Machine learning and artificial intelligence are all the rage today in venture capital circles. We've seen spectacular exits in the past few years, from Google absorbing Deepmind in 2014 for $500 million, to Twitter buying TellApart in 2015 for $533 million, and Intel swallowing Nervana in 2016 for $400 million. But these were all IT plays.
What happens when machine learning meets biology?
Berkeley-based Lygos is engineering and designing microbes that convert low-cost sugar into high-value, specialty chemicals. At the core, they're developing and exploiting a number of tools, both software and hardware, and applying them to biology. Ultimately, the ability to design and optimize microbes, or program them, is becoming faster and cheaper than ever before. This is being fueled by cutting-edge advances in data science and biotech, and the rapidly dropping cost of reading, writing, and editing DNA (a trend that's even faster than Moore's Law for computing).
In other words, the latest advances in software, big data, machine learning, biotech, and chemistry may be combining to quite possibly start a new industrial revolution.
A new spin on machine learning
Cloudera co-founder Jeff Hammerbacher once declared, "The best minds of my generation are thinking about how to make people click Ads. That sucks." He was right about the "suckiness," but perhaps some of those best minds are working on things that are far more substantial in their impact.
Take, for example, Eric Steen, co-founder and CEO of Lygos, an interesting new startup that aims to become the next DuPont, but not in the traditional way. Lygos develops microbes to convert sugar into high-value specialty chemicals, focusing its flagship product on malonic acid (derived from petroleum), which is used in a diverse set of industries, including flavor and fragrance, electronic manufacturing, and coatings.
But, how Lygos got to this point is what makes it interesting.
Microbes have evolved over millions of years to become hyper-efficient factories. A microbe has amazing computational and machine learning ability because so much is written into the genetic code. Evolution is nature's machine learning algorithm.
Lygos is unlocking the ability to control and guide evolution to reprogram a microbe to produce its products. A microbe can do a computation every time it divides and grows itself, which it does every 20 minutes. Lygos has millions of them going in a single vat at a time. Using technology that harnesses evolution, they have a more powerful machine-learning platform in nature than a computer could ever deliver, and are developing and deploying a range of these technologies to design microbial factories.
Data science beyond Silicon Valley
This is important, super-cool stuff, and it's yet another reason to look beyond Silicon Valley for the most interesting innovations. In the case of Lygos, it's based in the Valley, but its impact is not.
I've written before about John Deere revolutionizing farming through big data, and continue to believe the most important work in data is not happening within 45 miles of San Francisco. As McKinsey & Co. has noted, industries like manufacturing and retail have huge incentives to embrace data science and plenty of data to work with. And, though they will borrow tech from the titans of Silicon Valley (e.g., TensorFlow from Google), and cloud vendors like AWS will lower the bar for developers dipping their toes into machine learning, the biggest impact of big data will not go toward ad-clicking strategies. Thankfully.
- Big data's biggest impact is not on Silicon Valley (TechRepublic)
- Why AI and machine learning are so hard, Facebook and Google weigh in (TechRepublic)
- The cloud war moves to machine learning: Does Google have an edge? (TechRepublic)
- Machine learning: The smart person's guide (TechRepublic)
- Google's problem with the enterprise cloud is that it's too innovative and not practical enough (TechRepublic)
Matt Asay is a veteran technology columnist who has written for CNET, ReadWrite, and other tech media. Asay has also held a variety of executive roles with leading mobile and big data software companies.