As access to advanced neural-network capabilities broadens, we’re seeing it used by the amateur community in new and novel ways — from $200 self-driving R/C cars to cat-recognizing laser spotter toys. And now, a new group is training their machines with images of celebrated actor Nicholas Cage and inserting them onto the faces of iconic characters in film like James Bond and Indiana Jones, replicating the movement and emotion .

The output isn’t perfectly seamless yet, but is surprising in how effective it is for just one person to do at home, which leads to a lot of questions about what happens as technology and access advances just a bit further. It also comes from an unsavory initiative, which taints the fun of these vids a bit.

Make’s own Jon Christian reports about the trend on the Outline — check out the article for the story and more Cagey clips.