“We don’t even know what it is yet,” says Facebook co-founder Mark Zuckerberg’s character in The Social Network. “We don’t know what it can be. We don’t know what it will be. We know that it is cool. That is a priceless asset I’m not giving up.”

Investors have expressed similar feelings when assessing the potential of artificial intelligence (AI). The applications seem limitless, but for every Facebook the sector has its share of My Spaces.

Which companies are driving the truly transformational and which will be transitional? And how can institutional investors get in on the ground floor? Our Semiconductors & Automotive Technologies Analyst, Vijay Rakesh, recently convened Mizuho’s 2018 Silicon Valley Seminar to spotlight some of the most interesting, privately-held firms.

Among them was Clarifai, a leading image recognition company that is poised to drive a paradigm shift in search capabilities. Its technology allows you to expand beyond text searches to use images to return matching content. The applications for the retail/fashion industry are apparent, but this “computer vision” company also helps our government’s missile targeting weaponry improving the accuracy of drone strikes. The technology is also helpful in policing social media sites and preventing the uploading of offensive video or visuals.

Mountain View, California-based H2O.ai is on a mission to “drive AI democratization” by supporting the growing data science community around the world. H20.ai is building a broad, deep learning AI platform it calls ‘unsupervised, driverless AI technology.’ By “driverless,” they don’t mean autonomous vehicles. Driverless AI refers to the automation of machine learning – essentially, platforms that use AI to do AI. Driverless AI may see substantial traction in insurance, credit card defaults and fraud, healthcare and cancer research, among others.

When it comes to the self-driving car type of AI, JingChi is at the forefront, with an autonomous taxi fleet of approximately 30 vehicles in China and plans to expand to 200+ by the end of 2020. Autonomous driving technology and its implementation is a main point in China’s Thirteenth Five Year Plan (2016-20), meaning that JingChi may be in prime position for key collaborations among cities and local governments across the world’s most populous country.

Remember when IBM’s Watson beat Jeopardy! All-Time Champion Ken Jennings in a game? It was most people’s first demonstration of the instantaneous comprehension and computing power of AI and a glimpse of things to come. Loop AI Labs wasn’t even formed when the historic 2011 competition took place, but it has since developed its own software for natural language understanding across the automotive, telecom, banking, insurance and fast food markets. LoopAI executives at our Silicon Valley Seminar discussed how their technology drives an unsupervised learning algorithm on unstructured data, without the need for data scientists. And its technology can be implemented in 1-3 months at 1/10 to 1/100 the cost of Watson.

Vienna, Austria-based Cortical has developed Natural Language Understanding (NLU) technology that can not only identify data but can also understand its meaning. The technology is inspired by neuroscience, with algorithms imitating how the human brain processes information. The company’s Retina Engine converts words or text into numerical codes or semantic fingerprints.” Cortical believes that with NLU, it can create a fingerprint with as little as 10 text examples, while Natural Language Processing requires thousands.

If you are interested in learning more about this year’s AI Conference or would like to learn more about other upcoming Mizuho events, please email Mizuho Corporate Access at USEquity.CorporateAccess@us.mizuho-sc.com.