I am speaking tomorrow at CogX on the Ethics Stage about The Urgency For More Human Education in an AI Future, Hosted by Big Change.
Come down!
Over the last 12 months, I have been writing a 15,000 words postgraduate thesis on AI, racial bias and gender bias which has played a big role in shaping the research focus at Comuzi.
One of the challenges, we have encountered exploring this area of concern (AI ethics & AI’s impact on society) is the role that the media and movies has played in the mainstream amplification of potential and current problems associated with AI.
This had led to AI being one of the trendiest, least understood and most debated technological breakthroughs, I think in history (shade to Black Mirror).
After seeing the efforts by Deepmind and RSA to facilitate public engagementon the impacts of AI, Akil, Sekyeong and I spent a week to create AI Cheatsheet — a tool for educating the community outside the field of AI to find, understand and share simple definitions about the technology.
We were heavily inspired by Do You Speak Human? by Ikea’s Space10 and Sideways Dictionary by Alphabet’s Jigsaw & Washington Post.
I curated twenty words that are associated with AI, and then I gave Akil the challenging task to develop definitions for each word within a 140 character limit (which he did brilliantly).
We also created a section on the website explaining what can AI do in 2018 as ‘AI’ or sophisticated algorithms have been powering our favourite products and services for ages now (AI is always framed as this faraway future thing).
We also created a survey with questions focused on what impact AI shouldn’t have on society.
Impact
We had a fabulous launch on Product hunt in the first week of May (Much love to Abadesi for hunting).
At the time of writing, AI Cheatsheet has been used approximately by 4,471users globally and collectively the definitions created have received 32,201page views. We got insights from 229 people on our survey (The goal is 1,000).
This is amazing for a project that we created in a couple of days. We plan to promote it more as we would love to get a more broader and inclusive view in regards to race, gender and age of what relationship we want as AI becomes more sophisticated and a staple in everyday life.
I also plan to introduce more features & more ideas as I’ve been sent loads of suggestions but to be honest, I don’t want to over complicate everything.
Reflections
Reflecting on this project and its moderate success, I would like to construct the argument that technology companies can employ the use of educational tools to engage the community outside those who work in the technology field in order to educate and address social and ethical concerns that the general public may have about emerging technologies through human-centred approaches.
Further reflecting, we want to improve the survey questions as for a participant to answer these questions, looking back it could be suggested that we may have assumed that whoever conducted the survey may have a high level knowledge about AI — notwithstanding the definitions provided exactly.
So I would like to better aid survey participants by implementing formats such as scenarios or case studies to help participants to think through their responses.
