Conversation at the MIT Media Lab about cybernetics with Paul Pangaro, Nathan Felde, Mike Bove, Iyad Rahwan, Edith Ackermann, Joi Ito and Lorrie LeJeune.
A few background posts:
Chat posted live on Facebook Mentions at facebook.com/joiito/videos/961545600598042/
Conversation at the MIT Media Lab about cybernetics with Paul Pangaro | Joi Ito | March 18, 2016 at joi.ito.com/weblog/2016/03/18/conversation-at.html
[Warning: the audio recording level on Youtube is low. An audio amplifier will be helpful for listening]
[09:45 As we started to think about the future of Media Lab … and the Media Lab has recently been getting into machine learning, synthetic biology, and other complex adaptive systems. If we go back to the roots of Media Lab, there were some roots of cybernetics there, but for a while a lot of our focus is on computer-human interface and things that were a little more objects. And then we started to get into networks, and systems. But now, we’re shifting into a lot of the hard sciences in self-adaptive complex systems.
[10:30] And then, as I looked at people like Kevin Esvelt who has been doing CRISPR gene drive, he thinks much more about how do we think about who should decide, and how should that enter the system, and less about what is the specific technology that is in the gene drive. And there he can see the design across scales, where at each scale you have a complex system that interacts with other systems across scales.
[10:55] As we start to put all of the science together, what we’re realizing is that the traditional disciplinary science rewards a very focused, single object, rather than systems connected. Now, there is systems thinking going on. But also multiple systems across scales. If you look at Ed Boyden’s lab, he’s got about 50 people working — at very interdisciplinary and antidisciplinary — we touch multiple systems, we create tools that look at multiple systems, and we perturb multiple systems.
[11:30] And then we’ve got people like Kevin Slavin who have come in, sort of talking about participatory design. You’re not stuck in traffic, you are traffic. You put that together. We had a conference last year called Knotty Objects, which is about design. We brought Paolo Antonelli and they talked about critical design. One of the things I realized is that there’s critical design, which was a lot about people not doing it, but critiquing it. And then we had the Media Lab kids who tend to do things, but not being as self-critical as you might want. And they thought more in objects, than systems.
[12:05] So, we were grappling with how do we think about what we were all doing, and how could science be more responsible. We launched this Journal of Design Science, which was to try to break design into science, which means thinking about all of the systems, thinking about things as iterative interventions in an unpredictable complex system, rather than how do I make those predictable things more efficient. Then, how can we change design so it’s less focused on the customer, and more on nature, or the system, across scales.
[12:35] So, as I was searching for the right word to describe what we were doing, I found I was repeating on second order cybernetics.
[12:50] We were having a discussion at the faculty meeting about being interdisciplinary / antidisciplinary. The Media Lab was where all disciplines were brought to work together. Well, there actually was a thing called cybernetics where all of the discipline had come together in the Macy Conference, where was this wonderful moment where it felt like we were going to go transdisciplinary. But, sort of in retrospect, somehow it disappeared.
[13:12] So, how did it disappear, where did it go? Stewart Brand would say “it got bored to death”. Some people said it got too academic, some people said the Macy Conferences ended and disappeared. Some people said the applications got so compelling that it ended up being applied. Theoretical. So there were a lot of reasons why it disappeared.
[13:40] So, I traced someone holding a torch. So now, what I want to explore is a couple things. One is, is there a way to connect cybernetics into what we’re doing in research? Now that I’m turning 50 this year, as an old-timer, I hear the crypto-currency people saying the same things that I was saying back in the 90’s, and making the same mistakes we made when we were building the Internet. I don’t like to repeat mistakes. I don’t like to rehash stuff that’s already been done. So, what can we can learn from cybernetics’ successes? But there are also learning from its failures. What could you or we have done better?
[14:15] We can look at the met-catalyst(?) movement in architecture, which was about bringing biology and architecture together. It was Tungay in Japan that did it. But it kind of died. I think it was because we didn’t have the tools to bring biology into architecture. Today, we do. There are a lot of things, like machine learning, and other things, where the technology has caught up to the theory, so we can apply it.
[14:40] So is it possible that maybe some of the things that we thought about in second-order cybernetics are more relevant and more possible now? [….]
[15:05] So one thing right now that is an argument — a disagreement — at the faculty level is about whether we should grab cybernetics, or not. Shouldn’t we just use the word “design”? Cybernetics comes with a lot of baggage. There a lot of people who are practitioners of cybernetics. When I wrote a little bit about cybernetics, I found some very enlightening comments coming in. But some people who had done stacks of work that seemed a little bit too … like the tools they were using to think about it, they weren’t using the new tools. I didn’t want to diminish what they were doing, and get on their turf, but I didn’t think I could grab the whole lot. [….]
[15:50] Since all of you are in touch with where we are with the Media Lab, or at least you know the DNA of the Media Lab, and you also know the history of cybernetics, I’d love to find — talk about the history of cybernetics, and talk about the forensics of it, and where it is today, and then maybe talk with about how you think it might apply to the future. [….]
[19:00 Paul Pangaro starts the discussion with “what does cybernetics offer”?]
[59:25] So where is this discussion, now, around, what is the terminology?
[59:30] This has been a useful conversation. We’re launching this idea around — we’re using the words “extended intelligence”, to be what we’re using instead of AI, to talk about this environment. We’re also going to be doing this meeting around AI and governance. And we’ll be launching what it means to put society into the loop.
[1:00:00] In terms of doing, we’re doing. I think a lot of the idea and the words you’re talking about resonate with what we’re thinking. We launched this Journal of Design Science, and are now working on how we describe it. We’re trying very hard — and Stewart has been helpful in thinking about this — to keep it a conversation, rather than be a series of peer-reviewed papers. The way we’re trying to do it is to have a conversation; have the online — we’re using a platform called PubPub that allows versioning and commenting — for the output to also be a reflection of the conversation. […]
[1:00:55] So, we’re moving forward. What would be interesting — talking about cybernetics — would be to learn from cybernetics, and also to see, if you and other people working in cybernetics can take some of the things that we’re working on — like say, machine learning, or evolutionary biology [….]
[1:01:50] The tricky part for me is, how do we have this conversation? I do think that kids don’t read books any more. It’s an ongoing conversation. The words are very fluid. Some of them mean different things to different people. [….]
[1:02:25] The world is much more global, now. Macy was kind of confined, in the cultural context. You talked about the influence of the Austrians. But now we have the impact of everyone. How do we have this conversation across languages?
[1:02:40] To me, that may be the harder thing.
[1:10:35] So that’s my personal meta place. What is the institution? I look at Bauhaus, I look at Black Mountain, I look at RLE, and all of the other institutions. And for some reason, the Media Lab has sustained over 30 years. I think a lot of it is the approach. [….]
[1:11:00] That could be what we learn about the forensics of cybernetics. [….] Some people had a tremendous amount of impact, but not sustainability. The Media Lab has a weird thing. We like orthogonality and disagreement, we build tools, we’re not obsessed about theory, although we have it, not as a primary output. And then, we’re happy to move along, as new tools come, and new technologies come.
[1:14:00] I feel like the best designer designs themselves, as the intervention. And, so, it’s me personally, and then the institution. The journal is for us, but I don’t want to create the church. So you have to make the membrane permeable, but humble in that I want to affect myself. In affecting myself, I may affect the environment, in a responsible way … but not to be evangelical about it.