According to the Guardian, Web Summit is “Glastonbury for geeks”. This year Web Summit – an annual technology conference in Lisbon, Portugal – convened over 60,000  technology nerds people and more than 1,200 speakers from 6 – 9 November.

I was one of those  technology nerds people.

I’m a Transformation Manager at Methods, specialising in digital transformation and – more recently – figuring out what the ethical application of Artificial Intelligence (AI) can look like. I’ve just completed a MSc in which I conducted exploratory research examining the relationship between AI and gender stereotypes. As part of the research I created and successfully tested a design process to for creating ethical, AI-powered chatbots. My work at Methods is focused on how we can harness AI to support effective public services.

The topics at Web Summit this year ranged from driverless cars, the role of government in promoting innovation, what’s next in data science to the ethics surrounding robots (a live conversation between two Hanson Robotics robots was particularly unnerving). Speakers included the European Commissioner for Competition Margrethe Vestager, CEO of Slack Stewart Butterfield and Sairah Ashman, CEO of the brand consultancy Wolff Olins. In addition, there was a very strong focus on AI at the conference – what it is, how it’s transforming business and technology, and what we can expect from it in the future.

 

As is the dilemma at any interesting conference, there were too many brilliant speakers and events to go to and so I suppressed my fear of missing out focused primarily on AI and its likely impact in the next few years. In between researching where to find the best pastéis de nata, I attended sessions like ‘Designing AI into your business’ by Kevin Bandy from Cisco Systems, ‘If it looks like a human and sounds like a human’ with x.ai CEO Dennis Mortensen, and ‘Future applications of AI’ by Jérôme Pesenti from BenevolentAI. Below is a summary of what I believe to be three key themes we need to consider as we apply AI to our public services: augmenting human capabilities; developing AI for all; and the immediate opportunities for applying AI.

 

Augmenting Human Capabilities

A lot of people when they hear the words ‘Artificial Intelligence’, immediately react with ‘robots stealing my job’ or ‘end of the human race’. At Web Summit a lot of speakers considered instead how AI can be used to augment human capabilities and decision making, and what this might mean for the future of work. As with any previous technological revolution, technology has been most useful and empowering when applied to either reducing or eliminating the need for humans to do work that we don’t particularly like (i.e. repetitive, manual labour) or that augments our ability to apply our very human skills of empathy, critical analysis and problem solving (i.e. social and collaborative tools like Slack or Basecamp).

 

Speakers at Web Summit encouraged us to imagine how AI can augment the capabilities of an existing workforce or free people from mundane tasks so they can focus on more strategic objectives. This is especially important for government services. For example, if a repetitive process which follows a well-defined decision tree can be automated – how does that free up or augment people’s ability to spot fraud, solve upstream issues with another part of government, or respond more effectively to complex, out of the ordinary cases? AI is less about replacing humans, and more about augmenting the impact humans can have when delivering services and solving problems.

 

Developing AI for all

A second, recurring theme was democratising how AI is developed and applied. First and foremost, speakers like the European Rapporteur for Artificial Intelligence Catelijne Muller discussed how AI needs to be developed with everyone in mind. Max Tegmark (Future of Life Institute) and Sairah Ashman gave examples of the Asilomar AI Principles and the recent Copenhagen Letter, respectively, which are commitments to applying AI to benefit humanity. While these principles and commitments are important, it is still too early to see how they translate into new types of products or services, or challenge the current commercialised forms of AI being developed (for example, I’m not convinced Amazon Echo Look solves a meaningful human problem). In terms of the types of AI products and services that are built, Government has a role in shaping the kind of AI it requires. AI needs to be able to align with the Government’s unique responsibilities, such as safeguarding citizens’ digital identities or solving wicked policy issues.

 

In addition, speakers such as Danny Lange from Unity Technologies demonstrated how open source tools and platforms are critical for ensuring the actual development of AI can be influenced by everyone. My own research into designing ethical chatbots also illustrated this point – AI needs to be developed by both technical specialists (i.e. data scientists, engineers) and people who understand the potential social, cultural and political impacts (i.e. policy makers, social scientists). By building or applying AI without both these perspectives, we are building it blind – blind to either the human consequences of utilising the chosen technology, or blind to the technological advances available to us to solve a particular problem.

 

Immediate opportunities for applying AI

One of the best things about going to an industry-wide conference like Web Summit, is that it’s possible to hear from industry leaders what the immediate opportunities and trends are. First, Cisco’s Kevin Bandy extolled the virtues of a ‘just get it started’ approach to applying AI, encouraging the audience to aggregate their data in order to begin automating routine, business processes. Second, as AI technology such as natural language processing becomes more sophisticated we will find more and more ‘voice first’ technologies like Amazon Echo on the market. What this means for user interaction design is that voice commands are likely to replace gesture or keystroke commands.

Third, as AI services become more sophisticated, Bandy predicted that it will enable us to deliver mass personalisation of services to individual customers, enable real-time customisation to their needs, and even anticipate what they might require. The combination of the second and third opportunities will see us move into an era where technology can become more humanised than ever before – for example, the x.ai CEO Dennis Mortensen demonstrated how their scheduling bot has many layers of empathy, whereby if a meeting is rescheduled time and time again, the bot adopts more apologetic and empathetic language. Compared to the unilateral nature of a telephone tree, this type of humanisation makes technology more enjoyable to use.

The last opportunity is one of powerful computation – AI machine learning algorithms are able to process large volumes of data and information in order to make a new discovery or identify patterns. The ability to analyse and pull insights from huge amounts of information will profoundly augment our human ability of critical thinking and problem solving. For example, being able to set up an AI algorithm to analyse whole of government data in order to identify patterns of child poverty could support policy makers to better solve such a problem.

All in all, Web Summit was an overwhelming feast of technology, pasteis de nata and inspiring speakers. The countdown is already on for Web Summit next year, and in the meantime I’m looking forward to applying what I’ve learned to our work.

If you would like to learn more about AI and the ethics and opportunities around harnessing it, don’t hesitate to get in touch at josephine.young@methods.co.uk 

Center Stage Web Summit
Center Stage at the 2017 Web Summit in Lisbon.