Let’s educate our kids for the future of Artificial Intelligence
The first image of Artificial Intelligence for our children of the future could be human-like robots, carrying boxes, watering the plants or moving metal bars. Elon Musk unveiled his prototype "Optimus", at Tesla's AI day last year, showing footage of it doing just that. But we're still a fair way off having one of these in each of our homes, particularly when these robots would need to deal with all manner of unexpected situations (like your dog kicking over its water bowl). What’s plain is that applications of true Artificial Intelligence will be all around us, but that means our workforce of the future needs to be prepared to build these tools, operate alongside them and understand how they work when we engage with them across our civil societies.
That means we're going to need to make sure our education system adapts and includes the right building blocks.
We’re already behind
Looking at other economies, we’re already a bit behind. China is competing with the US in producing the most STEM PhDs in the world (both around 40,000 in 2018) compared to the UK’s producing just over 17,000. It’s expected that China will develop a significant lead over the next five years, although given the freedom of thought allowed out of UK and US institutions, there’s room to ensure an edge with quality while still driving quantity. We need to develop the competitive edge of the education in an increasingly competitive world. So where should we start?
It's both quality and quantity
We've seen the push by our government, companies, research institutions and charities to inspire and drive students into studying STEM (Science, Technology, Engineering & Mathematics) subjects at university to help build the UK's highly skilled workforce of the future. What we don't always hear about is what needs to be done to prepare them across primary and secondary school. This is an extension into broader design and ethical principles, as well as real world applications which tie them all together.
We need to teach the beginnings of ‘human-centred design’, focuses on building AI in a way which delivers value to us all as human beings, read much more broadly than just addressing a problem, but accounts for our wants and needs and the issues we face. Ultimately, this means having empathy for our citizens people and societies.
It doesn’t stop there: students need to begin the broader reading and understanding of ethics, so these can be considered when building, testing and applying AI to different uses.
These design ideas and ethical principles should be rounded out by practically exploring existing issues and concerns within the development of AI, including privacy concerns, while challenging students to think of how they might be mitigated.
This type of learning is not just for STEM students who’ll be building AI, but hugely important for all of our awareness as we interact with AI tools and decisions on a daily basis. It’ll mean we can participate in the debates in our public forums and parliaments on the use of AI too.
Getting resources into the right hands
The challenge is getting the right resources for this education, into the right hands. Accessibility across the schooling system is hugely important, and it’s not just for the selective Public or Grammar Schools who will need to be provided with resources.
We know from experience that without diverse representation in our teams who build AI solutions for real problems, it’s more likely that we could end up building unintended bias into our AI tools or even entrenching existing inequalities. There are plenty of examples where this has already happened, from AI hiring tools which have marginalised women and people of colour, to automated benefits assessment tools which have incorrectly impacted some of the poorest in society.
In the short term, there are teams from charities like the Raspberry Pi Foundation and AI leader's Deepmind leading the charge to ensure our children have access to the right resources for AI and STEM. They’re bringing the latest thinking and resources, including lesson plans and teacher training, and ensuring they're freely accessible to schools across the UK. More widely it’s on Department for Education to ensure this is fed into our national curriculum.
We need our STEM-educated workforce to be as diverse as possible
Don’t forget about equality across our kids
The final piece in the puzzle is ensuring that those engaging in STEM education across our schools are as diverse as possible; in 2017, 27.1% of girls A level entries were for STEM subjects, compared to 45.6% of boys in 2017. At an industry level, the Engineering industry is dominated by White men (93%); and according to the All Party Parliamentary Group (APPG), Black people face underrepresentation in all fields of STEM employment.
This means getting as many people, from different backgrounds, ethnicities and genders involved in STEM education as possible when they’re kids. Recommendations have been made by the APPG to push forward changes which help advance equity in STEM, and there are some great non-profits like STEMettes, the Brilliant Club and the British Science Association who are tirelessly working to help bridge this gap.
It's for the UK, and the wider world
We’ll need AI to help solve our biggest societal problems, including drug discovery, safer and more efficient transport, climate modelling and disaster prediction; the list goes on. The only way we’ll be able to do this effectively, ethically, and with the interests of us and the planet at heart, is by driving forward a rounded education for our children so they can build and apply the tools we need to secure our future.
Originally published in the Evening Standard