AI investment

April 2017  |  FEATURE  |  FINANCE & INVESTMENT

Financier Worldwide Magazine

April 2017 Issue

April 2017 Issue


Artificial intelligence (AI) presents one of the most exciting and potentially transformative opportunities that mankind has faced in its recent history. In some quarters, it has been heralded as the next industrial revolution.

Through AI, technological developments such as autonomous vehicles, augmented reality and connected homes have evolved from pipedreams into realities poised to disrupt a range of industries in years to come. Companies including Alphabet, Tesla, Ford and Microsoft have made great strides in the commercial application of AI. 2016 saw more than 20 AI start-ups acquired by some of the world’s largest IT players. In May 2016, ARM Holdings announced a $350m deal for Apical. In June, Twitter acquired visual-processing start-up firm Magic Pony Technology for $150m and in August, chipmaker Intel announced that it had acquired Nervana, a startup that has been developing AI software and hardware, for more than $350m.

AI will be a key investment target in the coming years, with myriad organisations hoping to capitalise on its potential. In 2016, the AI market was worth just $644m, according to Tractica. In 2017, however, the market will likely double and then experience exponential growth. The market for the enterprise application of AI could surpass $30bn by 2025. The potential offered by the AI industry is a direct corollary of the way people are embracing technology and data. With nearly 40 percent of the world’s population now able to access the internet, and 8.4 billion connected devices set to be used globally in 2017 according to Gartner, up 31 percent from 2016, the amount of data assets being created will facilitate the growth of AI. As the cost of data storage decreases and learning systems, architecture and software infrastructure become more advanced, the promise of AI becomes clearer.

A maturing market

The AI market, though still nascent, is maturing rapidly alongside the growth in computer processing power. According to Moore’s Law, the processing power of a $4000 computer will surpass that of the human brain in 2019.

As the industry evolves, certain sub-sectors have emerged. Enterprise AI, for instance, is rapidly developing in areas such as healthcare diagnostics, automatic trading, business processing and advertising. “The market is evolving very quickly, and fragmenting into specific, focused use-cases,” explains Daniel Domberger, a partner at Livingstone Partners. “From consumer-facing online ‘chatbots’ to Watson’s tax advice and more broadly to self-driving cars, the excitement has moved from AI as a generic solution to its applicability to specific situations which are more or less well-defined – such as tax – or open-ended – such as navigating a car alongside other vehicles driven by humans. The areas of focus are often complicated niches with direct monetisation potential or cost saving opportunities through streamlining and automating existing processes. Or they are blue-sky transformational projects which redefine our understanding of the product or the problem, such as self-driving cars.”

It is these ‘blue-sky’ ideas that are likely to drive the industry, and be the catalyst for future investment.

R&D

Given the quantities of data being generated annually by smart devices and the Internet of Things, an increase in AI R&D was inevitable. Already in 2017, both Microsoft and Google have committed additional resources to their R&D departments. In January, Microsoft announced that it was following up its acquisition of Canadian language processing startup Maluuba by doubling the size of its R&D offices in Montreal, Canada and by investing $7m in academic research around AI.

In the long term, AI is expected to have a key impact on R&D processes – though the initial cost outlay will be significant, according to Alaistair Peet, a partner at Shoosmiths. “As AI develops we are beginning to see almost a chain reaction across various industries. The cost of AI investment for companies is and will remain high and there is no getting away from that. But AI is a necessary cost many companies across various industries such as law, accountancy and manufacturing will have to bear in the future to stay competitive,” he says.

However, other organisations will look to grow their offerings though M&A, rather than develop products and services in-house. With increasing levels of competition for the best assets, dealmaking momentum is likely to continue. Before startups can attract suitable would-be acquirers, however, expect many smaller firms to look to venture capital investments to help them grow and become more attractive targets.

AI will be a key investment target in the coming years, with myriad organisations hoping to capitalise on its potential.

Yet, certain countries are encouraging firms to engage in R&D. The UK, for example, introduced an R&D tax credits scheme for small and medium enterprises in 2000 and for larger companies in 2002. In the US, in October, the outgoing Obama administration’s National Science and Technology Council Subcommittee on Machine Learning and Artificial intelligence launched the ‘National Artificial Intelligence R&D Strategic Plan’ which established a set of objectives for federally funded AI projects related to research occurring both within the government, as well as federally-funded research occurring outside of government. The ultimate goal of these objectives was to produce new AI knowledge and technologies that provide a range of positive benefits to society, while minimising any negative effects. The Subcommittee proposed seven objectives: (i) to make long-term investments in AI research; (ii) to develop effective methods for human-AI collaboration; (iii) to understand and address the ethical, legal and societal implications of AI; (iv) to ensure the safety and security of AI systems; (v) to develop shared public datasets and environments for AI training and testing; (vi) to measure and evaluate AI technologies through standards and benchmarks; and (vii) to better understand the national AI R&D workforce needs.

R&D investment in AI is a packed playing field. To stay ahead of the competition, companies are investing considerable resources – both financial and human – into their R&D agenda. However, more needs to be done. According to the Obama administration, in the US, “current levels of R&D spending are half to one-quarter of the level of R&D investment that would produce the optimal level of economic growth”.

Risk and reward

AI’s ability to disrupt and remake entire industries is a key driver of investment. However, it has also led to the creation of entirely new industries. Driverless cars and personal assistants such as Amazon’s Alexa and Apple’s Siri, for example, are indicative of AI’s ability to break new ground.

But with the rewards come considerable risks. As Shigeki Tatsuno, a partner at Anderson Mori & Tomotsune, notes, in the AI industry, as with many other sectors, cyber security must be a key concern. “Notwithstanding the general optimism about AI’s effectiveness in improving defence against cyber attacks, some have cautioned that overreliance on AI could heighten vulnerability to information leakage and unauthorised access to confidential data,” he says.

AI’s relationship with cyber security is complex. Many cyber criminals use automation processes to exploit network weaknesses and launch attacks. Yet, AI can also revolutionise cyber security provisions. Machine learning, for example, can be a useful tool for companies in the fight against hackers. AI can be trained to learn from patterns of attack in order to identify and respond to any deviations. It can differentiate harmful activity on a system or network from permitted activity by making inferences from historical data. This data can be used to improve both the AI’s own functions and company’s cyber security provisions much faster than human responses. The future of cyber security is preventing attacks, and the learning capabilities of AI are complementary to this goal. Given the potential applications, it is little surprise that a number of AI cyber security startups have emerged. Moreover, these firms are attractive acquisition targets. In early February, for example, harvest.ai was acquired by Amazon Web Services for around $20m. This deal was likely just the tip of the iceberg. A 2016 report from 451 Research suggests that startups which specialise in AI and machine learning will be top acquisition targets in 2017, particularly for chip manufacturers, software firms and the automobile industry. ‘The Visionary Innovation (Mega Trends) report’ from Frost & Sullivan also named AI as the top investment trend for 2017.

As investment continues, we are entering a crucial period for the regulatory environment surrounding AI and cyber security. The European Union’s General Data Protection Regulation (GDPR), will create difficulties for companies that rely on gathering and processing user data for their businesses. The GDPR’s stance on profiling, which is essentially the ability of companies to use automation to determine certain characteristics of their individual users, may challenge AI. Once the GDPR comes into force in 2018, companies will need to manage and store their data in a compliant manner.

However, though companies will need to remain compliant with cyber security provisions in the EU and elsewhere, Jonathan Epstein, the chief marketing officer at Sentient Technologies, believes it is unlikely that the AI space itself will generate much in the way of specific legislation. “The industry is stepping up into self governance and almost every practitioner in the field has a strong belief in the need for ongoing discussion and self-regulation about the potential impacts of AI, whatever they might be. We do think that AI will cause a rethink of the regulatory regimes within the industries it transforms,” he says.

Long-term gains

AI could transform virtually every industry. The number of applications will continue to grow and investors are looking to enter the AI industry early. But, at the start-up stage at least, AI investment can be risky. “The risks to investment in AI start-ups are clear, because it is so capital intensive,” says Mr Peet. “To get cutting-edge AI products off the ground requires huge cash injections and it is not a sector that can be monetised very quickly or easily, either. Often these products require R&D teams and specialist engineers in machine learning, which would come at a cost. But, with great risk comes great reward. AI products will truly alter the fabric of how we live our lives, through virtually everything we touch. Leading businesses in most industries are now looking towards AI to transform their products and gain market share – and are willing to pay accordingly.”

Though it may take some time to see returns, the potential is there. “If a company is embracing the opportunities AI and technological advancement brings, the initial large investment is likely be paid off by longer term rewards,” says Nick Winters, head of technology at Kingston Smith. “As part of its competitor review process, a business will have an idea of the current technology being used in its industry, and any potential developments. A business should aim to continuously review and assess the technology it uses, determining whether it is fit for purpose and to provide high quality products and services to its clients.”

With competition for assets set to be fierce in the years ahead, AI will be an influential industry. Given the progress made in the sector, 2017 could be the year of AI.

© Financier Worldwide


BY

Richard Summerfield


©2001-2017 Financier Worldwide Ltd. All rights reserved.