Welcome to episode 151 of The Digital Life, a show about our adventures in the world of design and technology. I’m your host, Jon Follett, and with me is founder and co-host Dirk Knemeyer.
Yes, indeed it is. On this happy day we are going to discuss yet another area of humanity’s output that can be taken over by artificial intelligence.
Yes. We are going to talk a little bit today about The Next Rembrandt project, which was actually put together by an advertising film, if you can believe it, plus some very smart technologists who analyzed a host of Rembrandt’s paintings, and then based on that, allowed an artificial intelligence to pick a particular subject that Rembrandt might have painted. In this case, a Caucasian male with a goatee, very styling, and actually have the AI go ahead and paint this painting using a 3D printer with translucent paints in the style of Rembrandt.
If you go online you can see a video about this. Additionally you can see various critiques of the work. There will be some grand unveiling at a certain point in time down the road. The critiques of ranged from, “Wow. This is amazing and it looks just like a Rembrandt painting,” to, “Oh yeah, well if Rembrandt ever did this it would be a really mediocre Rembrandt painting” Nonetheless, it seems like if you want an original Rembrandt these days, all you need is a super computer and a 3D printer and voila, you have something that the master would have probably never done for you personally. If you’re an advertising agency with some bucks to spend, and some technologists on board, well, then you can resurrect the dead artist and have them create something at your bidding.
Which is entirely disturbing to me for a couple of reasons, which is that this derivative work, which it is very derivative, is also … I wouldn’t say it’s devalued Rembrandt’s artwork, but really, I think the idea that the AI is developed enough to create a mockery painting or a “close enough but no cigar” painting, it just kind of rubs me the wrong way.
I think it’s mainly the idea that … You know, when you watch a great movie versus a B movie, right?
B movies are great because they’re B movies and they’re either not plotted well or the dialogue is bad or something. You can appreciate them for what they are. This is really a B movie Rembrandt posing as a technological wonder, right? I don’t think it is quite … It’s an interesting experiment, but I don’t think it rises quite to the level of artwork.
Human artists have been doing B versions of Rembrandt style art for centuries. Do those bother you similarly, or is it the fact that it’s a machine?
No, because I think the B grade Rembrandt paintings enhance the reputation of Rembrandt as being the master, right? You know that you’re looking at a B grade painting. If it were the AI trying to approximate a so-so artist who is imitating Rembrandt, then I’d be okay with it. Maybe I’m just a little more disappointed with the hype surrounding it than the actual accomplishment, which is quite incredible.
Is it incredible or not? What I’m hearing you say is what you are … I don’t know, upset might be the wrong connotation. What you’re upset about is the fact that the media has said this is such a big deal, and you think it’s just sort of a crappy version.
You think it’s a really impressive thing at the same time.
I think it’s impressive technology, but I don’t think it approximates the human artist by a long shot.
There are many human artist that might approximate —
Yes, the masterful human artist. A few weeks ago we were talking about … Which is not to say that it’s not coming, which I guess is another nagging indicator in the back of my head.
That very soon a … You know, don’t even bother to try and do artwork because the AI is going to do it better than you. We talked about the AI mastering, or at least getting very good at the Asian board game of Go, which was thought to be at least a very difficult feat for an AI to take.
I think mastering is fair. I mean, If you say the top [inaudible 00:05:42] players in the world have mastered the game, this AI has mastered the game.
Right. Now when we see something on the art side, I guess I’ll say that in comparison perhaps this is not the Go level artistry, but it’s getting close, right? It’s on its way.
Yeah, five years ago they thought that beating a master in Go would take decades, plural, and it’s already done. I looked at the art done by the Rembrandt AI, for lack of a better way to put it, and I actually thought it was pretty good. Was it Rembrandt quality? No, but, I don’t know, I thought it was pretty good. I bet they’re not that far from basically being able to do it at a level comparable-ish with Rembrandt.
How does that make you feel, Jon?
Yeah, I don’t know how that makes me feel. I’ll have to come back on the show in a couple weeks and revisit that emotion. I think that human creativity which very well could be turned now towards creating more and more incredible artificial intelligence engines that do all sorts of amazing things. For me, creativity sums up a lot of the human experience, at least for myself. It’s, for me, something that differentiates human beings from machines and animals and the rest of the world, frankly, is the creative spark.
There are animals with the creative spark.
Sure, and we could talk about that too. For myself, the artistry and creativity is a defining characteristic of the human experience. To have an AI that also nominally displays that, of course, it’s not going to go create a painting in its own style, right? Which Rembrandt did, and that’s why he’s Rembrandt and we are not, right? It’s not the same because it’s imitative and it’s an approximation of what the master was thinking, right?
Certainly Rembrandt’s paintings evolved over time and he had a certain artistic muse, or something to say. The AI is approximating none of that, but at the same time it’s a little bit out of the armor, a little bit of a break there that says, “Hey, this thing that I thought made humans special, and maybe a lot of other people think the same way.” That’s chipping away at that. That’s probably why it’s getting an emotional rise out of me than some of these other tech topics normally might not.
It’s more than chipping away at it, right? I mean, if you think about how art schools work, they focus first on foundations and skills and fundamentals and you begin to learn a lot of different ways to create. The AI is learning different skills and different fundamentals and different tools and ways to create. The interesting question is how far are we away from the AI creating things that are uniquely its own based on the synthesis of many different styles and skills that it has developed beforehand? It probably is decades, meaning low ten to twenty years, as opposed to three to four, but it’s coming. Not too far away. It’s definitely coming, and how will you feel then?
Yeah. It just occurred to me that we have the Next Rembrandt now, and there will be the Next Picasso or, who knows, the Next Michelangelo, right? There will be this ability to replicate certain famous artists, which will also chip away at the market for artwork. If you think that you’re going to be a fine artist some day in the mold of any of these guys or gals, perhaps you should spend some time learning how to program a computer or artificial intelligence. The idea that there could be a synthesis of all these artists … Here’s what the intersection of Michelangelo and Rembrandt and Picasso would look like.
Who knows? I would actually be interested in seeing what a computer came up with, or what an AI came up with, in that respect. Sort of a super mash-up of the the greats. At the same time it makes you wonder where the value, just in terms of … Artwork in a broad sense is a a reaction to the world around us, right? At least it is … My understanding of music and art has always been in reaction to the things that are happening in the world, right? From that perspective you could then ask the AI to, “Hey, why don’t you react to the world? To the news?” Provide that additional input to make it even more human-like.
For me, artwork is always an expression of human condition. That piece of it is not quite sacred, but it feels close, right? Once again, I don’t know if you’re talking about having a computerized mind replace that. I would probably have a very, not a violent reaction, but something close to it.
Well, start violently reacting because it’s coming. Look, Jon, I’ll reframe it. Art is … I don’t think it’s so much an expression of the world around us as it’s an expression of the context that we have been exposed to and how we process and react to it. It happens that as humans and the way we live our life as humans, the context that we’re exposed to encompasses a great deal of data that seeps into many aspects of the world, at least in some limited geographical, cultural context.
The amount of context that machines will have in the future is going to grow exponentially. Right now the context of expression of the world that a computer would turn into art is a mash up of Picasso and Michelangelo and Rembrandt, or whomever, without the greater understanding of the world and place and the many things that inform human art. It’s inevitable that that context will be a part of artificial intelligence in the future.
At the point it has all of that context to go with the skills, it’s going to be doing what we do. I believe in hard determinism, so I believe that the things that we do are an inevitable, and if we had machines powerful enough to do so, even predictable synthesis of basically all of our nature and nurture up until that point, and as we interact with the world itself. The idea of free will and the human spirit, I think that those are clumsy ways of explaining things that we don’t have computational power enough to understand.
I think that AI and machines are being built in the model and mode of humans but with computational power that will continue to be orders of magnitude higher than we ourselves have. They’re taking our place as operational actors in many different ways, from a technological perspective, is an inevitability. The question is will things be done artificially from a social, legal perspective to prevent that operational eclipse of humans guiding the world. To me that’s the big question and unknown.
As far as computers being able to create original art in this dramatic, idealistic, lovely way that we think of human artists creating, I think it’s asked and answered and it’s coming. It’s just how long will it take the nerds to take to program the machines properly to be able to do that? It will probably be faster than we think.
In addition to artwork, I would very much like to hear a new Beatles album, as I’m sure many folks would. For the advertising agencies who are looking for the next big stunt, may I suggest that you use an AI to create a new Beatles tune.
Listeners, remember that while you’re listening to the show you can follow along with the things that we’re mentioning here in real time. Just head over to thedigitalife.com, that’s just one “L” in thedigitalife, and go to the page for this episode. We’ve included links to pretty much everything mentioned by everybody, so it’s a rich information resource to take advantage of while you’re listening or afterward if you’re trying to remember something that you liked. If you want to follow us outside of the show, you can follow me on Twitter @JonFollett. That’s J-O-N F-O-L-L-E-T-T. Of course the whole show is brought to you by Involution Studios, which you can check out at goinvo.com, that’s G-O-I-N-V-O.com. Dirk?
You can follow me on Twitter @dknemeyer, that’s at D-K-N-E-M-E-Y-E-R, or email me at firstname.lastname@example.org.
That’s it for episode 151 of The Digital Life, for Dirk Knemeyer, I’m Jon Follett, and we’ll see you next time.