Can education systems anticipate the challenges of AI?
UNESCO EDUCATION SECTOR
IIEP Strategic Debate: Can education
systems anticipate the challenges of AI?
Steve Vosloo, UNESCO, 15 May 2018, IIEP-UNESCO Paris
UNESCO EDUCATION SECTOR 2
A brief history of AI
UNESCO EDUCATION SECTOR 3
AI is already in our lives. It is here, and yet it is only beginning …
UNESCO EDUCATION SECTOR 4
AI is the future, and it’s setting off a panicked race amongst countries …
UNESCO EDUCATION SECTOR 5
… with the emergence of national strategies
UNESCO EDUCATION SECTOR 6
… with the emergence of national strategies
UNESCO EDUCATION SECTOR 7
… and also a race between companies
UNESCO EDUCATION SECTOR 8
Computer vision: Facial recognition systems
UNESCO EDUCATION SECTOR 9
Natural language processing: The next interface
UNESCO EDUCATION SECTOR 10
Data mining and pattern recognition
UNESCO EDUCATION SECTOR 11
What does it mean, for
example, for customer
services, supply chain
UNESCO EDUCATION SECTOR 12
Right now, AI in Ed is mostly promise and potential, but this is changing
UNESCO EDUCATION SECTOR 13
AI can improve the quality of education
• Continuous use and assessment
• Big data for learning analytics
• Personalised, adaptive learning
• Virtual mentors and chatbots
• Real-time data for policy
UNESCO EDUCATION SECTOR 14
Q1. Can AI itself help education systems anticipate the challenges of AI?
How can AI itself shorten the distance between skills supply and demand, e.g.
World Bank and LinkedIn partnership?
“AI will make forecasting more affordable, reliable and widely available” (The
UNESCO EDUCATION SECTOR 15
What AI doesn't do well … or what is the human value add?
UNESCO EDUCATION SECTOR 16
Q2. What do we know about educating for an AI in Ed future today?
In formal, non-formal, informal and
work-based learning environments:
- “Human” skills and competences
include: critical thinking, creativity,
social skills, emotional intelligence,
- Data literacy Carnegie Mellon to
offer the first undergrad AI degree in
the US (with a focus on ethics)
- Focus on lifelong learning
- Retraining on work taken over by AI
UNESCO EDUCATION SECTOR 17
Q4. Is it only IA vs us? The opportunity for intelligence augmentation (IA)
UNESCO EDUCATION SECTOR 18
Q3. What does this all mean for developing countries and equity?
Will AI create growing inequalities between developed and
developing countries, between those who own and/or
create the technology, and those who only use it, as well as
those people whose jobs may be impacted by AI?
Is there a new adaptation divide?
On the other hands, AI can support those with low literacy and digital skills livelihoods.
UNESCO EDUCATION SECTOR 19
Gartner Hype Cycle 2017: AI at peak hype
UNESCO EDUCATION SECTOR 20
Steven Vosloo, Senior Project Officer
In Greek mythology, Talos was a giant automaton made of bronze to protect Europa in Crete from pirates and invaders. He circled the island's shores three times daily. 400BC.
In 1956, officially an academic domain.
Lots of predictions, waves of optimism and AI “winters”. 1970, Marvin Minsky (in Life Magazine): "In from three to eight years we will have a machine with the general intelligence of an average human being."
On 11 May 1997, Deep Blue became the first computer chess-playing system to beat a reigning world chess champion, Garry Kasparov.
The Economist, 2018: “AI will put an end to traditional ways of doing things and start a new era for business and for the world at large. It will be pervasive, devastating and exhilarating all at the same time.”
In smartphones: Predictive text, voice to text, Route suggestions Amazon recommendations Voice assistants, e.g. Alexa and Siri
Questions: What is 1 + 1? Who is the DG of UNESCO? How do we achieve world peace?
GOVERNMENT China betting on it – wants to be global AI leader by 2030 in commerce, government and military The Pentagon is going to Silicon Valley to hire data scientists
COMMERCIAL Amazon calls itself an AI company Race to provide cloud services between Amazon, Google and Microsoft These companies are offering AI services in the cloud, e.g. image tagging and turning speech into text, open-source AI software and, increasingly data sets.
Remember: AI is the umbrella term. Within that is machine learning, which since the advent of big data, has become increasingly powerful.
You need an algorithm to teach, and you need data for it to learn. These are what are being introduced into the cloud.
Image: CC NASA Earth Observatory https://flic.kr/p/jPnc4t
Work 30 percent of “work activities” could be automated by 2030 and up to 375 million workers worldwide could be affected by emerging technologies. James Manyika, Susan Lund, Michael Chui, Macques Bughin, Jonathan Woetzel, Parul Batra, Ryan Ko, and Saurabh Sanghui, “Jobs Lost, Jobs Gained: Workforce Transitions in a Time of Automation,” McKinsey Global Institute, December, 2017.
From Pearson and University College London report on AI in Ed https://www.pearson.com/content/dam/one-dot-com/one-dot-com/global/Files/about-pearson/innovation/Intelligence-Unleashed-Publication.pdf Today: One-on-one digital tutoring to every student Tracking of student performance and providing analysis Voice interfaces and translation
In the future: Lifelong learning companions powered by AI that can accompany and support individual learners throughout their studies – in and beyond school. New forms of assessment that measure learning while it is taking place, shaping the learning experience in real time. AI and IoT for better management of school resources
The dream: quality, personalised learning at scale
AI is good at predictable, repetitive tasks. It is very good at searching, data mining, looking for patterns. But it has no people skills, it can’t do understanding or rationalisation.
Tarek R. Besold, City, University of London, UK https://stevevosloo.com/2017/12/14/online-educa-berlin-2017-rough-notes/ Intelligent tutoring only works well on well-defined, narrow domains for which we have lots of data. Learning analytics is best used to track learner and teacher activities so as to identify individual needs and preferences to inform human intervention.
Humans are good at understanding, empathy and relationships.
In fact, of all the sectors McKinsey & Company examined in a report on where machines could replace humans, the technical feasibility of automation is lowest in education, at least for now. Why? Because the essence of teaching is deep expertise and complex interactions with other people, things that AI are not yet good at. Besold proposed the “human touch” as our value proposition.
One encouraging consensus from the research is that, while there is concern that AI and robots will ultimately take over certain human jobs, teachers are safe. The role relies too much on the skills that AI is not good at, such as creativity and emotional intelligence.
Question: how can AI and humans complement each other? What can we hand over to AI to help us, free our time and “augment” our intelligence?
Image: World Bank CC https://www.flickr.com/photos/worldbank/8785371212/in/album-72157601441433631/
One exmaple of how it can be different is by the Allegheny County Department of Human Services (DHS), Pennsylvania, USA, screen calls about the welfare of local children.
To help, the Allegheny Family Screening Tool was developed. It’s a predictive-risk modeling algorithm built to make better use of data already available in order to help improve decision-making by social workers.
Drawing on a number of different data sources, including databases from local housing authorities, the criminal justice system and local school districts, for each call the tool produces a Family Screening Score. The score is a prediction of the likelihood of future abuse.
The tool is there to help analyse and connect a large number of data points to better inform human decisions. Importantly, the algorithm doesn’t replace clinical judgement by social workers – except when the score is at the highest levels, in which case the call must be investigated.
Given the sensitivity of screening child welfare calls, the system had to be as robust and transparent as possible. Mozilla reports the ways in which the tool was designed, over multiple years, to be like this:
A rigorous public procurement process. A public paper describing all data going into the algorithm. Public meetings to explain the tool, where community members could ask questions, provide input and influence the process. Professor Rhema Vaithianathan is the rock star data storyteller on the project. An independent ethical review of implementing, or failing to implement, a tool such as this. A validation study.
The algorithm is open to scrutiny, owned by the county and constantly being reviewed for improvement.