This is the first time I have written about the AI craze that has swept the web over the past 6 months or so. I’ve been intentionally resisting the urge given so many of the hot takes, hysteria, and complaints seem pointless. The only way to start wrapping your head around this post-humanistic beast that cannot die is through art, and so it happened when I was chatting with Michael Branson Smith (MBS) about what he might present at Reclaim Open. We had a wide-ranging chat, as always, and he started talking about using AI to generate a “fake” presentation, all he would need is someone who has written about a topic extensively and we can get access to samples of both their video and audio likeness. I think at that point MBS said something to the effect of, “You know, someone like Alan and his SPLOTs.” BINGO!
The rest is kind of history, I was thrilled to play some small role as a sounding board for MBS’s epic vision. In fact, I almost torpedoed it by suggesting we do it entirely without letting Alan know beforehand. I was fascinated by what his reaction would be after discovering his likeness and ideas were used to present at Reclaim Open without any prior head’s up. I mean it would give him another reason to rail against AI, right? But this was not so much because I wanted to piss him off, but because I think it would underline the blurry ethical lines where open, available content and credit get deeply complicated in the realm of AI generated content. That said, I also recognized that making a statement about AI and ethics was less important than offending a friend, so we looped Alan in, and he provided video and audio for MBS to train for A.I. Levine’s 5-minute talk on SPLOTs.*
The session was amazing, and you can watch it above in all its glorious hilarity. At the last minute we had the idea to loop Alan into the presentation with the idea of him acting irate and faking indignation at the whole thing. Alan, as expected, was even better than his A.I. likeness, and it was really a joyful, thoughtful session that underscored the serious implications of this new technology without being pedantic or sensational. It was also amazing to have Alan play such a powerful role at Reclaim Open from afar and, of course, he already blogged it.
This session also helped me understand the best way, at least for me, to come to terms with the artificial elephant on the web, namely to think of it as a creative challenge to explore the limits and possibilities of this tech. This spawned the idea of shopping around a new iteration of ds106 that’s centered on AI, which a subsequent session that put ds106 on trial cemented, but more on that in another post.
____________________________
*According to MBS, the audio and video he found elsewhere were even more effective, and I am hoping MBS will blog a post-mortem in which he narrrates his process, because it was pretty amazing.
The response of the audience (via the sound recorded) was the sign, the tell. At first it was kind of raucous laughing and fun and enjoyment. Then it got a little more nervous the more people realized, how much had be generated from the training. It was fun, and funny, and thought provoking.
Yeah, it was definitely an awesome session because it was fun, but also deeply concerning all at once. I appreciate Cogdog playing along like a champ, and anything MBS imagines is good enough for me.
Pingback: Accused! ds106 on Trial | bavatuesdays
Pingback: Rise of the Machines | bavatuesdays
Pingback: AI106: Long Live the New Flesh | bavatuesdays