The prolific investor admits he likes unconventional companies in the deep learning space. So it’s not a surprise that in late March, DFJ led a $9 million investment in Mythic, an Austin-based company that claims that its platform frees local artificial intelligence (AI) from the cloud and conventional on-device technologies that limit sophistication.
The Series A round also included participation from Lux Capital, Data Collective and AME Cloud Ventures. (But more on Mythic later.)
The AI Sector
While there might be a fair amount of debate as to what companies are actually in the AI space, it’s clear the sector is on fire.
In 2016, reported funding into US-based companies in Crunchbase’s artificial intelligence category hit new highs. 233 companies raised fresh capital, which totals roughly $1.06 billion into 180 companies. It should be noted that over 40 percent of that capital was raised by the top ten companies ranked by the amount of money raised. In 2012, only 44 companies in Crunchbase’s artificial intelligence category raised rounds. Nearly $108 million in equity financing is represented in the 26 rounds for which 2012 funding amounts are known.
In the first quarter of 2017, 51 companies in Crunchbase’s artificial intelligence category raised rounds. Over $1.37 billion in equity financing is represented in the 42 rounds for which funding amounts are known, the overwhelming majority of which is attributable to the $1 billion investment by Ford in Argo AI, a Pittsburgh, Penn.-based startup aiming to develop a virtual driver system.
Matt Murphy, a managing director for Menlo Ventures, joins Jurvetson in being a big believer in AI—particularly SaaS-enabled AI companies. Since he joined Menlo in 2015, he estimates the firm has invested some $75 million in about six AI-related companies. Murphy led the funding rounds of Veriflow, Usermind and Clarifai and followed on in an investment in Heap, a data analytics platform.
“I think the first wave of SaaS was digitizing workflows. But then, in the last couple of years, we’ve seen the playing field shift to the intelligence layer,” Murphy said. “People have started realizing that if you integrate AI into the workflow, then you are able to make applications much smarter and thus make a company much more competitive.”
Jurvetson said via email his firm is seeing the powerful application of deep learning and directed evolution across its portfolio, ranging from molecular design to image recognition, cancer research, and autonomous driving.
Back in 2014, DFJ led the investment in Nervana, which focused on digital deep learning chips. Eighteen months later, Intel acquired the startup. Overall, Jurvetson believes machine intelligence (AI) is of fundamental importance to every business over time.
“The application of iterative algorithms (e.g., machine learning, directed evolution, generative design) to build complex systems is the most powerful advance in engineering since the Scientific Method,”Jurvetson asserts. “Machine learning allows us to build software solutions that exceed human understanding, and shows us how AI can innervate every industry.”
Beyond the chip companies (Nervana and now Mythic), DFJ has been investing in the application of AI in various industries. Other portfolio companies include Cylance and D-Wave Systems.
In Mythic, Jurvetson saw “a much larger opportunity… as the application of intelligence at the edge (the IOT) is much greater than the opportunity to do learning in the data center.”
“I like to invest in companies that are unlike anything I have seen before,” he continued. “Mythic is taking a radically novel approach to building low-power, low-cost neural nets that can make any product more intelligent.”
Mike Henry, Mythic CEO and co-founder, says his five-year-old company makes it easy to put powerful, local speech and vision AI into any product from wearables to cars. “We can deliver the processing capabilities of 10 mobile GPUs in a small, low-cost form factor,” Henry, who has a PhD in electrical and computer engineering, told Crunchbase News.
Since its March raise, Mythic has more than doubled its staff to 11 employees. It plans to double that by this summer.
“The challenge of the AI space is that it’s so broad, and also a lot of companies are using the term loosely,” Henry elaborated. “Once you’ve already got some trained algorithms, the next step is how do you deploy that out in the field and use it in the real world. That’s where we come in.”
Just how many devices can benefit from having AI inside of them? The number is huge.
“We basically identified back in 2012 this was going to be big and predicted there would be a bare-knuckle boxing match between the big chip companies to deliver a local AI solution,” Henry says. “As a small company, we knew we needed to come up with something unique, different and not easy to replicate. We came out of stealth mode with proven technology when the market was white hot. The timing was right.”
While it remains to be seen how long investor interest in artificial intelligence will last, it’s reasonable to think that the startups who move the quickest while their category holds investor attention will make off with the lion’s share of the capital.