Rethinking automation

Introduction
SCVO is keen to help the Scottish third sector think about the implications that artificial intelligence (AI), automation and current and future technologies may have on the sector’s work. We’ve created this page as a way of engaging on these issues. The aim of this engagement is to create a briefing on the key opportunities and challenges for colleagues across the sector.

Context
Our world is changing. Artificial Intelligence (AI), automation, and new technologies are already impacting our organisations, our jobs, and our lives. But what does the future hold? Will new technology enable the third sector to provide better services, support and campaigns? Can automation reduce inequalities and help build a better, greener society? And could a robot take your job?

What is Artificial Intelligence (AI), automation, and new technologies?
AI is the concept of non-human entities possessing human-level intelligence to perform intelligent tasks. For many of us this means technology focused on self-learning to mimic the human mind. We are likely still decades away from machines capable of mimicking humans on the entire range of tasks, known as, general artificial intelligence. In various fields, however, there have been rapid improvements in the performance of AI-powered technology. Self-driving cars, virtual personal assistants, and smart home devices are all examples of AI. Similarly, fraud detection services, purchase predictions, image and voice recognition software, and even some security surveillance, all utilise AI.

AI technology has and continues to open new markets and new opportunities for progress in areas such as health, education, energy, and the environment. This briefing aims to prepare the sector for the growing role of AI, automation and other new technologies in society. Over the next decade and beyond, these new technologies have the potential to be a valuable tool. However, SCVO and our members believe that these tools can only realise their potential if the third sector, wider civil society, government, and industry work together to develop the technology, manage its challenges, and ensure that everyone can equally take advantage of its benefits.

Political Context
The UK Government estimates that by 2035 Artificial Intelligence could add around £630 Billion to the UK economy. To capitalise on this and secure Britain's position as a world leader in technology and innovation Chancellor Philip Hammond’s 2017 Budget included: £75m for artificial intelligence; £160m for 5G mobile networks; £100m for an additional 8,000 fully qualified computer science teachers supported by a new National Centre for Computing; a retraining partnership with the TUC and the CBI to boost digital skills in the workforce; and £76m to boost digital and construction skills.

At Westminster, an All-Party Parliamentary Group on the Fourth Industrial Revolution (4IR APPG), Chaired by Alan Mak MP, was formed in October 2016. Similarly, the Lords Select Committee on Artificial Intelligence was appointed in June 2017 to consider the economic, ethical and social implications of advances in artificial intelligence, and to make recommendations. In 2017, Artificial Intelligence was also identified as one of the ‘Grand Challenges’ in the, Industrial Strategy: Building a Britain fit for the future.

At an EU level, AI is one of European Commission Digital Vice President Andrus Ansip’s three priorities for the rest of his mandate. The Commission are expected to outline an AI initiative in April 2018. The Commission previously published the Digitising European Industry strategy, launched in April 2016, identifying Robotics and Artificial intelligence as cornerstone technologies.

How are Artificial Intelligence (AI), automation, and new technologies likely to effect the third sector?

 * Offer new ways for charities to achieve their mission
 * Change the way charities operate as organisations
 * Present new challenges for charities to address.

Challenges
Artificial Intelligence can solve many societal challenges but it can also become a societal challenge itself. AI raises many ethical and philosophical questions: who will be liable if algorithms make a mistake? Will systems be transparent? Who benefits? Values and principles must therefore be embedded in every application of this technology.
 * Upskilling new and existing staff to understand the potential of    AI, new technologies and automation in their role and workplace will be a     key challenge for the sector. The sector, like other sectors, will     require data literate staff with the skills to read, use, interpret, and     communicate about data, and participate in policy debates in areas     affected by AI.
 * The third sector must ensure that new technologies will benefit    society as a whole.
 * A few private companies control much of the knowledge about new    technologies

Inclusive AI and Algorithmic fairness
The advances in AI in the past few years have enabled the automation of decisions in a range of areas: education; hiring; lending; and criminal risk assessment. For example, machine learning can be used to make predictions in job candidate screening, issuing insurance or making loan approvals. These decisions effect people's lives. As is the case when human judgement is applied, these decisions are not free from bias.

AI is designed by people, as such it is reflective of its creators. It is important that the demographics of AI creators reflects the demographics of people from across society who will use this technology. Ultimately, this should be everyone. The creators of new technologies do not, however, reflect the diversity of our societies. Currently, only around 13.5% percent of those working in machine learning are women (Women in Machine Learning). Bias in training data provided by creators can therefore create biases that result in judgments that are discriminatory. For example, AI programs could recruit staff in the image of their creators. Some examples of this already exist: In a recent investigations ProPublica found predictive software used by US courts was biased against African Americans, calculating their risk of reoffending at almost twice the rate of white defendants.

The third sector must be aware of the potential impact of these biases both upon their beneficiaries and when utilising new technology.

Potential Roles for the Third Sector

 * Ensure that everyone can benefit from emerging technology
 * Build public trust
 * Seek alternative pathways to make use of new technologies by creating space to connect the knowledge already existing within communities (collaborative innovation)

Third Sector Initiatives
In March 2017, Arthritis Research UK introduced an AI-powered ‘virtual personal assistant’  to answer beneficiaries questions quickly and easily. Initially the virtual personal assistant will provide general information about arthritis and exercise. However, this is expected to be expanded to include answers to a broader range of questions on subjects such as diet and treatment options.

Age UK recently completed a pilot program with Accenture that uses artificial intelligence (AI) and voice enabled technologies to help older people navigate their care delivery and well-being. The AI-powered platform can learn behaviours and preferences, allowing the platform to spot abnormalities in behaviour and send alerts.

Useful Resources
Gunson, R and Thomas, R. (2017). Equipping Scotland for the future: Key challenges for the Scottish skills system, IPPR Scotland https://www.ippr.org/publications/equipping-scotland-for-the-future

Corlett, A. (2016). Robot wars: automation and the labour market. Resolution Foundation http://www.resolutionfoundation.org/publications/robot-wars-automation-and-the-labour-market/

Executive Office of the President National Science and Technology Council Washington. (2016). Preparing for the Future of Artificial Intelligence. https://obamawhitehouse.archives.gov/sites/default/files/whitehouse_files/microsites/ostp/NSTC/preparing_for_the_future_of_ai.pdf

House of Lords Select Committee on Artificial Intelligence (2018). AI in the UK: ready, willing and able? https://publications.parliament.uk/pa/ld201719/ldselect/ldai/100/100.pdf