KoolKill

Ethics in AI

Ethics in AI

DeepMind researchers propose rebuilding the AI industry on a base of anticolonialism – “The researchers detailed how to build AI systems while critically examining colonialism and colonial forms of AI already in use in a preprint paper released Thursday. The paper was coauthored by DeepMind research scientists William Isaac and Shakir Mohammed and Marie-Therese Png, an Oxford doctoral student and DeepMind Ethics and Society intern who previously provided tech advice to the United Nations Secretary General’s High-level Panel on Digital Cooperation.”

The researchers posit that power is at the heart of ethics debates and that conversations about power are incomplete if they do not include historical context and recognize the structural legacy of colonialism that continues to inform power dynamics today. They further argue that inequities like racial capitalism, class inequality, and heteronormative patriarchy have roots in colonialism and that we need to recognize these power dynamics when designing AI systems to avoid perpetuating such harms.

“Any commitment to building the responsible and beneficial AI of the future ties us to the hierarchies, philosophy, and technology inherited from the past, and a renewed responsibility to the technology of the present,” the paper reads. “This is needed in order to better align our research and technology development with established and emerging ethical principles and regulation, and to empower vulnerable peoples who, so often, bear the brunt of negative impacts of innovation and scientific progress.”

The paper incorporates a range of suggestions, such as analyzing data colonialism and decolonization of data relationships and employing the critical technical approach to AI development Philip Agre proposed in 1997.

The notion of anticolonial AI builds on a growing body of AI research that stresses the importance of including feedback from people most impacted by AI systems. An article released in Nature earlier this week argues that the AI community must ask how systems shift power and asserts that “an indifferent field serves the powerful.” VentureBeat explored how power shapes AI ethics in a special issue last fall. Power dynamics were also a main topic of discussion at the ACM FAccT conference held in early 2020 as more businesses and national governments consider how to put AI ethics principles into practice.

some of DeepMind’s machine learning fairness research…

also btw…

Softlaw: “law that is software coded before it is passed.” (A very direct and literal take on @lessig’s “code is law”)[1,2]

posted by kliuless (38 comments total)

45 users marked this as a favorite

Open Democracy: Reinventing Popular Rule for the 21st Century

Open Democracy: Reinventing Popular Rule for the 21st Century

Politics Without Politicians – “The political scientist Hélène Landemore asks, If government is for the people, why can’t the people do the governing?” (via)

Her forthcoming book, due out next year and currently titled “Open Democracy: Reinventing Popular Rule for the 21st Century,” envisions what true government by mass leadership could look like. Her model is based on the simple idea that, if government by the people is a goal, the people ought to do the governing.

“Open democracy,” Landemore’s coinage, does not center on elections of professional politicians into representative roles. Leadership is instead determined by a method roughly akin to jury duty (not jury selection): every now and then, your number comes up, and you’re obliged to do your civic duty—in this case, to take a seat on a legislative body. For a fixed period, it is your job to work with the other people in the unit to solve problems and direct the nation. When your term is up, you leave office and go back to your normal life and work. “It’s the idea of putting randomly selected citizens into political power, or giving them some sort of political role on a consultative body or a citizens’ assembly,” said Alexander Guerrero, a professor of philosophy at Rutgers who, in 2014, published an influential paper arguing for random selection in place of elections—a system with some precedents in ancient Athens and Renaissance Italy which he dubbed “lottocracy.” (It’s the basis for his own forthcoming book.) In open democracy, Landemore imagines lottocratic rule combined with crowdsourced feedback channels and other measures; the goal is to shift power from the few back to the many.

To many Americans, such a system will seem viscerally alarming—the political equivalent of lending your fragile vintage convertible to the red-eyed, rager-throwing seventeen-year-old down the block. Yet many immediate objections fall away on reflection. Training and qualification: Well, what about them? Backgrounds among American legislators are varied, and members seem to learn well enough on the job. The belief that elections are a skills-proving format? This, too, cancels out, since none of the skills tested in campaigning (fund-raising, glad-handing, ground-gaming, speechmaking) are necessary in a government that fills its ranks by lottery.

Landemore was taken with the unorthodox idea that normal people, in a group, could be trusted with big, scary decisions.

Open Democracy: Reinventing Popular Rule for the 21st Century with Hélène Landemore – “Hélène’s third book–Open Democracy: Reinventing Popular Rule for the 21st Century (under contract with Princeton University Press)–develops a new paradigm of democracy in which the exercise of power is as little gated as possible, even as it depends on representative structures to make it possible. In this version of popular rule, power is equally open to all, as opposed to just those who happen to stand out in the eyes of others (as in electoral democracies). The book centrally defends the use of non-electoral yet democratic forms of representation, including ‘lottocratic’, ‘self-selected’, and ‘liquid’ representation.”

also btw…

Continuous elections – “For a variety of reasons, I think it a good idea that we introduce into our voting system a greater element of stochasticism, of structured, intentional randomness.”[1]

We select juries largely by lottery, on the theory that this is a good way to get a representative sample of ones “peers”. There have been a variety of democratic experiments with sortition, simply choosing by lot, picking random names from the phone book. Perhaps overcynically, many of us might consider that an improvement over our present, professional political representation.

I do not favor sortition for the constitution of our legislatures. There is a lot to be said for choosing among representatives who express an interest in and commit to doing the work, and to some kind of voting process that ideally filters for quality. What I do favor is an idea called “lottery voting” or “random ballot”. I really encourage you to read the first link, a very readable academic note by Akhil Reed Amar which introduced the idea. You should also read this essay by David MacIver (ht Bill Mill). In a nutshell, everybody votes in the way they currently do for their preferred candidate. Then we throw the ballots in a big hat, and draw the winner like a bingo hall door prize.[2] You’d never want to use lottery voting to elect a President. Who knows who you might pick? It’d be totally random. But for a large legislature, lottery voting will predictably yield proportional representation along whatever axes or characteristics are salient to voters, not just formal political parties. Further, lottery voting is immune to gerrymandering, and every vote always has equal influence. (This is decidedly untrue of conventional “first-past-the-vote” voting, where the statistical effect of a vote — the difference in the probability of a candidate winning with and without an additional vote — depends very much on the closeness of the election.) There are lots of reasons to love lottery voting, including the conventional case for proportional representation, which I very desperately endorse. (See Matt Yglesias and Lee Drutman.) Not to let the best be the enemy of the good, Id favor American experiments in more common forms of proportional representation (multimember districts, party lists), but lottery voting really is the gold standard. It is simple, effective, resistant to entrenchment of incumbents or capture by political parties. The US House of Representatives should be selected by lottery voting today. At the very least, we should start experimenting in some state houses.