From Risk to Regeneration: Prioritizing People and Planet as the Metrics for Responsible AI
John C. Havens
Author, Heartificial Intelligence and Hacking Happiness. Director, IEEE Planet Positive 2030. Founding E.D. of IEEE AI Ethics program and IEEE 7000 Standards Series.
The following is text from my remarks at the end of the first day of The Athens Roundtable Event, Artificial Intelligence and the Rule of Law 2023 Edition . Thank you to The Future Society who organized the event for the opportunity.
To get an immediate sense of how to implement Wellbeing metrics (measures that include accountable ways to prioritize people and planet for GenAI and any other AI systems), please check out IEEE's recent publication, Prioritizing People and Planet as the Metrics for Responsible AI written by over a dozen experts in areas of business, policy and academia.
**********
What gives you a sense of worth that connects you to yourself, other people and the planet??
My Dad got his sense of worth from being a psychiatrist.? His life’s purpose was based on helping people remove negative pain from their lives.?
When he died in 2011 I merged my passion for technology with the study of wellbeing that led me to write my 2014 book, Hacking Happiness: Why Your Personal Data Counts and How Tracking it Can Change the World .? It was in researching this book I discovered how thoroughly our data represents our identity in the world of technology and AI.?
It’s also where I discovered the science of positive psychology.
Whereas traditional psychoanalysis asks, “How can we remove someone’s pain,” positive psychology asks, “While we remove the pain, how can we also increase wellbeing?”?
Or in other words, how can we move beyond simply addressing harms to help someone flourish??
The science says this can happen for a person when they practice gratitude for themselves or engage in acts of altruism for others.
Living in this way brings connection to self and to others.? In joy.?
This is true for all humans, anywhere in the world.?
In my last book, Heartificial Intelligence: Embracing Our Humanity to Maximize Machines from 2016 I did a lot of research on the metrics of wellbeing.? These metrics provide quantifiable and qualitative instruments to account for and improve long-term human flourishing while also honoring nature.? The UN Sustainable Development Goals fall into this category as does the excellent work of The OECD in their seminal Better Life Index and new Centre on Well-being, Inclusion, Sustainability and Equal Opportunity .
These types of metrics account for things not measured by economic instruments of growth in isolation.?
But what you measure matters and who and what you don’t measure are not accounted for.? Literally.
And typically by not being measured they’re harmed.??
Using these wellbeing metrics, however, is a way to connect all people and the planet in complement to existing measurements of GDP.? Nobel laureate in economics Joseph Stiglitz noted in his 2009 report that, The time is ripe for our measurement system to shift emphasis from measuring economic production to measuring people’s well-being .
2009.
It is high time we adopt this thinking regarding the design, creation and regulation of Artificial Intelligence Systems.? Frontier, dual-use, or any type of AI.?
And we need to use the term, “AI Systems” versus just “AI.”? Because that reminds everyone these amazing technologies are powered by human and planetary data we must always account for.
While we have to understand the risks that can happen with these systems, we must also recognize the overarching harms that currently exist from not prioritizing ecological flourishing and human wellbeing as the societal metrics of success for these tools.?
I work for The IEEE Standards Association , and in terms of technical connection, we are all connected to wi-fi because of an IEEE Standard called 802.11.? Creating standards is hard work.? It involves lots of intense discussions over years where people must ensure they’re best communicating how certain technologies can be interoperable anywhere in the world.?
In 2016, IEEE brought together what would become over seven hundred global experts in a similar process of interdisciplinary consensus building to create a compendium document called, Ethically Aligned Design: A Vision for Prioritizing Human Wellbeing for Autonomous and Intelligent Systems .?
The 2017 version of the document inspired organizations like the OECD, IBM and dozens of others to help create their AI Principles and by the release of the 2019 version of the paper The IEEE 7000 series of standards focusing on the intersection of technological interoperability and applied ethics had grown to encompass over a dozen working groups and standards.?
IEEE is also proud to have co-founded The Athens Roundtable with The Future Society for its first iteration, inspired by the Law Chapter of Ethically Aligned Design .?
In terms of how standards can help govern AI systems with a people and planet perspective, I’ll give a few quick examples.?
The IEEE 7010 Recommended Practice for Assessing the Impact of Autonomous and Intelligent Systems on Human Well-Being launched in 2020 provides a wellbeing impact assessment that Frontier and all AI systems designers and users can implement to discover the innovation and value that comes from scenario planning utilizing wellbeing metrics.?
For instance, it’s well known that LLMs use a great deal of water to cool servers in data centers.? But that water won’t be available for more than a few years if water scarcity issues and aquifer depletion aren’t measured at the outset of design before funding is even allocated for a specific data center.? Working with local communities, farmers and indigenous populations using wellbeing metrics means creating the accountable basis for long-term business and AI development.?
Below is an image of the Wellbeing Impact Assessment flow chart from IEEE 7010-2020 which provides a visual complement to how this example of water could be analyzed using the standard:
In terms of using metrics to embrace people who are often not accounted for, IEEE has a Standard Working Group creating a Recommended Practice for Provenance of Indigenous People’s Data , that we believe may be the first international indigenous data standard.? We must avoid the erasure and colonization of people’s language and culture with these AI systems.? Free, Prior and Informed Consent and indigenous sovereignty must be honored for a flourishing future.
And finally in terms of the people we universally must include and prioritize in all our AI Systems governance, The IEEE 2089 Standard for Age Appropriate Design released in 2021 provides organizations a framework to practically orient design processes for age appropriate digital services toward responsible technological innovation inclusive of children.? It’s based on regulation launched in the United Kingdom that was recently approved as regulation by Governor Newsom in California .?
Because human children exist now.? Their rights, data, and identity need to be protected whether or not AGI ever comes into being.?
Loneliness and isolation have been declared an epidemic in the U.S. by our Surgeon General and globally by the World Health Organization .?
Kids need to be told by this community that they have worth simply for being born.?
And they need to know our generation is prioritizing the planet over profit.?
Finally, IEEE has a new program called Planet Positive 2030 and a new series of 7800 standards focused on the concept of “regenerative” sustainability and design.? The idea of regeneration involves a transformation of the heart and mind to understand that “sustainability” doesn’t just mean minimizing emissions to avoid regulation.?
It means giving back more to the planet with everything we do.? With all the technology we create.
Our kids and the planet.? The living land and our ancestors.?
These are the stakeholders that matter most. We must see them.? And account for them.?
I want to thank The Future Society for this chance to talk and for all our speakers and panelists.? I especially appreciated hearing from the Global South from a region so critically important to listen to in terms of planetary safety to help gain perspectives beyond Global North or Western ethical traditions that tend to prioritize rationality as the cornerstone of intelligence.?
领英推荐
But it’s also the relationality and connection of community and caregiving that will help us ensure the value of these AI systems does not come primarily with metrics of growth or speed that risk devaluing human and planetary worth.
Rather, we must prioritize people and the planet as the metrics for responsible AI.
Together.
**********
The following charts are from Prioritizing People and Planet as the Metrics for Responsible AI and the first publication in the same series, A Call to Action for Businesses Using AI .
The logic from the Committee is you can use the first paper and AI Ethics Readiness Chart to help implement Responsible AI throughout your organization, then utilize Wellbeing Metrics to drive impact prioritizing people and planet for AI Systems, tech and policy overall.
**********
IEEE's Wellbeing and AI Systems work (2016-2023)
Please note I have used people's titles as I found them on LinkedIn / other sources today (Dec 3, 2023) and members have worked at various jobs during the time these documents below were written (from 2015 until 2023). Where I have missed any work at IEEE focused on AI Systems and Wellbeing, please let me know and I can add them to this list. I am included work I was directly involved in creating, shepherding, or supporting.
Thank you in advance to everyone who I've included below. I should note that my comments in my speech for The Future Society do not necessarily reflect those of my employer (as IEEE is a large organization) or the experts listed below.
If I've forgotten anyone - my deep regrets and please email or contact me right away so I can add your name / info below. Where I couldn't find prominent bios on LinkedIn or via longish searches, I've left a few names / bios blank.
In 2019 The Board of Directors of IEEE did approve a Position Statement however, that, "IEEE endorses the principle that the design, development and implementation of autonomous and intelligent systems (A/IS) should be undertaken with consideration for the societal consequences and safe operation of systems with respect to: A/IS developers should consider impact on individual and societal well-being as the central criterion in development." Noting that Board positions can change, my point in listing this fact is that the nature of wellbeing in relation to AI Systems was elevated to the highest levels of IEEE in 2019.
I can't speak for all contributors below in terms of their current thinking, re: the nature of wellbeing metrics and AI Systems, but do want to deeply thank them for their participation in this work that has helped me and many others fundamentally consider why ecological flourishing and human wellbeing are so critical to prioritize for AI systems and all technology development.
**********
Prioritizing Human Well-being in the Age of Artificial Intelligence
On 11 April 2017, IEEE hosted a dinner debate at the European Parliament in Brussels called, Civil Law Rules on Robotics: Prioritizing Human Well-being in the Age of Artificial Intelligence.
A primary subject of debate for IEEE’s event at the Parliament was to ask how existing Beyond GDP metrics that do measure emotion or other AI-relevant factors could increase the efficacy of these technologies while inspiring ethically driven innovation overall. In the same way the notion of liability must evolve to best define accountability for a robot and its operator,
...the hypothesis for the evening was that the world’s top metric of value must move Beyond GDP to holistically measure how intelligent and autonomous systems can hinder or improve human well-being.
Watch the video here , (seven minutes) featuring the following people:
**********
Thank you to the Co-Chairs of the committee that drove the work of the two Ethically Aligned Design for Business Papers. The first paper in the series is, A Call to Action for Businesses Using AI:
And to the Committee members that drove this work:
**********
Thank you to the Co-Chairs of the committee that drove this work. The second paper in the Ethically Aligned Design for Business series: Prioritizing People and Planet as the Metrics for Responsible AI:
And to the Committee members that drove this work:
**********
Thank you to the members of IEEE 7010-2020 who helped inspire all this work:
**********
This work got its inspiration from the Wellbeing Chapter of Ethically Aligned Design, a committee that released two reports in 2017 and 2019 . Below is the list of members from the Committee :
?
?
AI Ethicist | Author of Responsible AI | Keynote Speaker | Founder
11 个月I love the wellbeing and flourishing focus John, and have loved getting to work with you on it!
REGENERATIVE PUNK // ??keynote speaker, advisor, author, entrepreneur, board member, professor // #AI #data #leadership #(bio)diversity #health #food #wearabletech #climate // Lisbon + Vienna (formerly New York)
11 个月John C. Havens #AI and #regeneration or at least #sustainable computation. Let’s explore impact of environmental changes caused by and impacting digital technologies. Wonderful to read your summary of thoughts.
THIS RANDOM SENSE OF WONDER Writing | Research | Science & Technology | Future | Business Culture
11 个月Excellent! Thank you, John, for championing this message from your unique blend of position, gift, and passion! So gratifying to see...
Climatetech Entrepreneur, Investor, Board Member, Schwarzman Scholar, Forbes30U30
11 个月Incredible sharing and picture
Co-director FARI - AI Institute for the Common Good (ULB-VUB) in Brussels, Adjunct Professor - Université du Québec à Montréal
11 个月Bravo John C. Havens