Emad Mostaque

speaker

1,351 appearances 1 recordings 1 series first heard Dec 2024 last heard Dec 2024

Emad Mostaque’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

No recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.

newest first · ▶ plays the moment

Then anyone can take those because we release everything MIT permissionlessly, which is the other important thing.

Like you don't what if someone you break the

Rules.

And they shut off your AI.

You're a bit screwed then, right?

Like for whatever reason.

And then I think again, that can move towards standards because the highest value AIs will be decision supporting and making AIs.

And what do those look like and how do we understand what those do?

And are those got repeatable f frameworks?

The most widely used AIs will be the AIs that are available, especially if they're permissionless.

I mean, we have three hundred million downloads of our models at Stability AI, you know?

And that's a lot of freaking downloads because they're just available.

When the people built around them, 'cause they didn't need commission.

So I think that again it's necessary but not sufficient because

Like otherwise they look at anthropic and open their eyes models, like how will they incorporate conceptualizations of

Chinese or Japanese ethics.

Is anyone even taking that seriously?

But then what if they use to make decisions in the Japanese government?

as they kind of are right now.

Like I think you need to be very deliberate in the way that you figure this out on the underlying principles and the feedback loops there.

Showing 701–720 of 1,351 · page 36 of 68