Environmental justice aspects of the (un)intended consequences of AI
What is the socio-ecological impact of AI-enabled systems?
–
AI has consequences much more far-reaching and multi-dimensional than many other technologies. How do we even start to think about the impact of AI on the Environment?
Beyond the carbon footprint of training a large machine learning model, it is helpful to consider how a model is utilized as part of a social context. From a systems thinking perspective, it will inevitably impact all other human and non-human actors already participating in that context.
The Environmental Justice movement emerged as a response to the disproportionate exposure to pollutants and environmental hazards borne by marginalized and racialized communities. The US Environmental Protection Agency defines environmental justice as "the fair treatment and meaningful involvement of all people regardless of race, color, national origin, or income, with respect to the development, implementation, and enforcement of environmental laws, regulations, and policies." [1]
Around the world, we've seen the Rights of Nature movement - "the recognition and honoring that Nature has rights. It is the recognition that our ecosystems – including trees, oceans, animals, mountains – have rights just as human beings have rights." [2]
Puzzling through this myself, I think that any investigation of the socio-ecological impacts of AI needs to consider the following:
-
Data ownership and governance frameworks
-
The similarities and differences between data governance and climate governance models
-
Could new business models enable equitable futures for people and the Planet?
What do you think?
References:
[1] https://www.epa.gov/environmentaljustice
[2] https://www.therightsofnature.org/what-is-rights-of-nature/
Sveiki, [bobi.rakova], nuostabu, kad esate čia, ir ačiū, kad pasidalijote klausimu, kvestionuodami PG ir aplinkos sankirtą!
_ // Dirbtinis intelektas turi daug platesnių ir daugialypių padarinių nei daugelis kitų technologijų. Kaip mes net pradedame galvoti apie dirbtinio intelekto poveikį aplinkai? _
Iš tiesų, tiesiog paieškos rezultatų reitingavimas turės nelinijinį poveikį mūsų kasdieniams sprendimams: politikos formuotojams, įmonėms, kurios dėl tų sprendimų priims skirtingus sprendimus. Kokį kelią pasirinko vairavimo krypties navigatorius ar kokia medicininė procedūra buvo grąžinta, rezultatai gali būti labai skirtingi. Idėja: Taigi, galbūt paieškos sistemos turėtų sukurti neprivalomus „saugios paieškos“ rezultatus, kurie būtų traktuojami taip pat atsakingai kaip ir finansinės ar medicininės konsultacijos.
Teisingai paminėjate mokymo modelių anglies dvideginio pėdsaką, tačiau nesu tikras, kiek jis skiriasi nuo, tarkime, kriptovaliutų kasimo anglies pėdsako. Siūlymas: būtų gerai pridėti keletą skaičių teiginiui pagrįsti.
Tačiau ką intelektas turi bendro su judėjimu „Aplinkos teisingumas“ ir „Gamtos teisės“? Kaip jie susiję su trimis jūsų paminėtais klausimais? Klausimas: Ar galėtumėte paaiškinti, kaip, tarkime, duomenų nuosavybė ir duomenų valdymas veikia klimato valdymą? Koks status quo ir kaip turėtų būti?
P.S. Kadangi išsaugojote temą prieš pasirinkdami „tyliai“ arba „juodraštį“, todėl gavome pranešimą apie tai, nors šiuo metu jos negalima rasti rodyklėje, ji vis tiek pasiekiama tiesioginiu URL. Taigi, tikiuosi, kad mūsų komentarai bus naudingi norint pagerinti jūsų rašymą.
Welcome, [bobi.rakova], amazing to have you here, and thank you for sharing the issue questioning the intersection of AI and environment!
// AI has consequences much more far-reaching and multi-dimensional than many other technologies. How do we even start to think about the impact of AI on the Environment?
Indeed, simply the ranking of search results will have non-linear effects to our daily decisions: of policy makers, of businesses, that, because of those decisions shall make different choices. What path driving directions navigator chooses, or what medical procedure search engine returns may entail very different outcomes. Idea: So, perhaps search engines should develop an optional "safe search" feature to display only results that are treated with the same responsibility as financial or medical advice.
You are correctly mentioning carbon footprint of training models, though, I'm not sure how much it is different from carbon footprint of say, mining crypto-currencies. Suggestion: it would be good to add some numbers to substantiate the claim.
However, what does AI have to do with Environmental Justice movement, and Rights of Nature movement? How are they connected with the three issues you mentioned? Question: Could you elaborate on how, say data ownership and data governance affects climate governance? What's the status quo, and how it should be?
P.S. As you've saved the topic before choosing "silently" or "draft", therefore, we got the notification about it, while it is currently not discover-able in the index, it is still accessible via direct URL. So, hope our comments are useful in improving your write up.
■ ■ [+]
Dėkojame už apgalvotą atsakymą [Mindey]!
Geriausia nuoroda, susijusi su AI anglies pėdsaku, yra Strubell, E., Ganesh, A., & McCallum, A. (2019). Energetikos ir politikos aspektai giliam mokymuisi NLP. https://arxiv.org/abs/1906.02243
Pavyzdžiui, mokant „Google“ kalbos modelio (BERT) versiją, kuri yra bendrovės paieškos variklio pagrindas, Strubello vertinimu, 1438 svarai CO2 ekvivalento buvo beveik tokie patys, kaip skrydis pirmyn ir atgal tarp Niujorko ir San Francisko. Praktiškai vienas modelis yra mokomas daug kartų, kol jis pradedamas gaminti.
Turėčiau patikslinti, kad turiu omenyje ne dirbtinį intelektą, o mašininio mokymosi naudojimą kuriant aplinkosaugos sprendimus. Puikus su tuo susijęs apžvalginis straipsnis yra: Rolnickas, D. ir kt. (2019). Spręsti klimato kaitą mokantis mašinų. https://arxiv.org/pdf/1906.05433.pdf Interaktyvią jo versiją žiūrėkite čia: https://www.climatechange.ai/summaries
Pavyzdžiui, dirbtinio intelekto sistemos buvo naudojamos miško anglies sekvestracijos potencialui prognozuoti naudojant palydovų ir dronų vaizdų duomenis; kompiuterinės regos algoritmai naudojami nustatant tinkamas sodinimo vietas, stebint augalų sveikatą ir analizuojant tendencijas; Dirbtinis intelektas taip pat naudojamas siekiant nustatyti, kur miškų kirtimas galėjo būti vykdomas neteisėtai, taip pat įvertinti riziką dėl gaisro, ligų, vabzdžių ar kitų priežasčių. Visų šių naudojimo būdų nuorodas galima rasti Rolnick ir kt.
Mane domina duomenų valdymo modeliai aplinkos AI modelių atveju. Pavyzdžiui, dirbtinio intelekto atveju, naudojamo miško atkūrimo projektams:
kokios duomenų nuosavybės ir duomenų valdymo sistemos galėtų suteikti teisingumo ir įtraukti vietinių pasaulėžiūrų į technologijos naudojimo būdą?
koks gali būti tokių dirbtinio intelekto modelių poveikis ekosistemų paslaugų ir paslaugų ekosistemoms koncepcijoms? Comberti ir kt. (2015). Ekosistemos paslaugos ar paslaugos ekosistemoms? Vertinti auginimą ir abipusius žmonių ir ekosistemų santykius. Visuotiniai aplinkos pokyčiai, 34, 247-262. https://www.sciencedirect.com/science/article/pii/S0959378015300145
Redagavau įrašą pažymėdamas parinktį „paskelbti“ ir tikiuosi, kad padariau jį dabar atrandamą, tačiau vis dar nematau pagrindiniame „Puzzles“ puslapyje.
Thank you for your thoughtful reply [Mindey]!
Related to AI's carbon footprint, the best reference I know about is: Strubell, E., Ganesh, A., & McCallum, A. (2019). Energy and policy considerations for deep learning in NLP. https://arxiv.org/abs/1906.02243
For example, training a version of Google’s language model (BERT) which underpins the company’s search engine, produced 1,438 pounds of CO2 equivalent in Strubell’s estimate to nearly the same as a round-trip flight between New York City and San Francisco. In practice, a single model is trained a large number of times before it's put in production.
I should have clariefied that I don't mean AI in general but rather the use of Machine Learning in the development of environmental solutions. A great overview paper related to that is: Rolnick, D., et al. (2019). Tackling climate change with machine learning. https://arxiv.org/pdf/1906.05433.pdf See an interactive version of it here: https://www.climatechange.ai/summaries
For example, AI systems have been used to predict forest carbon sequestration potential through the use of satellite and drone image data; computer vision algorithms are used in identifying appropriate planting sites, monitoring plant health, and analyzing trends; AI is also used to identify where deforestation may have been conducted illegally as well as assess risks due to fire, disease, insects, or other causes. References on all of these uses are available with the paper by Rolnick et al.
I'm interested in data governance models in the case of environmental AI models. For example, in the case of AI used for forest restoration projects:
what data ownership and data governance frameworks could empower equity and inclusion of indigenous worldviews in the way the technology is used?
what might be the potential impact of such AI models on the concepts of Ecosystem Services and Services To Ecosystems? Comberti et al. (2015). Ecosystem services or services to ecosystems? Valuing cultivation and reciprocal relationships between humans and ecosystems. Global Environmental Change, 34, 247-262. https://www.sciencedirect.com/science/article/pii/S0959378015300145
I edited the post by checking the 'publish' option and I hope I've made it discoverable now, however, I still don't see it on the main Puzzles page.
// AI has consequences much more far-reaching and multi-dimensional than many other technologies. How do we even start to think about the impact of AI on the Environment?
-Hmm, to me it seems we're talking about the same Human impact on the Environment (because Humans create AI to improve lives for themselves). So from here, I invite to zoom out, and add a puzzle on top of this about questioning Human perspective on life in general. When humans see themselves as the centre of the Earth (and probably, the Universe), we create technologies to improve life for ourselves only, e.g. we design say AI to have all knowledge at the click of a button without realising that training AI creates pollution. And then in the design process we further zoom in and divide humans into different stakeholders (like, users, funders, doers, etc) instead of zooming out and having the planet Earth as a stakeholder for example.
So, zooming out here as a thought exercise:
How would the Internet (and any web tech) be designed if we see humans only a specie as part of the living network on Earth and in the Universe? Could we imagine that the planet Earth itself has needs and not only humans do? Could we map out species from the environment as stakeholders in the design process of technologies?
What we could do even here on o2oo: when a new idea is described (before it goes into production), it could share an estimate of the pollution it will create and estimate the impact/value it will produce.
Let's take o2oo system as an example. So by running our website we probably pollute too ([Mindey], maybe you could help me think through this? :), and then we promise to get the humanity connected, so that all new ideas are discussed before they go into production. A promised value (with all stakeholders in mind) would be - that if an idea/project is discussed early, then it can be adjusted, while existing industries are so complex, and are not flexible anymore..
So I wonder how about estimating waste and value/impact for ideas and projects?
Another question is about ethics: Some ideas will promise to create a greater knowledge, even with an expense of creating pollution. What then? How could pollution transform into no-waste (and have a circular model of zero waste)? Say, perhaps when in the ideation process a CO2 emission is projected, then a part of that idea should involve solving that problem in general.
■ [+]
Thanks for your response [Ruta]! I agree, I think this idea of enironmental justice pushes us to imagine a future where Nature is a stakeholder.
I think it is important to differentiate between AI and technology in general. In my experience, often times, the engineering of an AI model start from the idea of full automation because an AI system can do what people do - faster, cheaper, and better than humans. I think this is why it becomes important to consider the cnsequences as automating away the human introduces new kinds of challenges.
I think it'd be easier to start with the impact assessment of AI models which directly interact with environmental systems as then we could more directly measure the potential impact. Does that make sense?
// For example, training a version of Google’s language model (BERT) which underpins the company’s search engine, produced 1,438 pounds of CO2 equivalent in Strubell’s estimate to nearly the same as a round-trip flight between New York City and San Francisco. In practice, a single model is trained a large number of times before it's put in production.
[bobi.rakova], a few more thoughts:
1,438 pounds (=652 kg) of CO2 for Google language model? Training AI models from big data is expensive, but trained once, a model can be reused very cheaply large number of times. Google may re-train the model frequently, so it depends on the frequency, which your comment doesn't mention. Also, it doesn't mention how they are training this model -- if they have something pre-trained already, and training only the top layer, the statistic of expensive entire training would misrepresent the reality of relatively cheap top layer training (transfer learning) on top of the base model. Suggestion: It would be good to mention the total amounts.
In general, because it's expensive to train, but cheap to use, it would be great to have large public model zoo, or model market, so as to prevent people re-training expensive models. I wonder, what's the largest public model zoo that humanity has. Having something like "GitHub" for AI models would also be very liberating to people who have no hardware on which to train large models, but still want to benefit from AI models. I think I had this as a separate idea somewhere.