Tech Index 2022: AI and RoboticsThe business case builds
Increasingly capable algorithms, the proliferation of data and huge increases in computer power mean that the benefits of AI/robotics are no longer theoretical but a reality.
Recent analysis of patent filings1 shows that between Q2 2018 and 2022, the biggest growth in patented technologies was in AI. There were nearly 20,000 AI filings in that time, an increase of 25%, ahead of 5G, which increased by 17%.
It’s clear evidence of just how fast new ideas are coming to the market.
“Companies rank AI/robotics second only to the Internet of Things and connectivity as a key area for potential growth.”
As a result, businesses of all types – from retail to aerospace, life sciences, financial services and fast-moving consumer goods, as well as tech companies – are looking to embrace these technologies to create competitive advantage, or for fear of being left behind.
In our survey, companies rank AI/robotics second only to the Internet of Things and connectivity as a key area for potential growth.
And they continue to see a range of benefits from AI/robotics, including greater flexibility, improved efficiency, faster processes and increased competitiveness.
Cost and skills remain a concern
Some concerns persist, however, with the level of investment required, cost of implementation and proving how AI will provide a return on investment all continuing to be seen as potential drawbacks.
Lack of skilled staff also features high on the list of concerns – and this may link to perceived skills gaps in related areas like data science (see the section on Data Monetization). However, each of these appear slightly less important in 2022 than in 2020, perhaps suggesting that the market is maturing.
Making sense of regulation
One significant concern for many of our clients, but not one highlighted in the survey, is regulation, which is still under development and likely to continue evolving as AI technology advances.
Until recently, the regulatory landscape has been shrouded in fog and it’s been a real challenge for companies to get a handle on the likely shape and scope of regulation in this area. But the fog is now beginning to lift, with the overall approach to regulating AI in different jurisdictions, at least, becoming clearer.
The EU has, for instance, indicated that it favors creating a single, cover-all AI regulatory regime. By contrast, the UK and the US look set to take a sector-by-sector approach.
Many of our clients are now wrestling with how to take account of these differing approaches in their compliance processes.
A priority for many is to put solid foundations in place now in terms of compliance structures, ethics, ways of working, policies and processes that will stand the test of time.
Above all they want to avoid having to rewrite contracts and supply chain agreements in a few years’ time as the fine detail of AI regulation becomes apparent. The memory of adjusting to the EU’s GDPR data protection regime is still fresh, where many were late to implement compliance programmes, and paid the price with extensive remediation exercises later.
Some are taking a bolder approach, however. AstraZeneca, for example, has not only publicly declared how important AI is to its drug discovery and development work, but has set out how the use of AI will be governed within an ethical framework, linked to its approach to ESG and sustainability.
There are real benefits from making the business case so clearly. It empowers people within the business to explore, invest and implement these technologies with a clear focus on achieving a significant return on investment.
Where AI is having most impact
Our survey reveals that IT, customer services and product development continue to be seen as the areas of the business most likely to benefit most from deploying AI/Robotics.
HR is seen as the area least likely to benefit by our respondents. This is surprising finding given that we are seeing significant use of AI in this area, for instance in recruitment and talent management systems.
It is also an area that legislators are focusing on intently given the potential ethical issues around unintended bias in algorithms used to profile individuals and given widespread worries that AI could lead to significant job losses.
Spain, for example, has already introduced legislation giving employee representatives rights to be informed about AI implementation where it may affect work conditions, hiring or job security. The new legislation also sets some parameters around how algorithms are trained when they are used to profile employees.
The split between companies relying on in-house teams to push forward the adoption of AI/robotics and those turning to external advisors for help, remains pretty consistent between 2020 and 2022, although the number looking in-house only has declined slightly from 23% to 20%. Almost half are using a mix of both, compared with 70% in 2018.
In general, the companies that are taking greatest advantage of AI/robotics are those that are working with specialist external providers.
Understanding the deep computer and data science behind these technologies is no mean feat, especially some of the transformative ideas coming out of organizations like OpenAI and Google DeepMind. Even with a high proportion of the technology in this area being released on open-source terms, it is difficult for all but the most expert to keep up with what is currently state of the art.
Companies continue to sign partnership agreements with external suppliers, reserving the right to acquire the business to bring the expertise in-house further down the road. Only an extraordinarily large organization would have the resources to attract and retain the sort of talent needed to deploy cutting-edge AI properly all by itself.
1 Source: GlobalData report on Patent Statistics and Analysis Q2 2022.