June 14, 2023/ / ARTICLES/ Comments: 17

There’s been a lot of talk surrounding AI recently, but I honestly haven’t paid too much attention to all the fuss. Beyond watching Ex-Machina and occasionally pondering how robots will inevitably take over the world, AI technology has been pretty absent in my thought-rolodex.

At least that was the case until a few weeks ago.

Adobe recently released its “Generative Fill” feature in the latest Photoshop update, which can convincingly fill in empty space in photos, drawings, or whatever type of image you feed it. The heated debate of “AI art” seems to be the latest divisive fad among internet nerds right now, so I decided to take a crack at it myself and see how far I could push this futuristic feature.

Being that many of skating’s most iconic photos are in a portrait orientation (often with text overlayed on top for covers, ads, or whatever else), it seemed like a perfect opportunity to see what sort of landscape-oriented alternate universes AI could dream up. Although these weren’t generated by just one simple click (like most videos out there will show) this experiment has shown that it didn’t take very much effort to fabricate a passable reality.

*DUMBASS DISCLAIMER* While these AI edits seem somewhat passable, we know there is no replacement for actual human skateboarding photographers. Any disgruntled photographers can direct their quarrels in a strongly worded e-mail to [email protected]

Jeremy Wray, Water Tower Gap Ollie, 1997. Photo by: Daniel Sturt

Simon Woodstock for Big Brother Issue 14, 1995. Photo by: Mike Ballard

Peter Bici, Backside Ollie, 1997. Photo by: Dimitry Elyashkevich

Jamie Thomas, The Leap of Faith, 1997. Photo by: J. Grant Brittain

Shiloh Greathouse for Big Brother Issue 1, 1992. Photo by: Spike Jonze

Phil Shao, Front 5-0 at Fort Miley, 1996. Photo by: Luke Ogden

Brian Anderson, FS Blunt at Hubba Hideout, 1996. Photo by: Bryce Kanights

Eric Koston, Pizza Ollie, 2002. Photo by: Mike O’Meally

Daniel Castillo for Big Brother Issue 2, 1992. Photo by: Spike Jonze

Mark Gonzales at Alcatraz, 1988. Photo by: Bryce Kanights

Related Posts


  1. Justin Figueroa

    June 14, 2023 1:48 pm

    Obviously AI has never been to Hubba Hideout

    • Samus Aran

      June 14, 2023 4:40 pm

      Hey, speaking of Hubba Hideout – this is totally off-topic, but who cares! You know in Fulfill the Dream, where Steve Olson does… whatever he does… back tail shove? And then he jumps over the ledge, and then they show the dummy? How high was that drop in real life? And if you’re the real Figgy, big up. I’m a fan!!

  2. Franky O'Penis

    June 14, 2023 4:59 pm

    Finally, someone does something cool with AI.

    • 100% talkboarder

      June 16, 2023 10:14 am

      Right, all the pearl grabbing is unnecessary. If it’s used as a tool and doesn’t replace the three chip minolta or Kodak camera, its just better for the reminders.

  3. ??

    June 15, 2023 10:38 am

    what was the point of this?

  4. Alex

    June 15, 2023 10:43 am

    I find it interesting how the AI was able to correctly continue the leap of faith handrail. It must have some “knowledge” about the location, probably from dozens of other photos. Otherwise the result couldn’t be so similar to the actual thing.

    • Derrr

      June 21, 2023 9:53 am

      I mean it’s a stair rail and stairs not that hard to assume it continues down to the ground.

    • Andy

      June 23, 2023 4:37 am

      I think a lot of the AI programs “learn” by harvesting data/images etc from the ‘net. So this theory mages a lot of sense

    • Link

      June 30, 2023 11:56 am

      AI simply “repeats” the image. Look at the stairs on the LOF behind the side-wall. Its a jumbled mess where you can’t see the status within the perforation.

      Also, look at the cloud on the top right off the hubba hideout photo. Makes a small cloud into s big cloud

Leave a comment