AI, data and design skills: outputs from a Design Council workshop

Modern design is no longer confined to particular sectors or occupations. The skills, principles and practices of design are widely used across many parts of the economy, from protecting public health, to guiding the growth and development of our towns and cities. Good designers put people first and use their creativity to solve problems, challenge thinking, and make lives better.

But the design profession is changing. Big data and advances in artificial intelligence (AI) have greatly changed how people interact with the world, from how we shop to how we access healthcare. Designers are acknowledging that we are complex beings, in relationship to one another, and not separate from the planet in which we all live.

This work builds on our 2017 report Designing a Future Economy which investigated the skills used in design, the link between these skills and productivity and innovation, and how they align with future demand for skills across the wider UK economy. To explore these trends further, we commissioned some speculative research looking ahead to 2030. We then ran a speculative design workshop on 9 January 2020 to think about the future and the challenges of today around the untapped potential of design skills.

A multi-disciplinary approach

The workshop brought together a multi-disciplinary group of AI specialists, designers, data scientists and people working in the retail, planning and healthcare space. Some had experience of taking products or services to market. Others were finding better ways to help people live healthy and happy lives, or investigating people’s emotional experiences of the built environment.

We focused on retail, town planning and healthcare because these are areas in which we are active at the Design Council (for example, see our work with the High Street Taskforce, on Architecture and the Built Environment and Healthy Placemaking).

We also invited illustrator and mural artist Cristina Guitian to represent the discussions visually, to make them easier to understand and modify and to stimulate new ideas. Her illustrations are featured throughout this blog post.

Speculative design

“While science aims to explain how things are, design aims to explore how things should be by finding a solution to a problem and improving the current status quo.” Nesta

We decided to take a futures approach to explore the role of tech/digitation, big data and AI and what this might mean for the skill sets that people need.

The designers at the workshop used provocations as a starting point to consider how retail, town planning and healthcare could change over the next decade. We used a matrix and they came up with a number of scenarios with distinct and divergent visions of the future. These were not exhaustive portrayals, but they helped to capture a wide range of outcomes and present them in a way that is easy to imagine.

Some discussions went further than others! For example, the retail group all had different preconceptions of AI and decided to use it as a proxy for a broader discussion about data, digitalisation, automation and technology. The planning group felt that their sector wasn’t even in the digital space yet, let alone looking to AI solutions.

Below is a summary of the discussions that we had at the workshop, including where we are now within each sector, and where we might want to go.

RETAIL

Where are we now?

Data has long been used to understand consumer behaviour. But, perhaps more so than other industries, retail brands are moving aggressively forward with a range of data and artificial intelligence projects. Our provocations and discussions highlighted the following developments:

• Some fashion subscription boxes use a combination of artificial intelligence and style experts to select new outfits.
• There is an “epidemic” of facial recognition surveillance technology at privately owned sites in Britain. There is little discussion of what racial and gender bias these technologies may perpetuate, and particularly how this will affect people of colour.
• Design invention studio COMUZI invented the “Invisible Mask” to confuse the computer vision models and make the face unidentifiable.
• Usage of Amazon’s self-service parcel delivery service is growing steadily, in part due to the introduction of AI-driven enhancements to improve the volume of successfully filled deliveries.
• Could these go one step further and use unsupervised learning (a type of machine learning algorithm that is given no labels, leaving it on its own to find structure in its input) to predict what consumers might need?

Where do we want to go?

In this scenario:

• Good design has been redefined to incentivise quality products over profit margins
• Retail brands inspire people to make more sustainable choices
• AI is used as a tool to offer substantial improvements in product design, operations, and infrastructure optimisation
• Consumers seek products that are zero-waste, reusable and repairable

Scenario #1: Coming Full Circle

In the early 2020s there was an AI winter sparked by social backlash. Back then, AI was still advancing through deep learning and machine learning, building up systems from the bottom by training them on mountains of data. These data-hungry systems led us to the ethical and moral dilemmas that reached a head in 2022, when extinction rebellion and ‘techlash’ came together to force a roll back of growing tech intensity that was threatening our planet.

The consequences were drastic. The economy shifted from a linear to a circular system, aimed at eliminating waste and the continual use of resources. Retail companies overhauled their sustainability efforts, facilitating the simple return of their products and packaging at the end of what their owner thinks is their useful life. Recycling became the norm and retailers leveraged the reuse of materials as new products as part of their commercial business. It became unacceptable for companies to offload unwanted materials. Consumer behaviour completely shifted too. Consumers saw ‘cool’ products as those that no longer contribute to global warming.

By 2028, the AI winter thawed as we realised that AI could help us be more effective. Combining the power of AI with a vision for a circular economy represented a significant, untapped opportunity to harness one of the great technological developments of our time. But, as creating an AI was five times worse for the planet than a car, this time around we chose to apply it more wisely: to design better and consume less.

TOWN PLANNING

Where are we now?

In stark contrast to the retail industry, the attendees at our workshop felt that the planning sector has been slow to capitalise on big data and AI. The planning system is ripe for digitisation and machine-learning, but the first step is supporting the digitalisation of materials — for example, making documents in a way that can be machine read. They felt:

• There is increasing financial pressure on councils and communities
• Planners face the practical challenge of interpreting growing stores of data from increasingly smart, connected environments
• Some data-driven companies (like Sidewalk Labs) argue that planning is unnecessary in an era of AI, automation and responsive, smart cities.

Where do we want to go?

Some academics and businesses are exploring the ways in which AI can help us better understand our cities, uncovering information that was previously too laborious to quantify, and producing new design solutions. And with this knowledge, new opportunities for designing urban spaces will emerge.

In this scenario a town planner is facilitating a discussion between the community and some built environment professionals. The planner is presenting designed scenarios for discussion and they are all using VR headsets to visualise what their local built environment/community could be like. The community have provided their own data e.g. on local road use and commuter behaviour.

In this vision, town planners:

• generate a shared vision for a place, that is owned by the community and realised by the planner;
• experiment and take risks to generate different scenarios for their town;
• navigate complexity in an uncertain environment;
• are data gatekeepers, helping to make data easily accessible and draw out facts, figures and stories from the community;
• scrutinise the decisions being made by algorithm, especially those that are making consequential decisions in society or producing unexpected outcomes; and
• have knowledge of local communities and strong relationships and ties.

Scenario #2: Shared Visions

In 2030, digitisation of processes and AI have freed up town planners from much of the paperwork that used to be associated with assessing and pushing planning applications through the system. Simple applications, such as requests for extensions, are all automated through AI-enabled technology. Complex applications still need a level of human scrutiny, but they are less time consuming than they used to be. This is primarily because planners play a different role in the run up to decision making. For example, AI-enabled technologies check off the basics in each application, so that planners can shift their focus away from outputs (e.g. numbers of applications processed) towards outcomes (e.g. how quality of life can be enhanced by the decisions being made, does the scheme fit with the shared community vision etc).

HEALTHCARE

Where are we now?

The idea of doctors ‘being replaced by machines’ in widespread, but this can obscure the many positive applications of AI. For example, it can free up time for specialists to meet with patients and help doctors to diagnose disease. Researchers at Google and Moorfields Eye Hospital have developed deep learning algorithms that can automatically detect eye conditions with a great deal of accuracy.

But to get there, there are also the privacy and data hurdles. The use of big data analytics and new technologies in the health sector has considerably changed the way health data is being used, accessed, analysed and shared between health professionals and individuals. Organisations handling health data that embrace these new techniques and practices have to maintain a high standard of security and privacy.

There are also some key challenges currently facing the NHS, such as the ageing population, evolving healthcare needs (e.g increasing cases of obesity and diabetes), or antibiotic resistance and closure of local services due to centralisation drives.

Where do we want to go?

In this scenario, a healthcare worker is using AI to identify those at risk of developing medical issues. The technological advancements allow the healthcare worker to spend more time on the ‘human’ aspects of care, such as helping patients digest and cope with bad news.

Healthcare workers of 2030:

• use AI to identify those at risk of developing conditions;
• treat the cause as well as the symptoms;
• coach patients about different choices to adopt healthier behaviours; and
• translate the evidence base on what works and being a bridge between the public and research.

Scenario #3: Predictive precision

AI is being used to learn and identify those who are risk of developing conditions, as well as learning what motivates different types of people to take up preventative, healthy behaviours. Design is built in throughout, for example:

• Into how data is collected, seeing data donation as an experience that can be improved, and clearly visualising the benefits of providing a more holistic data profile.
• Into how the models are created, using human insight to widen the data variables that help computers understand different contexts.
• And into designing the types of intervention at the end, which are empowering and based on a human (or human-like) conversation at the end, which coaches the ‘patient’ about different choices they have available to adopt healthier behaviours.

Machines are helping us understand our own data (and health), and translating the large evidence-base of what works into easy action we can take, depending on our geographic location, interests, aspirations — and links us up with people in similar situations to create communities that support and motivate each other. AI is able to identify, triage and guide people through 90% of decisions, leaving health professionals to lever in their intuition, ethical discernment and feelings — alongside clinical practice — in their toolkit. As people are living for longer, there is a need for more conversations about living wills and good deaths.

Fostering design skills

Where are we now?

The UK has transitioned from an economy powered by might and machine, to one increasingly powered by services and technology. Design has played a key role in all of these developments, evolving with economic shifts, boosting productivity and instigating innovation — from the industrial designers that pioneered post-war aviation to those designing robots and artificial intelligence today. But the design profession today is in a state of flux. As the fourth industrial revolution takes hold, boundaries continue to blur between disciplines.

We see humans using their design skills alongside machines to be more effective, productive and innovative workers. But much of the design economy’s potential to contribute to future UK growth remains untapped, and there is an emerging risk of growing inequality between firms accessing design and those that do not, as well as between people who have such skills and those who don’t. As we found in our 2018 Design Economy report, design can generate significant value for local and regional economies: London remains the powerhouse of UK design, with almost one in three design firms now based in the capital.

In the last few years there has been a major decline in the take up of subjects like Design and Technology. GCSE take-up of D&T has reduced from over 400k in the year 2000 to just under 100k today. Between 2011/12 and 2015/16, the number of people leaving Higher Education with undergraduate or postgraduate qualifications in Creative Arts and Design subjects fell by 7%.

The education system in the UK is very narrow. Subjects are not taught in an interconnected way or linked across different disciplines. Pupils will not learn about the theory and application of the design of products, services and environments.

Finally, while many projections of how many jobs will be lost, gained, or changed by AI have been published over the last five years, a consensus has begun to emerge that 10–30% of jobs in the UK are automatable. Many of the skills the workforce of the future will need are closely correlated with design, and creativity is likely to be even more important in the future job market.

Where do we want to go?

Workshop guests wanted to see more “specialist generalists” (it does not matter whether these people identify as designers or data scientists or AI specialists)! These people will support the culture change that is needed to define roles in an era of greater automation. They also wanted to see multi-disciplinary teams; evangelists for design and the power of technology; and translators and honest brokers to act as intermediaries. They wanted to see technologists with a great understanding of design processes (perhaps using design to revisit the process by which data is collected, and models and interventions are designed).

Finally, they wanted to see more personal accountability for unintended consequences of poor design, more discussion on we are measuring if we aren’t measuring profit (satisfaction? or levels of personalisation?) and a good enough understanding of data and AI to be a critical client.

This article was originally published on Medium