By Lu Zhang, Fusion Fund
To say artificial intelligence is old news would be an exaggeration, as we’ve only just started to see its implications: It’s transforming industries across the business landscape, and PricewaterhouseCoopers predicts AI will grow the global economy by $15.7 trillion over the next 12 years.
Needless to say, venture capitalists should be investing in AI. But it has become more sophisticated since the early 2000s, and investment strategies need to follow suit. We can now stop focusing on whether the technology itself is good enough. It is. Instead, investors must consider how a company is applying technology and whether the market is ready for that application.
Where to invest
Innovative technology is no longer a distinguishing factor for businesses. Open-source projects created by bigger companies, such as Google’s TensorFlow or Microsoft’s Cognitive Toolkit, have made underlying AI technology available to any sector. And, because these companies are using huge amounts of data collected from those programs to constantly improve, it’s now unlikely that a venture-stage company would be able to assemble a competitive underlying technology.
Instead, investors must seek startups that are using AI in genuinely disruptive ways, such as those that have identified a specific market where AI can solve problems and be easily integrated into the existing infrastructure.
These are the sectors most likely to adopt disruptive AI solutions in the next decade:
Manufacturers already use industrial-grade robots in their production plants, such as the robotic arm used by Tesla. But the big winners in this field will be companies that use AI as a full-package solution, rather than as a partial one.
The healthcare industry is already beginning to adopt robotics in surgical applications, and the adoption rate will likely increase as underlying medical technology improves. One particularly exciting opportunity is in “soft” robotics: Imagine a robot that is intelligent and gentle enough to wrap around a beating heart while a doctor is operating.
AI is beginning to make other inroads in healthcare, too. Doctors often have to review hundreds of medical images per day, closely examining each one for potential irregularities. By pre-screening these images with computer vision technology, however, doctors could more easily access only the images that AI flags as potentially problematic. Thus, saving valuable time.
While this makes most VCs starry-eyed, we must remember that integration costs are important. In some instances, a tech developer may introduce a robotics solution that’s far too advanced for a certain industry. When this happens, the industry likely can’t or won’t redo its entire infrastructure to adopt this new technology. Thus, it’s important to note that new tech innovation, especially in robotics application, must allow a low-integration cost to achieve success.
2. Customer Service and Sales
The improved function of natural language processing (NLP) to better understand and respond to human speech is one of the most exciting tech advances of the past few years. A widely available NLP device like the Amazon Echo would have been considered unrealistic a few years ago. Naturally, industries that are most communication-heavy will see the biggest payoff from NLP solutions.
Many companies are already turning to NLP in the form of customer service chatbots or automated phone assistants, but the real potential for NLP is in enabling human interactions, not replacing them. Imagine if an NLP program could monitor a customer conversation and help the customer service rep respond. It could authorize a discount, present in-depth product information, or offer to connect the customer to the appropriate person at the company.
NLP technology could also help a sales representative make real-time decisions when talking to a lead. The companies that can decode these types of high-value AI solutions have a very high ceiling.
Machine learning enables programs to become more intelligent by “learning” from data sets and user feedback. Google is determined to set the industry standard in machine learning with AlphaGo and TensorFlow, and Microsoft launched an internal open-source machine learning platform to spark application innovation among its developers.
Investments in machine learning can be highly strategic because there are so many uses for it. From disease detection to data security, machine learning can conduct important processes at exponentially faster speeds than any human, leading to improved outcomes in every field in which it might be used.
There are no shortcuts in AI investments. To understand the opportunities, you must analyze them by industry. What does that market need at this moment? AI investments only produce results if you have the right application at the right time. Even with as promising a technology as this, investors need to redouble their efforts to find the signal through the noise.
Lu Zhang is the founder and managing partner of Fusion Fund, which promotes early-stage venture capital for entrepreneurs. She is also an elite member of Forbes’ 30 Under 30 list and was nominated as World Economic Forum’s 2018 Young Global Leader. She can be reached via LinkedIn at lu-zhang-vc.