Ethan He

speaker

722 appearances 1 recordings 1 series first heard Jun 2026 last heard 1 Jun

Ethan He’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

Recordings per month over the last 12 months — 1 in all, peaking in Jun 2026 with 1.

Appearances

newest first · ▶ plays the moment

Latent Space: The AI Engineer Podcast · Why Video Agent models are next — Ethan He, xAI Grok Imagine · 1 Jun 2026 podcast

Yeah, actually there's a Grok Imagine agent beta.

That was the first attempt in that direction.

Yeah.

So I think the process would be similar.

It's just agent mode.

Yeah, you can ask it to maybe generate a one-minute video, which is not possible if you ask the same prompt to video models.

But this model will literally call different tools to do that.

So, yeah, this is actually an interesting thing.

So when they first released a video editing model, I see on some people try the video editing feature.

It's like, edit this video to be one minute.

Because they didn't understand how video editing works.

Video editing typically is just a removal, add, replace, style transfer, this kind of thing.

But that's actually a valid request under the assumption of video agents.

So these agents should be able to understand these kind of long-horizon tasks to be able to iteratively create a long-form video.

I think this is really fascinating because it's kind of taking the same direction as first you have these AI-assisted coding, kind of like tab completion, GitHub Copilot.

And from there, you gradually evolve to Codex and Cloud Code, where you do things fully automated.

So in Garki Imagine Asian mode, you can still go in there and do stuff by yourself.

Gradually, as the model capability increases, it will be able to do everything fully automated.

Yeah, there, there are a bunch of things like we talk about distillation and if you talk about thinking, if you don't have any thinking budget, the model can just think three minutes and then come back to you.

And also like inference, the inference infra team was very talented and they were, they were able to accelerate a hell lot of these models.

Showing 601–620 of 722 · page 31 of 37