Robotic Alms: AI and the future of charitable giving
- 1. 1 Robotic Alms: AI and the future of charitable giving Rhodri Davies Programme Director, Giving Thought
- 2. 4 Ways AI could transform charity 1) Creating new problems for charities to address 2) Developing new ways of addressing existing problems 3) Offering new ways of working for existing charitable organisations 4) Creating new governance structures and operating models for achieving social good 2
- 3. AI & Grantmaking 3 Some have been thinking about how AI could improve grantmaking process by automating ID/selection of grantees, e.g. Nesta:
- 4. Philanthropy advice 4 From this… …To this
- 5. AI & Philanthropy advice 5 AI could transform philanthropy advice in a number of ways: 1) Mass marketisation: Automation through chatbots etc. could reduce costs & make mass- market service, whilst also driving up quality 2) Identifying need: AI analysis of data could identify most pressing areas of social or environmental need at given time 3) Picking interventions: AI analysis of social impact data could identify most effective interventions and organisations 5) Understanding donor aims: AI analysis of donor behaviour, peer group etc. could help to identify donor goals more effectively 4) Automated giving: AI could be used to automate matching of needs to interventions, and dispersal of money; thus creating “AI philanthropy”
- 6. Making Philanthropy Advice Mass- Market 6 “AI has the potential to become a great equalizer. Access to services that were traditionally reserved for a privileged few can be extended to the masses.” PWC (2017) Bot.Me: A revolutionary partnership: How AI is pushing man and machine closer together
- 7. A bit of context 7
- 8. How likely is this to take off? 8
- 9. No really, how likely? 9
- 10. Advice on Philanthropy Methods 10 • Use AI-powered chatbots to bring together existing advice in personalised way • Might include: • Vehicles (foundation, DAF, charity account etc.) Social purpose org structures (charity, CIC, socent, B Corp etc.) Tax incentives (Gift Aid, Payroll Giving, Share Gifts, SEIS etc.) • Strategies (place-based, social investment, venture philanthropy etc.).
- 11. Approaches to philanthropy advice 11 • Choice over whether to prioritise meeting donor aims or maximising social outcomes • May be aligned, but often there is tension between the two
- 12. Focus on outcomes 12 • AI could analyse data to identify most acute pressure points in terms of social or environmental needs at any given time. Identifying causes Identifying Interventions • AI could analyse data on social impact to identify most effective interventions/orgs Onus on data: Scale Openness Consistency Social Impact data
- 13. How far will this go? 13 Hyper-rational “AI Philanthropy: AI used to ID most pressing needs at given moment and match with most effective interventions. • So no need for human involvement • Sceptical p.o.v: “can never remove element of heart from philanthropy” BUT: could this change in the future…?
- 14. Why should we care? 14 The IoT market is going to be huge, with vast numbers of M2M transactions Q: Could we harness some of this for charity?
- 17. Identifying Donor Aims 17 • Subjective • Objective • Self-reported info on what donor wants to achieve • BUT: We are notoriously bad at self-analysis… • Objective Social: Facebook algorithm approach i.e. advice/suggestions based on past behaviour/peer group • Danger of entrenched bias • Filter bubbles • Reinforcing popular causes and orgs. • Objective Personal: Future in which AI can know us better than ourselves • Advice based on what objectively maximises donor’s satisfaction.
- 18. Challenges 18 • Lack of tech skills in sector: how do we built partnerships, get govt support etc.? • Developing social impact data: who pays for it? • Facebook-style algorithms: will they entrench social bias and create filter bubbles? • If algorithms are based on past behaviour, will popular causes/orgs win out at expense of less popular ones? • Black-box algorithms: how do you optimise if you don’t understand criteria?
- 19. What should charities do? 19 • Cross sector tech/AI forums (private sector, vol sec, govt) • Join debate over AI and put charity perspective (e.g. CAF Lords AI committee response.) • Adopt Open Data approach (e.g. 360 Giving) • Develop social impact data • Work with IoT industry: both to harness data, and to develop M2M giving
- 20. Where to find us 20 • CAF Giving Thought page: https://www.cafonline.org/about- us/blog-home/giving-thought • Archive Giving Thought blog: www.givingthought.org • Giving Thought podcast: http://givingthought.libsyn.com • Twitter @Rhodri_H_Davies
- 21. Rhodri Davies Programme Director, Giving Thought Charities Aid Foundation rdavies@cafonline.org