Interview with Joi Ito

Sketch Notes by Rob Hanson

Sketch Notes by Rob Hanson

During a Future Lab earlier this year, we had the opportunity to sit down with Joi Ito, Director of MIT's Media Lab, to brainstorm about the growing intersection between synthetic biology and the internet of things.

JI -     Synthetic Biology reminds me of the early days of the internet. The internet redefined a lot of fundamental assumptions on how telephone companies thought of circuits or how big computer networks thought of architecture. 

New people actually had kind of an advantage because they didn’t have to unlearn the old stuff. Stuff is changing so fast that, this isn’t strictly true, but it almost feels like it’s true; the more you know, the worse off you are. We are constantly finding out that everything we thought was true is actually not the way it works. We used to think that genes were the code for everything, which it turns out through epigenetics and through other ways there’s all kinds of metadata around the gene and that you can transfer.

What’s interesting is now with CRISPR-Cas, I think it was formally published around 2013, that now allows us to edit genes for $30. Just during the period of a Master’s students studies, something we knew we couldn’t do, suddenly we can do.

JB -     How do you see the relationship between man and nature evolving in evolving in the next 20 years?

JI -     Over the past 25 years, we have seen massive global urbanization and that will continue to happen. In a city, there are all kinds of things going on at the biological sphere that we currently adapt to after the fact: we change the “thermostat”. If you actually understood how the biology worked, you could design a city with, and around, nature rather than in opposition of it. I think the future of architecture, and the future of cities, is going to be a lot more about understanding nature and designing things for that.

Today, if you think about what a city is, a lot of what’s important in a city is the things that live in the city: they’re human beings. You have more microbes in a human being, trillions of them, than there are human cells and human beings are essential in a city. In a city, there are all kinds of things going on at the biological sphere that we currently kind of adapt to after the fact. 

Whether you’re talking about honey bees or microbes, you could design a city with and around nature rather than in opposition of nature. That requires a certain level of modeling, a certain level of computation, whether you’re talking about the shape of the structure, the materials that you use, or the way that you run transportation.

I think the problem in the past is we’ve done as much as we can to shield cities from nature in these little bubbles, and then try to pretend it wasn’t there, but it’s actually the whole reason why we’re there. As we build better better models of our bodies and our environment, the invisible relationships between ourselves and our environment will become much more clear.


"Whether you’re talking about honey bees or microbes, you could design a city with and around nature rather than in opposition of nature."

Joi Ito, Director, MIT Media Lab


JB -     What exciting, unexpected paths do you see in the bio-space?

JI -     We have this thing called city farm and what we do there is is aeroponics and hydroponics. It turns out by changing slightly the light or the nutrient levels; you can fundamentally change things like the size, the color, the taste, the nutrients. It turns out even more than the DNA, the environment in which the plant is grown has more effect on the fundamental things that matter which is nutrition and growth. 

DNA still is important, we’re collaborating with Howard Shapiro who is the chief scientist at Mars. He has a wonderful project called the African Orphan Crops Project where he’s sequencing all these crop varietals. The cool thing that he’s doing is, because Africa and Europe don’t like GMOs, working with the local academics and farmers that know breeding, giving them access to all of the state-of-the-art gene sequencing. They can understand what we think this is part of the genome that’s going to increase the nutrients, to help with cognitive development, or what we think helps with drought robustness. Then they work together and think about how can we use traditional breeding to do the same thing as GMO techniques. We find solutions designed outside without any local content almost obviously don’t work.


JB -     How are biology and computing starting to converge?.

JI -     We’ve got Marvin Minsky and the whole A.I. crowd. You’ve got people like George Church saying “why do A.I. on silicon, why not do it in the brain?” Then there’s other people saying well maybe the brain and circuits should be connected, or maybe there should be a better user interface between A.I. and human beings. So Karthik Dinakar is working on what we’re calling human in the loop computation, which is putting the human inside the training loop of A.I. So we’ve got this whole spectrum of pure A.I. and we work with Carnegie Mellon and the rest of MIT on that; all the way up to George Church and making our brains smarter and everything in between.


JB -     What key changes changes need to occur for the type of technology, bio-engineering that you’re talking about, to occur in scale? 

JI -     I’m an internet guy so I look at things like the internet. The internet started working really well because we unbundled the layers. You could continue to iterate making the fiber more efficient, you could iterate on the IP and the BGP protocol and each layer could be worked on without really disrupting the other layers. You didn’t have to know about the other layers to contribute at your layer and I think what’s key.


"The internet started working really well because we unbundled the layers."

Joi Ito, Director, MIT Media Lab


You can have people working on new technologies for doing the gene editing without affecting the next layer, which is to take and identify the genetic blocks, and then another layer which is a design tool for genetic blocks. So you unbundle the pieces and create APIs between them. An example would be if you looked at Ruby or some programming language. It’s all modularized in a really good way.


Follow us as we explore the BIoT in this ongoing series about what comes after the internet as we know it. Stay tuned for product concepts inspired by the proceeding conversation, and others like it.