Log in
Watch Our Video
I was fortunate enough to be invited to be part of the AI Panel at the Civil Society Trustee Exchange Conference in April 2024 at which I met Talk, Listen, Change who were part of the panel. They very kindly agreed to use their own work in developing charity AI to use as a case study to help others. They are one of very few charities currently building AI and what they learned in doing so will help all of us as we all move forward into an AI enabled world.
If you are considering building, buying or funding an AI system you may wish to have a look at the Charity AI System Design and Procurement policy and principles.
TLC: Talk, Listen, Change is dedicated to fostering safe, healthy, and happy relationships. Our commitment to innovation has always focused on meaningful change, whether small or big, that puts the individual at the core. It's not just innovation for innovation's sake; it's about making a real difference in people's lives. In this blog post, we aim to share our approach to AI, our ethical considerations, and the practical applications we've implemented so far.
TLC has always been fortunate to have a multidisciplinary team dedicated to developing and testing experimental tools and ideas that support our staff. Our focus has always been on innovations that make a real difference in the daily lives of our staff, ultimately benefiting our clients. Efficiency improvements help our staff focus more on their work with clients, and AI has been an essential tool in this regard. Language models, in particular, are a new addition to our toolkit.
AI helps us improve our reach and gives a voice to people who struggle to find the words, enhancing the quality of our work. However, we are mindful of the ethical risks that AI poses, both to our mission and to the sector as a whole. There's a danger that AI could trigger a race to the bottom, where people become mere components in an AI-driven process rather than meaningful partners in our mission.
Despite these concerns, we have seen significant benefits from AI. It has helped us overcome barriers by providing the extra expertise needed to push past our limitations. AI acts as a critical friend, highlighting areas we may not have considered. Efficiency improvements have been notable, boosting some teams' productivity by 50%, which allows them to focus on more complex challenges. AI has also been instrumental in thematic analysis and data processing, enabling us to understand our clients' needs more rapidly.
Ethics was the first thing we worked on when considering the use of AI. Language is unique to humans, and now that AI can use it, we are particularly vulnerable to words, especially the spoken word. If a human says "I love you," we assign meaning to it. If we have spoken to them a lot, laughed and smiled, we may believe they meant those words. But what if a machine says those words? With a voice you have chosen, having learned everything about you, it may not mean them, but it would evoke emotion in us—maybe fear, maybe joy. That's why we started with ethics and applied it to all our staff, clients, and their families. Words have meaning; how do we get the benefits of AI without surrendering the joys that make being human so special?
We decided on a set of ethical principles to protect against this. AI should promote critical thinking. People should not become tools in an AI process; that process should positively contribute to their growth and independence in the way they want. We should not use AI in a way that pretends to be a real human. We should provide rights to people's voice and likeness; no one can willingly give those away indefinitely. AI should adopt the ethics and values of the charity, promoting learning and independence.
We run AI exploration in two ways: an experimental arm with strict rules using fake data, focusing on experimenting with every AI technology they can access, and the applied side, which uses all the ethical problems and benefits to apply safe and useful solutions.
For instance, in our experimental side, I cloned my voice using artificial intelligence and created an AI avatar that looked like me. The idea was simple—it is painful making training videos with all the "umms" and "ahhs." AI could save me days or weeks a year in time. Then, when I plugged it into a word generator, it now had my appearance, spoke like me, and used my quirky language. But then the ethical questions hit starkly: Who owns this AI version of me? What might it be used for? Have I robbed myself of agency? An eternal version of me that could outlive me. We resolved this with our ethical principle: humans only have one body and one mind; we cannot swap that out. Identity is something special to us, so no one can give real consent to the perpetual use of their AI selves.
Ensuring our AI tools align with our mission starts with a conversation. We talk about real things that impact real lives while being clear that as a charity, our values apply to everyone. We must respect our shared humanity and the agency it gives. We are not afraid of making mistakes and listen to feedback, ensuring leaders understand the risks and benefits before it is too late to consider the negatives and the positives.
At TLC, we have seen significant benefits from integrating AI into our operations. One of the primary advantages has been overcoming barriers in implementation by bringing in the extra expertise that AI offers. It has been a critical friend, highlighting areas we may not have considered and providing efficiency improvements that boost productivity. For example, some of our teams have experienced a 50% increase in efficiency, allowing them to use the extra time to focus on more complex challenges.
AI has also been instrumental in thematic analysis and data processing. This capability enables us to more rapidly understand our clients' needs, providing insights that were previously difficult to obtain quickly. By leveraging AI for these tasks, our staff can spend more time on direct client interactions and less on administrative duties, ultimately enhancing the quality of our services.
One specific example is the use of AI for voice and language processing. Although this posed ethical challenges, as previously discussed, it also provided practical benefits. For instance, creating an AI avatar for training purposes helped streamline the creation of training videos, saving significant time and effort. This tool allows us to maintain high-quality training materials without the need for constant updates and re-recordings.
Moreover, AI has helped us bridge gaps in knowledge and expertise. By using AI-driven tools, we can access and implement solutions that would otherwise be beyond our current capabilities. This has been particularly useful in areas such as data analysis and thematic research, where AI's ability to process and interpret large datasets has provided valuable insights that inform our strategies and interventions.
Overall, AI has been a valuable addition to our toolkit, enabling us to enhance our operations and better serve our clients. By focusing on practical applications that align with our mission and ethical principles, we ensure that AI contributes positively to our work and the well-being of those we support.
Looking ahead, we are keen on forming an AI ethics group to bring more diverse perspectives into conversations about new innovations and technologies. We are also focused on upskilling our staff, preparing them for a future where AI is as integral to work and life as Google was in the 2000s. As a smaller charity without a large AI budget, we are concentrating on the impact, safety, and ethics of the AI products we use. This will be a constant process of experimentation and application.
Staying ahead of ethical issues as AI technology evolves means embracing the unknown and being comfortable with making mistakes. We need to stay close to the news and promote open discussion about the positive and negative aspects of AI and the anxieties they create. Keeping up to date with new technologies, experimenting and testing them, even those we may not use, will help us understand how less ethical competitors or antagonists, such as scammers, might use them.
One of the developments I am particularly excited about is our internal AI chatbot project. While it will be a long process, the potential of having an AI imbued with all of the charity's knowledge, values, and ethics is thrilling. Capturing a bit of the heart of the people who passionately work to make lives better is incredibly exciting. This project may be the start of something that could outlive even my time at the charity. It's a strange thought that something we create now could be guiding people 50 years from now, preserving our knowledge, experience, and ethics. It could explain why we thought the way we did and make history really valuable, giving hope to those running charities in the future that the people before them also faced and overcame similar struggles.
At TLC: Talk, Listen, Change, our commitment to fostering safe, healthy, and happy relationships drives our approach to AI. By prioritising ethics and focusing on practical applications that benefit our staff and clients, we ensure that AI enhances our operations without compromising our values. Our future plans include forming an AI ethics group, upskilling our staff, and developing innovative tools like our internal AI chatbot. We embrace the unknown and are ready to face the challenges and opportunities that AI presents, always with the goal of making a meaningful difference in people's lives.
We encourage other charities to explore the potential of AI, keeping in mind the importance of ethical considerations and the impact on their mission. Together, we can harness the power of AI to create a better future for our organisations and the communities we serve.
You can find out more about their brilliant work at either of these 2 websites.
The TLC: Talk, Listen, Change - A Relationships Charity (talklistenchange.org.uk).
