AI M&A just got real
With 1,114 TMT deals in the US in 2016 worth US$363 billion, the sector outstripped other industries as dealmakers pursued innovative intellectual property and battled to stay technologically relevant in an ever-advancing digital economy. While technology megadeals such as Microsoft’s US$25.5 billion acquisition of LinkedIn hogged the headlines, M&A in the nascent but rapidly growing AI arena fired the imaginations.
AI deals on the riseAI is the field of technology that enables machines to conduct operations that previously only humans could perform (for more on AI, see The lowdown on AI, below). In the past five years, investment in AI has rapidly accelerated. Research group CB Insights found that there have been 140 acquisitions of AI companies since 2011, with well over a quarter of these (40) in 2016 alone—an eightfold increase in AI M&A since 2011. Digital media company TheStreet reported that since 2011, a total of 1,098 deals attracted US$6 billion in global AI equity funding. Equity financings in AI climbed by 746 percent between 2011 and 2015, from US$282 million to US$2.4 billion—paving the way for a rush of M&A activity in this area. According to Arlene Hahn, a partner at White & Case: “It is not just the tech that is driving AI deals, but also the talent. Demand for top teams means that ‘acqui-hires’ [deals based on talent acquisition] are very much part of dealmaking strategy.” However, Hahn notes, “AI talent comes at a cost”. Twitter’s 2016 US$150 million acquisition of AI startup Magic Pony far exceeded the industry norm—landing at US$10.7 million per each of its 14 employees.
Why AI?Although modern research on AI has been underway since the mid-1950s, the last decade has seen prodigious leaps in the field. A number of factors have contributed to this growth.
The first is Big Data. Data is the lifeblood of AI and the success of any machine-learning technology depends on the quantity and quality of available data. Advances in the widespread adoption of Internet of Things (IoT) devices have resulted in exponential growth in data generation. As smart and connected devices rapidly infiltrate our offices, homes and cars, the volume of data multiplies. Indeed, researchers at the International Data Corporation have predicted that the total amount of digital data created globally would explode from 4.4 zettabytes in 2013 to 44 zettabytes by 2020 to 180 zettabytes by 2025. Putting that into perspective, 1 zettabyte is equivalent to around 152 million years of high-definition video.
AI deals—need to knowAI is becoming a central pillar of business across all sectors, and companies need to start investigating how they can apply it to their businesses. When it comes to absorbing AI into your business, two key considerations are whether you develop or acquire it, and ensuring data quality. AI is a highly specialized field. This means that talent is limited. William Choe, a partner at White & Case, notes: “Companies considering AI should determine whether they have the resources in-house to build and deploy deep learning or other AI technologies. If not, an acquirer will need to consider investing in talent via acqui-hires and/or acquiring partially developed AI technologies on which a platform may be built in order to augment its internal expertise and capabilities.” On top of this, the success of AI depends on the quality and quantity of its data set. AI can not only analyze and review structured data (i.e., data that is easily held in traditional databases such as payment ledgers) but also unstructured data (i.e., data that is not easily organized into traditional categories, such as the human language). And businesses now hold massive amounts of both data types. However, if companies want to capitalize on this data, Hahn notes, “first they need the technical capability to capture data in a way that informs cognitive analytics. And, second, they need to have sufficient legal procedures in place to ensure that its data collection is compliant with security regulations.”
The future is nowFrom chips to devices and analytics to applications, AI is infiltrating every segment of our lives, and therefore, the market. And it’s not just for tech titans anymore. Even industries that are traditionally low-tech can leverage AI technologies, such as natural language processing and predictive analytics, which can boost productivity and increase efficiencies. However, businesses will need to move fast because AI startups will be highly coveted acquisition targets in 2017.
AI deal open to allGiven the growth in the market, AI deal activity is poised to surge in 2017 and beyond. A survey by research firm Forrester indicated that there will be more than a 300 percent increase in investment in AI in 2017 compared with 2016. Meanwhile, Accenture recently forecast that the market for AI would grow from US$4.5 billion in 2014 to US$9.2 billion in 2019. However, the value that AI can deliver is by no means limited to technology companies. Large industrial organizations such as GE have been investing in emerging AI companies such as Bit Stew, whose technology is geared towards utilities, oil and gas, aviation, and manufacturing, to bring machine learning to the Industrial IoT. Furthermore, other startups such as BenevolentAI, are pushing deep learning technologies into new territories, including pharmaceutical R&D. As the collection and analysis of data becomes an integral part of all businesses, from fitness clubs to logistics companies to retailers, the reach of AI’s application to a wider range of industries increases—and therefore, more deals for its capabilities will inevitably follow.
The lowdown on AI
Artificial Intelligence (AI):AI (also known as cognitive computing) is technology that enables machines to perform tasks that previously required human intelligence, such as speech recognition, visual detection, learning, decision making and problem solving.
Machine learning:This is a subset of AI in which algorithms review and analyze data in order to make predictions or determinations. Rather than being limited to programming instructions, machine-learning algorithms improve and “learn” as they get exposed to more data.
Deep learning:This is a type of machine learning that uses networks of algorithms to simulate the multilayered neural networks of the human brain. Unlike machine-learning algorithms, deep learning networks are capable of “learning” from raw data itself—with little instruction from the programmer—and to continually scale with larger data sets.
While technology megadeals such as Microsoft’s US$25.5 billion acquisition of LinkedIn hogged the headlines, M&A in the nascent but rapidly growing AI arena fired the imaginations
Meanwhile, advances in chip technology and computational power have enabled machines to quickly and efficiently mine and analyze this data and to deploy deep learning networks. And innovations in the semiconductor industry continue to drive growth. Major chipmakers are making acquisitions in the deep learning chip technology space, while several startups are hoping to disrupt the industry by creating new chips that they claim are faster than graphic processing chips (the standard for implementing deep learning algorithms due to their high speed), and more efficient than field programmable gate arrays (popular for energy efficiency and adaptability).
In addition, social changes have contributed to the rise in the popularity of AI. The use of data-collecting applications and devices has mushroomed in recent years with the ubiquity of social media as well as innovations such as Apple’s Siri, Amazon’s Alexa and Fitbit, and services such as Uber and ZocDoc. Just like automated service providers and the cloud before it, the rapid adoption of AI into the daily lives of consumers has created an environment ripe for further development and investment.
SHARE THIS ARTICLE
Meet our partners
SHARE THIS ARTICLE
Breakthroughs in deep learning are set to drive a surge in M&A activity in the artificial intelligence (AI) space in 2017 and beyond
RECEIVE OUR PUBLICATIONS