Versatility drives venture dollars to AI/ML companies

Artificial intelligence and machine learning applications spread further into new industries as speed and efficiency prove data science isn’t just for IT

Artificial intelligence (AI) and machine learning (ML) have become ubiquitous across the global economy. These technologies, which provide enormous efficiencies and scale can be as basic as product recommendation models, or as mind-bending as quantum neural networks. Still, the space in between opens the door for the thousands of business applications that have attracted a surge of venture capital (VC) dollars.

Indeed, big-name deals grab the headlines, but much more is happening in the background. Since 2017, $188bn in venture capital has been invested in North American companies that apply AI or ML in some way (Fig. 1). This is seven times the amount invested over the six years prior. Moreso, this capital has been invested in nearly 7,000 separate funding rounds during that period. This growth has been fueled by many factors, primarily including the combined computing power and storage offered by cloud-computing platforms such as Amazon Web Services and Microsoft’s Azure, which have moved technology innovation away from traditional and expensive centralized locations. 

Last year alone saw $81bn in AI/ML investment, making up one-fifth of all VC deals made into US companies amid a record year for the industry.

What are the key drivers for AI?
Innovation, scalability, and diversity have been the key drivers for venture capital investment in artificial intelligence, and there has consequently been a large volume of capital put into companies applying it across industries. The information technology industry is home to most of the AI/ML investment today. From under this umbrella have come companies including Databricks, Stripe, and Aurora Operations, which have attracted a combined $8.9bn in venture equity (VE) over the past five years, but use AI/ML technologies in very different ways.

Databricks quickly understood this versatility. The company makes big data simple, allowing users to explore their own data, integrate data sets, and run AI algorithms to fit their own needs. But the core of its business model highlights the open protocol for secure data sharing of AI/ML. Anyone can use it for their own purposes. Unlike pharmaceutical products or electric vehicles, which theoretically have a finite number of customers, Databricks can be applied, also in theory, to an endless number of problems.

Since its founding in 2013, Databricks has attracted $3.5bn across nine funding rounds. As recently as August 2021, the company received $1.6bn in series H funding at a post-money valuation of $31bn, according to Preqin’s Company Intelligence.  

Healthcare and medical research were a natural fit for AI/ML. With this technology in their toolbox, researchers vastly improved their ability to quickly process large amounts of data and find patterns across convoluted datasets. Investment in AI/ML healthcare companies over the last five-plus years is second only to IT companies at $26bn. Companies focusing on molecular science and gene therapy have received notable venture interest, such as GRAIL, Inc. which has received over $1.9bn in funding since 2017.

GRAIL builds cancer screening tests for early detection across the cancer spectrum. After eight funding rounds between January 2016 and May 2020, the company was sold to genetic sequencing company Illumina for $8bn in August 2021.

Financial services have also jumped on the opportunities that AI and ML offer where massive companies have emerged. Stripe, Inc., a finance and accounting infrastructure provider, has received $2.9bn in funding since 2017 across eight deals. Specialized in building digital payment systems, Stripe’s notable customers include Deliveroo, booking.com, Hostelworld, and ASOS. Preqin’s Company Intelligence estimated Stripe’s Series H post-money valuation at $95bn as recently as March 2021. 

West Coast, best coast

California, the epicenter of the global technology industry, has remained at the heart of the AI/ML boom, particularly in the Bay Area. The most populous state in the union still draws billions more to its tech companies than other nearby states or provinces. Venture investment in AI/ML companies in California comprised more than half of all VC investment in North America over the past decade. 

California certainly lost ground in AI/ML investment over the past five years, but it’s still an inarguably high peak to reach. New York and Boston have increasingly become home to more tech companies, as has Texas. However, news that tech was leaving California by a significant margin is perhaps overblown. Indeed, large companies such as Hewlett-Packard, Oracle, and Salesforce relocated as a result of high taxes and living expenses, for example, but this hasn’t stopped smaller companies from continuing to consider the state home.

Arms race
Although the business case for AI and ML is apparent and proven, does it match the increase in valuations in 2021? In a year where VC funds added $600bn in AUM in just nine months, other forces are clearly at work besides logic-based discounted cash flow models. A look at the NASDAQ-100 Technology index, as a proxy, showed it reached its peak in October 2021 as its year-over-year return passed 46%. The period that followed was a market correction marred by persistently high inflation and heightened global unrest which has adversely impacted the tech sector as well as other corners of the market. 

Despite these headwinds, AI and ML are likely to remain key elements of the venture capital market, owing to their flexibility and diverse application. As such, we can expect funds to continue committing capital to companies focusing on these technologies.

What is artificial intelligence?
Briefly speaking, artificial intelligence (AI) is a system designed to predict outcomes and make decisions, much like the processes of the human brain. And like the human brain, which decisions are made can evolve as the system is fed new data. Efficiency, however, underlines everything with AI. AI-empowered systems can analyze huge amounts of data to identify patterns and make predictions at a scale the human brain is not capable of. For a more detailed description of AI, Oracle put together a useful overview.

What is machine learning?
Machine learning is a component of AI. Machine learning (ML) algorithms use statistical methods to train models to classify or predict outcomes. The models are built with a training set that is used to help determine the best model. The simplest ML models use structured, labeled data where the outcome is clear, but more complex ones use artificial neural networks that act like the human brain. While these models can perform well on a fixed set of data, how they perform on new data is a constant challenge. As cloud-computing infrastructure continues to expand, the data available, as well as inexpensive computing power, has lowered the barriers to entry to apply the tech to industries. 

Access granular company information and request a demo with Preqin Company Intelligence today. 

The opinions and facts included in the above do not constitute investment advice. Professional advice should be sought before making any investment or other decisions. Preqin providing the information in this content accepts no liability for any decisions taken in relation to the above.