The AI Of The Possible: Developing Scotland's Artificial Intelligence (AI) Strategy

The definition of AI for the purposes of Scotland's AI strategy is easy to understand but lacks some of the depth OECD who define AI as "a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions in influencing real or virtual environments".

It could also be suggested that AI is defined as the theory and development of systems that sense the environment, make decisions and act in a way that would normally require human intelligence.

It is important that the definition of AI recognises both "what" it can do but also "how" it does it in a responsible and trusted way.

The AI strategy should be people-centred and aligned with Scotland's National Performance Framework. Both align with North Lanarkshire's vision of inclusive growth and prosperity for all our communities.

A responsible and ethical AI agenda should be under pinned by a core mission - maximising the social benefit of AI by embedding fundamental human rights as a foundation to ensure moral and legal accountability and the adoption of people-centred AI for the "common good"
Therefore, AI requires a human centered approach to:
1. Ensure that users and humans are front and centre of all decisions
2. Address all human centric concerns such as dignity, freedom, equality, justice
and peace and reinforce fundamental human rights that would foster the kind of
society we want to live in
3. Identify and address grave ethical risks associated with the pervasive and
disruptive nature of AI, as most of the risks associated with AI have ethical
implications on people and the society.
4. Establish a culture of trust and integrity within the AI industry to help reduce
societal friction and anxiety.

Most countries including the US, Germany, Japan, UAE and the UK have underpinned their AI strategy with a people centered mission with particular emphasis on fundamental human rights and ethical outcomes.

A people centered AI strategy needs to be reinforced with appropriate governance and controls to mitigate potential economic, ethical and societal risks.
Aligning AI to Economic Strategy and KPIs. Embedding transformative technologies, such as AI is critical to the long-term success of the Scottish economy. AI has the capabilities to promote a step-change in productivity growth and improve standards of living for everyone in the country. However, to fully unlock this potential, ambition and forward thinking around the topic of AI is required. AI should be thought of as a technology that can play a significant role in helping Scotland achieve all outcomes identified in the national performance framework.

Policymakers, with the input of businesses and citizens, need to establish a clear understanding of the specific opportunities and challenges that AI will bring to Scotland. Developing the knowledge of how Scotland and its people will be affected by AI is a crucial step before considering policy responses. For example, AI has the power to create good jobs with greater earning power, but only if the right skills are developed across the Scottish workforce. This example demonstrates that the task will require collaborative thinking with other bodies such as Skills Development Scotland (SDS).

As part of the UK’s National Industrial Strategy, AI and data are identified as one of the four Grand Challenges facing the UK economy. The UK government wants to put the country at the forefront of the AI and data revolution. To achieve this ambition, government established the AI Sector Deal from both themselves and industry to support the adoption and implementation of AI across the UK.

This paper not only sets out commitments from government but also identifies specific interventions to achieve outcomes across the five pillars of productivity. The Scottish government now have the opportunity to demonstrate how they can play their role in that strategy as well as align their efforts to their own National Performance Framework.

In 2017, Price Waterhouse Coopers (PwC) presented a report on the economic impact of AI. PwC concluded that the economic benefit of AI to Scotland in 2030 was likely to be £16.7 bn with up to £2,209 additional spending power per household. However it should be recognised that there is an opportunity for more than just economic benefits - by targeted application of AI to societal challenges in Scotland it is also possible to improve the quality of life of people in Scotland.

The Art of the possible is compelling, but requires clear objectives and problem statements
The use cases of AI are wide ranging:
● In medical imaging deep learning has been applied to enable early detection of irregular tissues - facilitating better patient outcomes and reducing treatment costs.
● Scotland can leverage AI to design and plan smart cities with higher energy efficiency, increased crime prevention and improved service delivery.
● Within the Agriculture sector, AI can be used to identify and target weeds - reducing herbicide use.

It is important to first identify which societal and people issues need to be addressed to deliver a focused AI strategy with high impact outcomes. Issues that can be addressed by AI include unemployment, ‘in-work’ poverty and health inequality across Scotland all of which will require varied AI interventions.

In order to create shared benefit for all of Scotland's people we recommend adopting a robust approach that addresses:
1. What values should underpin AI for common good and who decides the values?
2. How do we estimate and calculate the social benefits of AI
3. What social policies (e.g., universal basic income and alternative policies) are required to distribute the wealth and benefits generated by AI?
4. What laws do we need to pass to ensure equal benefit access and distribution,
5. How do we empower All sections of the community and population to harness AI and harvest addressable benefits through future workforce planning, re-skilling, creation of new roles and accessible learning.

We agree with the vision and strategic goals as set out in the scoping document. We see potential for the inclusion of a goal related to AI adoption across government services to enable the people of Scotland to be better served by the central and local governments in Scotland.

This is a good summary of Scotland’s AI ecosystem, but would also suggest establishing research and innovation as a separate ecosystem to consolidate activities relating to futuristic applications of AI.

The following areas should also be considered to future-proof the strategy:-

Development of AI and AI Enabled Products and Services:
- Moonshot projects

Ethical and Regulatory Frameworks:
- Citizen monitoring
- Autonomous weapons

Skills and Knowledge:
- University funding

People protection:
- Data security
- Income security
- Digital anonymity

Business protection:
- Local companies
- Specific industry sectors
- Algorithmic governance

Specialised AI Tech:
- Drones
- Autonomous vehicles
- Service robots

Building trust and confidence in AI requires identifying and managing risks and adopting a responsible approach to AI implementation. There are a number of significant risks which need to be managed including:

● Ethical risks
● Economic risks (job displacement, winner-takes-all’ concentration of power risk)
● Societal risks (risk of intelligence divide and autonomous weapons proliferation)
● Performance risks (errors, bias, lack of stability)
● Security risks (cyber intrusion, privacy risks)
● Control risks (risk of AI going ‘rouge’, inability to control malevolent AI)

As North Lanarkshire Council advances its digital transformation and adoption of AI we would look to adopt a “Responsible AI Framework” such as that provided by our Digital Business Partner PwC. This framework will help guide our AI implementation throughout the strategy, planning, ecosystems, model development, deployment and operation & monitoring phases of an AI system. By using practical tools like an Ethical code of conduct, Data bias identification & correction, Adversarial analysis and a Risk and controls matrix we can mitigate against the risks

It is important to consider how AI can transform public service delivery and improve the quality of outcomes and support delivered through local and central government capabilities.
It is also important to learn from the successes that others have had and to accelerate the adoption of AI across Scotland and build trust across the public and private sectors. Scotland’s AI Strategy and approach should provide the opportunity for us to connect with other countries to understand how they are driving forward AI technology. Understanding what others are doing will help Scotland to attract and foster business investment and innovation, educate, train and create a skilled workforce and provide the benefits of AI to Scotland while protecting our society from the risks.

Our Digital Business Partner PwC has been supporting government entities across the globe in the development of National AI strategies (PwC is actively shaping 71% of national AI strategies). PwCs National AI Strategy Radar is used to automatically gather data and to understand government concerns on specific applications, priorities and concerns for AI. Using these resources will help Scotland make the right policy decisions to maximise opportunity and minimise risks.