Stanford University has released its 4th AI Global Index report. At 220 pages, it’s a big read, so we will be summarising the report’s key findings over the next three weeks. In the following two weeks, we look at the big AI trends the report identifies. This week we look at the current global economic impact of AI.
Follow the money
The total global investment in AI increased by 40% in 2020 relative to 2019 for a total of US$67.9 billion:
- M&A, not private equity, made up the majority of total investment in 2020, increasing by 121.7% over 2019. The report suggests that COVID felled many smaller AI companies, driving industry rationalisation
- More private equity is being invested in AI but in fewer start-ups. Private AI investment increased by 9.3% increase in 2020, compared to an increase of 5.7% in 2019, but the number of AI start-ups funded has fallen sharply, from over 4,000 in 2017 to less than 1,000 in 2020
- The US accounts for over the largest share of investment globally in AI, with China, the next biggest private investor, at less than half the US. However, the US and China are much more evenly matched on public investment in AI. Australia, as would be expected given the size of our economy, is down the rankings, but we are behind other similar economies such as Israel (no. 4), Canada (no.5) and Singapore (just ahead of us).
Not surprisingly given COVID, over a third of total global AI investment went into medically related AI – 4.5 times the level in 2019. Autonomous vehicles, fleet and road AI were the next largest category, with US$4.5 billion. Interesting, there were increases in AI used for language teaching driven mainly by China and in gaming AI driven by US and South Korean investment.
The Global Talent Race
The report developed an AI skill penetration metric that shows the average share of AI skills among the top 50 skills in a wide range of occupations, using LinkedIn data. India has 2.83 times the global average, followed by the United States at 1.99 times the global average, China at 1.40 times, Canada at 1.13 times and Australia lagging at 0.58 of the global average.
While the growth rate in AI jobs in the US was strong at twice of the global average, AI jobs grew the fastest in Brazil (at 3.5 times the global average), India, Canada, Singapore, and South Africa. In a small bright spot for Australia, our AI job hirings also grew by twice the global average. Overall, the global average of AI hirings itself had increased 2.2 times over the AI hirings in 2016.
Corporate use of AI: Missing-in-Action
The report’s analysis of commercial use of AI draws on a McKinsey study.
While neither Stanford or McKinsey put it this way, the ‘sizzle’ of AI at the development and investment level ‘fizzles’ in mainstream business:
- there is no increase in the use of AI in businesses over 2019 levels. Over 50% of survey respondents say that their organizations have adopted AI in at least one business function - in 2019, it was 58%
- companies in developed Asia-Pacific countries led in AI adoptions (at a little over 60%), followed by those in India and North America
- outside companies in high tech and telecom which are, unsurprisingly, the biggest users of AI, both financial services and automotive and assembly sectors ranked the next highest in AI use – and this was also not a change from 2019
- while mentions of AI and other digital technologies in earnings calls have been climbing steeply since 2013, the peak has passed. For example, big data mentions by companies peaked in 2017 and have since declined by 57%. But, if and when companies do talk about digital transformation, there has been a decisive shift to AI, with mentions of AI in earning calls now two times higher than mentions of big data, cloud, and machine learning combined.
In what should be a disturbing finding for corporate risk management and governance, the survey found that “[o]nly a minority of companies acknowledge the risks associated with AI, and even fewer report taking steps to mitigate those risks.”
Top of most respondent’s list of AI risks was cybersecurity – but the problem is that this was the only risk of AI that many respondents gave much thought to:
“Despite growing calls to attend to ethical concerns associated with the use of AI, efforts to address these concerns in the industry are limited. For example, concerns such as equity and fairness in AI use continue to receive comparatively little attention from companies. Moreover, fewer companies in 2020 view personal or individual privacy as a risk from adopting AI compared with in 2019, and there is no change in the percentage of respondents whose companies are taking steps to mitigate this particular risk.”
Most corporate AI investment is into B2C, such as mobile phones, autonomous vehicles or consumer credit approvals. After six years of growth, the number of new industrial robots installed worldwide actually decreased by 12% in 2020. The report suggests that the decline is a product of trade tensions between the United States and China as well as challenges faced by the two primary AI utilising consumer industries: automotive and electrical/electronics.
Despite the economic downturn caused by the pandemic, half the respondents said that the coronavirus had no effect on their investment in AI, while 27% actually reported increasing their investment.
The regulators cometh
Regulation of Big Tech in general and AI in particular is ‘top of mind’ in the US Congress:
“The 116th Congress (January 1, 2019–January 3, 2021) is the most AI-focused congressional session in history. The number of mentions of AI by this Congress in legislation, committee reports, and CRS reports is more than triple that of the 115th Congress. Congressional interest in AI has continued to accelerate in 2020. …during this congressional session, 173 distinct pieces of legislation either focused on or contained language about AI technologies, their development, use, and rules governing them. During that two-year period, various House and Senate committees and subcommittees commissioned 70 reports on AI, while the CRS, tasked as a fact-finding body for members of Congress, published 243 about AI or referencing AI.”
The UK Parliament was not far behind. There also has been a steep climb in mentions of AI by central banks in their pronouncements (although COVID crowded out everything else out in 2020).
So, while mainstream business maybe MIA on AI, governments and policy makers are not.
Read more: Ground the conversation about AI in data