For our latest expert interview on our blog, we’ve welcomed Alice Thwaite to share her thoughts on the topic of artificial intelligence (AI) as well as the reasons behind founding technology ethics company Hattusia.
Alice Thwaite is an AI ethics philosopher and ethicist who specialises in creating democratic information environments. As well as being the founder of Hattusia she is also the founder of the Echo Chamber Club, a philosophical institute dedicated to understanding what makes information environments democratic.
Subsequently, she has recently spoken on the BBC on the topic of ethics and their place in AI.
How did you begin your journey in becoming an AI ethicist, what piqued your interest in this field?
I came into AI ethicism through philosophy. I graduated with an MSc from the Oxford Internet Institute and an MA Cantab in History and Philosophy of Science from Cambridge University.
After university, I founded the Echo Chamber Club, which at the time, existed as a newsletter to help those who identified as metropolitan liberal/progressives access and understand different points of view.
This project ended in January 2018 after I realised that the theory underpinning echo chambers was misguided, as were the theories underpinning a huge amount of the technology we create.
Why is ethics necessary for AI?
A single technology has the ability to have a profound effect on thousands (if not millions) of people. For example, look at the decision-making algorithm that the Post Office used to incorrectly judge whether postmasters had stolen from the company, or the police surveillance technologies that have been shown to amplify bias and police overreach.
Anyone building AI should be making conscious, deliberate decisions about how their technology will interact with the world.
Ethics helps us to answer questions about the kind of world we want to live in. Do we want a world where the environment is sustainable for human and animal life? Where human rights are upheld and protected? Where organisations are designing with future wellbeing in mind?
Or do you want to live in a world where work is only meaningful if it makes a profit? Ethics helps us understand that growth is not everything. If a business grows in a way that exploits human and global resources, without giving anything back, eventually there will be nothing left.
What are some of the biggest ethical AI challenges we are currently facing?
I believe there are individual issues like bias and discrimination, or surveillance and our privacy, the automation of workforces, but I believe all of these individual issues would benefit from better awareness. Better awareness of what tech ethics/ai ethics is, how it affects peoples lives, and what organisations can do to mitigate the potential harms of these technologies.
How can we avoid bias arising in our AI-led projects?
We’ll never be able to eradicate bias from AI. We need to accept that technology is never ‘neutral’, and always makes a statement about what kind of world we want to live in.
There are a number of techniques for understanding the power structures and relations that technology produces, as well as methods in diversifying data sets and so on. We need to create measurement infrastructures to be able to tackle the problem’s intolerable bias.
How do we keep AI under control when it is continually evolving in ways that are unexpected?
We need to move away from the idea of ‘move fast and break things’. If we are building technology that is evolving too fast for us to control then that technology is not ready to be tested on the general public.
There is no harm in taking a slower, more considerate approach to launching technology that has the ability to impact millions.
What did your journey look like when founding Hattusia?
Hattusia started with an activist project called the Echo Chamber Club. I was working to understand digitally induced polarisation and echo chambers and through this process realised that the philosophical underpinnings of echo chamber theory was misguided.
This experience made me realise that the philosophy behind a lot of technology is very poorly constructed. By using the philosophical method, we could help technologies create worlds that we actually wanted to live in. And that is what Hattusia specialises in.
What are the typical clients of Hattusia, what is your typical workflow like?
We’ve worked with clients like Just Eat Takeaway.com, NHSX, Sky, General Assembly, the European Partnership for Democracy, the Centre for Data Ethics and Innovation, and many more.
Typically clients first engage with us through what we term our ‘Campus’. This is a series of workshops that give your workforce a grounding in the key ideas of tech ethics and how you can start thinking about embedding ethical methodologies into your products.
After the workshops, we start to look at consultancy work which can come in the shape of ethical audits, conceptual analysis & design, literature reviews, education programmes, and operationalising ethics.
How do you handle marketing activities in an ethical way whilst still running profitable campaigns?
We use privacy-focused analytics software on our website, and we do not use Facebook and Google’s tracking pixels and codes.
This can make things trickier, especially for a small business, but it also makes us more creative in our approaches to contacting potential customers.
We are also aware that as a consultancy we can rely on more traditional sales methods, which tend to be less reliant on data-centric advertising practices.
Do you have any other additional news or upcoming projects you would like to share?
I will be MCing the Ethics Stage for CogX in June, if you’re going please do stop by and say hello!
If you enjoyed this article then why not check out our previous blog post on Microservices vs APIs?