basketball association to start using, um, technology to record games, and supposedly for performance analytics.
And, like, these are kids, and a bunch of people raised concerns about it, and they said, "No, no, we've got security measures in place.
Only certain people can access it." They're working with an Israeli tech firm to do it.
And then, of course, almost, I think, a week or a little bit later, um, someone has complained about being identified by a former partner, uh, and receiving messages.
But basically, the exact problems that people had raised concerns about coming to pass.
And it just really brought home for me, as we're having these discussions around what good comes from these sophisticated forms of technology and how we might have to suck up the bad in order to get that good, that what also seems to be the default phenomenon is we are having to accept technology in all these different spaces for some s- often quite marginal benefit, and then downside arrives in quite a material obvious way, and that is just accepted now as part of life.
Uh, you watch these inquiries today, and you see Sam Altman saying elsewhere in the media that we just have to accept the bad 'cause there'll be good that comes with it, and that's why we need light touch regulation.
Um, one of the things that really struck me, apart from the fact that the so-called hauling was very, very friendly, um, he was the guy who'd flown out from the US, was given a chance to apologize, and then promised that he'd use better AI to find the breaches in future.
But Dan, one of the things that stood out was that OpenAI was given the opportunity to articulate the positive use case.
It was almost like this Dorothy Dixer, "Give me your 30-second elevator pitch." And he really struggled to go beyond the old tropes of, "Our, a researcher has discovered his dog's cancer." It was really revealing that these guys that are working really deep in the tech can't really tell you what it's good for.
And I feel, I don't know, like the missing part of the story, if there is a story here, is how do we make it less curing your dog's cancer or stalking you while you're playing basketball and something useful.

So you want me to come up with the, uh, the, the justification for the technology? [laughs]
[laughs] Lizzie, beyond the spectacle of the OpenAI mea culpa today, we've had Sam Altman telling Politico this week that we need to accept the bad of AI to get to the good.
Um, the reason why I wanted to raise it is it just seems a prime example of us importing sophisticated technology where it need not be.
So, um, Deep Cut News were reporting on this, that there was a proposal for a, a basketball association to start using, um, technology to record games, and supposedly for performance analytics.
And, like, these are kids, and a bunch of people raised concerns about it, and they said, "No, no, we've got security measures in place.
Only certain people can access it." They're working with an Israeli tech firm to do it.
And then, of course, almost, I think, a week or a little bit later, um, someone has complained about being identified by a former partner, uh, and receiving messages.
But basically, the exact problems that people had raised concerns about coming to pass.
And it just really brought home for me, as we're having these discussions around what good comes from these sophisticated forms of technology and how we might have to suck up the bad in order to get that good, that what also seems to be the default phenomenon is we are having to accept technology in all these different spaces for some s- often quite marginal benefit, and then downside arrives in quite a material obvious way, and that is just accepted now as part of life.
Uh, you watch these inquiries today, and you see Sam Altman saying elsewhere in the media that we just have to accept the bad 'cause there'll be good that comes with it, and that's why we need light touch regulation.
Um, one of the things that really struck me, apart from the fact that the so-called hauling was very, very friendly, um, he was the guy who'd flown out from the US, was given a chance to apologize, and then promised that he'd use better AI to find the breaches in future.
But Dan, one of the things that stood out was that OpenAI was given the opportunity to articulate the positive use case.
It was almost like this Dorothy Dixer, "Give me your 30-second elevator pitch." And he really struggled to go beyond the old tropes of, "Our, a researcher has discovered his dog's cancer." It was really revealing that these guys that are working really deep in the tech can't really tell you what it's good for.
And I feel, I don't know, like the missing part of the story, if there is a story here, is how do we make it less curing your dog's cancer or stalking you while you're playing basketball and something useful.

So you want me to come up with the, uh, the, the justification for the technology? [laughs]
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.