Technologist and VC Kai-Fu Lee has called AI the ‘singular thing that will be larger than all of human tech revolutions added together’ and said AI will probably replace 50% of human jobs. It’s appropriate to be concerned about how we manage the significant disruption AI will bring, but AI has enormous potential to accelerate businesses and elevate the human condition.  And AI is (finally) turning from sci-fi into reality as software eats the world, compute power gets cheaper/faster, algorithms and models mature, and massive amounts of data become available.

Huge investments in AI are driving this recent traction.  As a recent McKinsey study outlined, investment in AI has been accelerating at an unprecedented rate with $26 to $39B invested in 2016, which was nearly three times as much as in 2013.  Internal R&D ($18-27B) paced well ahead of external investment from P/E and VC ($6-9B), as the big tech companies (Apple, Google, Baidu, Amazon, Salesforce) and automotive companies (BMW, Toyota, Tesla) poured money into computer vision, robotics, speech recognition, virtual agents and machine learning to power their own applications.

The big tech companies are building a lot of the backbone of AI, and are rapidly advancing a variety of models and algorithms (many of which are being made available as open source).  As venture investors, we do not seek to fund AI companies to directly compete against these massive R&D investments.  Instead, we look for software companies solving business problems with compelling workflows that generate new datasets. These unique datasets can then be used to build upon the foundational AI technologies (e.g. computer vision, natural language processing, open source models) for a particular vertical.  We are in violent agreement with the comments by Accel’s Jake Flomenberg in his recent interview on Twenty Minute VC – vertical SaaS companies innovating in AI should crawl, walk, then run – build workflow first, harness a new and unique dataset from these workflows, apply models and algorithms and then iterate.

This is why we are excited to lead the latest round of investment in XOi. Aaron, Anthony and the team at XOi identified core problems in Field Service – a rapidly growing trade skills gap and a lack of trust between technician and end customer.  Their VisionTM product delivers compelling workflows that help bridge this gap, build trust, and accelerate sales.  These workflows enable the capture of millions of new data points – audio and video content of every job performed by a technician in the field.  The XOi team has been applying models and algorithms to this data to drive even more efficiencies and sales for their customers, and much more is still yet to come.

To read more about the XOi investment, see the full press release here.