Bull Session

AI + Human = Centaur

September 6, 2018          

Episode Summary

On The Digital Life this week we chat about human / machine hybrids — the combination of AI and people — into advanced teams that produce amazing creative output. From chess to photography, art to science, these so-called “Centaurs” may represent the future of work for many fields. Join us as we discuss.

Art by algorithm

Welcome to episode 274 of The Digital Life, a show about our insights into the future of design and technology. I’m your host, Jon Follett. With me is founder and cohost, Dirk Knemeyer.

Greetings, listeners.

For our topic this week, we’re going to talk about humans working very tightly with artificial intelligence in advanced teams that produce amazing creative output. These human machine hybrids have been labeled so-called centaurs, which I think is very interesting term for humans and machines working together and was inspired by a wonderful essay on Aeon which is aeon.co, the website if you’re interested. The essay is, “Art by Algorithm,” written by Ed Finn who has also written a book, “What Algorithms Want: Imagination in the Age of Computing.” I quite like Ed’s essay on Aeon and recommend that listeners check it out. Let’s start with sort of the fun naming of this term which I think for all the geeks in the audience sort of appreciate the centaur term for its mythological background.
This of course is a body or the upper torso of a man, the body of a horse, and there are these powerful, smart, majestic, fierce creatures that can sort of kick a lot of ass in mythology. Right? This idea that we’re bringing together AI and human beings to create a centaur, it’s just a funny term to me because I don’t know rather than call it some other mythical hybrid they come up with this fearsome creature that kicks ass and plays a prominent role in Harry Potter as well for all you Harry Potterites.

Yeah. It’s a really clever term, with terms like this, trying to coin a term, many clever terms are not adopted and don’t really go anywhere. Whether we ever use centaur again after a few months or a year from now, who’s to say, but it certainly is clever and would be a nice term for talking about these, the hybrid projects, the hybrid initiatives between humans and machines.

Yeah. I think there’s some history there as well. I believe Garry Kasparov has talked about these human computer teams that play chess. There’s a new form of chess play that involves these human computer or hybrid teams called freestyle chess and that’s way out of my depth in terms of the types of things that would happen in those games. Essentially, the human plays the strategic role and the computer sort of runs through all the tactics and options and anticipates possible negative repercussions for the human sort of initial strategy. At a really high level, that’s what freestyle chess is. This idea of the centaur, the human plus the enhancement of the AI to make these so much better. But you know, Ed Finn raises a really interesting point in his essay, which he’s basically saying, we kind of do this now, this is at the core of our creative output. Dirk, bringing this back to the idea of human beings being great creators, right, and then using AI as a tool to enhance that, that’s where I get really excited about it. It’s really just part of the continuum of what we’re already doing. Right. In a lot of ways we’re already doing this computer human hybrid to generate all sorts of interesting outputs, whether they be excellent photographs that with a regular camera. There’s no way I could take some of the photographs that I take with the iPhone X. I mean it’s got this lovely portrait mode and I swear every time I post something to Facebook, people are like, “What are you shooting with? That looks awesome.” I’m like, I’m just using the portrait mode in iPhone X.

It’s not Leica. No Leica.

Yeah, no, I mean like all of this accoutrement, the history and skill and working in the dark room, which I used to do. Just like all of these things, learning about cameras and lenses. When I was in seventh grade just figuring this stuff out, how to frame the shot. All of this is negated by smarty pants algorithms that level up all of your skills for no other good reason than, hey, like now it’s a beautiful photograph and that’s so much better than posting a crappy snapshot. In fact, I think part of the iPhone X marketing is just showing people in their ads like how fantastic some of the photographs can be, just sort of generated by this computer software. To Ed Finn’s point, like, yes, this has made me a much more expressive and interesting photographer and I didn’t really do anything to deserve that. I just bought an iPhone X.

Yeah, it’s interesting. Right? One of the things that I think complicates matters is the label of artificial intelligence. When we talk about AI and humans and computers playing chess together, people get a little … I think it’s a little odd, like certainly chess championships are still being conducted human versus human, right? The centaurs are not present in those competitions but in other contexts like photography, which you were just talking about, we take for granted how completely the technology has integrated into the humanity to better represent the world. Once upon a time before the invention of cameras, we would have had to draw each other. If I saw you or saw a scene and wanted to capture it to and sort of taken out of time asynchronously and share it with someone else, I would have had to use my hand which from one person to another the skill difference and how well that could be done is vast.

Now, that technology, that iPhone X, which is a really expensive phone, Jon, that iPhone X is available to anyone who can afford it, which is in the United States of America most people are able to buy it whether or not they should be buying it with the money they have is a different question. But it’s a widely available piece of technology that lets all of us capture reality for asynchronous sharing right there, we can all do it at this ridiculous level of quality far beyond the little drawing and nobody thinks about it because it’s a camera, it’s a phone, it’s a technology that we consider a thing, we consider inanimate and it’s just giving us more, more power. With the chess example, it’s a product of a particular type of software and software development that is falling into artificial intelligence, AI.

Because science fiction takes that into artificial wife and machines replacing humans and all of these other things, people are a little bit concerned about it. People are a little bit scared of it, but it really isn’t particularly different at this moment in time in technological evolution and for decades in the future minimally than a camera. It’s just instead of a piece of hardware with some software running in and very advanced technologies and aspects to that hardware, it’s software that is running a particular way that’s very advanced and is able to do different things. There’s real demystification that needs to happen for our being able to leverage AI and leverage sort of a centaur approach to doing things as well as we can that we need to overcome.

Yeah, I think that’s a great point. I think that’s sort of worth digging into from a slightly different perspective that I think I can share from the music side of things. What sort of modern music and production does particularly well is incorporate software or hardware with traditional creative means to generate very interesting artistic music releases. What I mean by that is there’s a ton of software that is, and some of it is even available more or less for free on your laptop that can make music production happen so much more easily than back in the ’90s when I was sitting in front of a mixing table and a reel-to-reel. I’m not complaining, but pretty much taking tens of thousands of dollars worth of equipment and shrinking it down into bits and bytes and giving it away for free. Additionally, adding all kinds of interesting filters and auto tune and effects and different processes and workflows where you can construct things now that would have taken a massive budget and a lot of know-how, if this were 20 years ago.

What I find interesting is the conversations around creating music. There’s certainly, there’s a hell of a lot of back and forth about whether or not a laptop can be considered a musical instrument. Whether or not all of these processes are really helping or not, because of course you can take someone who can’t sing a lick and tune up their voice and all of a sudden they sound reasonably acceptable, right? They sound like a rockstar, and add some distortion. You can even split up their voice into different tonalities. Now you have backing vocals. There’s so many things that you can do with the enhancement of computers to create musical art, to create musical releases that some people find that offensive, right? Some people find that difficult to digest because it’s not done in the way that it was previously. In some ways it’s reducing the amount of human skill required because the software really automates and pulls in some of the tasks that would otherwise fall to a human engineer or a recorder or what have you.

My point is that music is integrating all of these tools fairly well and one of the things that I as a music creator listen for now is how people are using the software tools in their music. When I hear songs now, I’m really interested in what filters and what harmonies and what interesting things producers have done to make their music more interesting. I think it’s bringing out from people, it’s giving them more creative opportunities and I don’t think that the controversy over the mechanics of generating this music is really worth paying attention to because I think the outputs still really represent great human creativity. That being said, there’s a ton of music being released now and a lot of it’s not getting listened to. I think part of that is because of this ability to just generate music on your laptop, record in a tiny studio in your house, and then reasonably have an album you can share online doesn’t require the same level of fiscal backing that it used to so for what that’s worth.

Yeah. I mean, it allows anyone to be an expert, right? I mean, there’s different levels of quality and final output, but you can take someone who’s not been trained musically who is just kind of figuring it out for themselves with the tools and monkeying around and they can produce music that sounds pretty good. It may not be quite as good as the best stuff that’s released commercially, but somebody who you wouldn’t expect can release something that’s pretty, pretty wonderful. You used the phrase earlier of replacing human skill. While that’s true in the most literal sense, I think we focus too much on that part of it, of, well, it’s coming in and replacing this thing a human could otherwise be doing. It is, and it’s allowing the human to do other things, to do things that are what I might say higher level things or things that allow us to extend our reach and our grasp even further instead of having to do manual laborious processes, the machine auto magically does them.

What can we do with that time? We can listen to more music to make our context more rich and more full. We can learn new techniques, we can research new software to bring in more software and make things even better. There’s a lot of new things that are enabled that will make the music better, that will make us smarter by virtue of the fact that those things are happening. Even though it’s technically correct that it’s coming in and replacing things that used to require the human skill. In the process, it’s enabling us to do much more and oftentimes more interesting things because typically the things that are being automated are not the things that people would choose to do if they could do anything that they want with their time. It’s usually the more funky stuff that’s being automated. So rock on automate away as far as I’m concerned.

Yeah, there’s a lot of exciting things that software can do for the artists. Be the musician, photographer, you know, you name it, that the software enhancements are there and are growing. Even if we think about sort of the broader array of human tasks and jobs and work, we think about all the writing tasks, all of the computational tasks around accounting or all of the research tasks in science. These are the kinds of things where software AI driven software is going to be making a huge difference around automating things that like you pointed out, Dirk, might not be super exciting. Maybe there will be better ways to vet the thousands of papers that get injected into scientific literature every year so that scientists can be better informed in their research and yet don’t have to spend quite so much time keeping up with their reading of papers.

From the writing perspective, we already know that there are a number of software programs that are sort of automated and able to generate stories that we can read about. You know, like minor league baseball stories are being automated based on the statistics that come out of the game. There’s so many areas where these opportunities exist and I think part of the framing of this discussion around what comes next for humans as they do work and as they do creative things, it’s going to be important to number one, look at AI and software as tools of enhancement and this centaur framing of the problem is I think pretty useful for that. Then also just to appreciate that there are people who are going to be very uncomfortable with introducing software at all levels of their work. Getting over that hump as well I think is going to be important. Whether it’s in science or art or writing like these software packages are going to inundate these fields and we can sort of already see that happening.

Something we need to get ready for. 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 the digital life and go to the page for this episode. We’ve included links to pretty much everything mentioned by everyone, 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. You can find The Digital Life on iTunes, SoundCloud, Stitcher, Player FM, and Google Play, and if you’d like 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 GoInvo, a studio designing the future of healthcare and emerging technologies 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 @D-K-N-E-M-E-Y-E-R and thanks so much for listening.

That’s it for episode 274 of The Digital Life for Dirk Knemeyer, I’m Jon Follett and we’ll see you next time.

No Comments

Leave a Comment

Your email address will not be published. Required fields are marked *