How AI Will Change Photography Forever

**Ren Ng** (0:00)
The critical moment, the decisive moment in photography, which is this thing that as photographers, we seek to capture, that moment in space and time is so fleeting.

**Cecilia Zhang** (0:11)
Moment is no longer fixed when the shutter is pressed. It's a very interesting new dimension of what photography is.

**Ren Ng** (0:19)
If you believe that premise, which I deeply do, then we are at a land shift moment for imaging a photography.

**SPEAKER_3** (0:26)
What if you could retake your favorite memories, fix the lighting, catch the smile, or even open your eyes years after the moment passed? In this episode, a16z general partner, Martin Casado and partner Yoko Li sit down with scientists and Lytro founder Ren Ng plus Phota Labs co-founders Cecilia Zhang and Zach Xia to trace the wild arc of computational photography. From Stanford camera arrays and light field refocus, to today's personalized generative retakes. Together, they unpack why identity and context matter more than filters and ask whether the future of photography even needs a lens. Let's get into it.

**Yoko Li** (1:05)
I'm super excited to have the group here together because we share a lot of personal history and then chatting about evolution of computational photography. One fun fact is Cecilia and I, we went to college together and we actually know each other for more than 15 years. I kind of watched her working first day of college. I don't know if you remembered, you kind of carried a camera and walked in. It's just so great to see you turning that into a top-notch research career to Adobe and where you met Zach. Another connection here is Cecilia did her PhD was Ren, who was the founder and CEO of Lytro camera, which is a light field camera, very revolutionary. Another fun fact is Ren and Martin, 20 years ago, 15 years ago, maybe 20 years ago.

**Ren Ng** (1:56)
20 plus.

**Yoko Li** (1:59)
More than 20 years ago. We're actually grad school classmates, and we did grad school together.

**Martin Casado** (2:06)
We were actually portfolio founders in the first fund of Andreessen Horowitz, and I was his reference.

**Yoko Li** (2:11)
You were his reference.

**Martin Casado** (2:14)
I was like, do you know this Ren guy?

**Ren Ng** (2:15)
I'm like, she's a genius.

**Yoko Li** (2:18)
Yeah.

**Ren Ng** (2:19)
The best entrepreneur I ever knew. He said something nice about me, so Ben decided to take a look.

**Yoko Li** (2:26)
Yeah. Maybe the right place to start is turning back to the history, thinking about historically, how did computational photography as a field evolve? How did you come up with Lytro, the story behind us? Funding Lytro too, and I'll let you start, Ren.

**Ren Ng** (2:41)
Exciting topic, exciting times, I think, for computational photography, photography at large. And I'd say that, I think computational photography has been in the mix since the beginning of digital photography, right? I mean, you can't have a digital picture without doing an immense amount of processing and very basic things or things we would think are very basic, like making it make the right colors, involves some of the most studied algorithms early on in computational imaging. And so I'd say there's been a ton of computational imaging. Where I arrived on the scene was that I was really into photography. Like Cecilia, I carry a camera. Actually, I got into photography because all my best friends were rock climbers. And one of them was really into photography, and he said, oh, you should try my camera. And so we climbed up the top of a route and then tried to take pictures of the next person climbing up and make ourselves look cool, you know? So that was my introduction to photography. And I got hooked. And then I decided for my PhD that the stuff I was working on, which I thought was great, was not going to go where I wanted it to go. And I was like, I want to do photography stuff. So I tried to marry that stuff, which is computer graphics, with photography. And so for me, it was really an attempt to bring more computation into photography by bringing all of graphics, how we make pictures with the computer, into the core of the question of how we're going to do photographic imaging. And that was the start for life in photography.

**Yoko Li** (4:07)
And Martin, you were there when we made the investment to Lytro. And what was it like for computational photography back then?

**Martin Casado** (4:15)
So I'm a systems guy, and so here's what I remember. What I remember is you'd go into Gates, and I think it was Marc Lavoie had this camera, which is not one camera, but a whole bunch of cameras.