View profile

futuribile / curating futures - Issue #14 - #Robotsplaining

Aloha, Here we are with a new round of updates about tech role within our society: a selected storyte
October 31 · Issue #14 · View online
futuribile / curating futures
Here we are with a new round of updates about tech role within our society: a selected storytelling of the best (or worst!) facts and reads from the past couple of weeks.
Next week, as part of the NGI initiative, I will be in Lisbon for the Web Summit: you can find me at the Social Innovation Village on the 7th, or the same night join this workshop I am organising for millennials (1980-2000) together with Jennifer Veldman. “You have ten years to save the Planet” is the (not-so-unrealistic) provocation we will be moving from. Register or drop me a line if you are interested in hosting the workshop elsewhere. Speaking of social innovation, the 12-13 in Seville there is the Social Innovation Community event. I won’t be there (holidays!), but it is definitely worth a trip. You still have time to sign the #SIDeclaration and put some pressure on the EU long-term budget negotiations, so that money is put into social innovation initiatives.
Next issue might be delayed a few days, the perks of being a one-woman-band.
Marta Arniani

Blockchain ta mère
In issue #13 I reported about the Civil startup blockchain-based model to fund journalism. Well, luckily enough for the 14 newsrooms already engaged in the experiment, ConsenSys committed to buy tokens in the second ICO for 3.5M $, because the first one was a total flop. Why such a debacle? First and foremost, buying tokens was way too complicated, buried in a convolution of (“only” 44) steps. Also, major newsrooms declined the offer to join the platform.
The lesson? “You can’t simply say “blockchain” to raise funds anymore”. But, you never know, better be prepared. Demand for blockchain engineers has grown of 400% since end of 2017: it’s like that self-reinforcement phenomenon that makes us buy the ultimate gear when we get to a new sport or hobby, before even knowing the fundamentals (people who barely can swim, but trasform fins and paddles in weapons of messy-mass-destroy at the swimming pool, I am talking to you). 🙅🏻
On a positive note, some argue that blockchain can help making AI better by allowing artificial intelligence networks to access large stores of data without any big company in control of the data or the algorithms.
Artificial Intelligentsia
Ladies and gentlemen, we had the first robot to co-teach a university-level class. Previously on its curriculum: being the first-ever robot to complete a college course. It is called Bina48 and held a course at West Point about… AI ethics. #Robotsplaining 🤖
With how many people directly affected by AI bias have you talked about it? Not many, I bet. As with any ecosystem, marginalized people are likely to suffer the most in an AI-led context that didn’t consider their voice in its design.
We must be wary of replicating the exclusionary practices in the international cooperation model and instead, move towards a paradigm that sees technology design, transfer, and debate as an act of solidarity.
The article is also a mine of references to conferences and documents on the topic.
For sure the more people you include in a debate, the more it gets complicated. As the MIT points out, the whole debate about ethical AI doesn’t take much into account cultural diversity. Take the trolley problem (the thought experiment where you have to decide who to kill). The MIT has a website where you can play with it, Moral Machines, and has finally came to analyse those data:
Participants from collectivist cultures like China and Japan are less likely to spare the young over the old—perhaps, the researchers hypothesized, because of a greater emphasis on respecting the elderly. (…) The results showed that participants from individualistic cultures, like the UK and US, placed a stronger emphasis on sparing more lives given all the other choices.
Evolution of leadership
“I don’t give command nor take command, but I just act as a very reliable channel to contextualize policymaking, so that the civic society can know exactly what’s going on”
Meet Taiwan’s digital minister, Audrey Tang: transgender, hacker, protester, a “conservative anarchist” bringing the g0v philosophy into institutions.
If you are looking for new reads, this article about the remarkable rise of feminist dystopia connects the dots between many books of the genre, arguing that more than producing possible future scenarios they disclose the structure of the present. It insists on the 2016 novel The Power, which conveys how entirely the world is built on male power and male privilege, to the extent that societal structures topple as soon as women are given the advantage. 💪🏽
The current spate of speculative works by and about women is surely a response to a present that itself feels grossly distorted. The process of examining how—and why—our own reality became so troubling is a valuable one, even if only for readers who were already compelled to undertake it. But being absorbed in a world, fleetingly, where women don’t have to be afraid is more than a thought experiment. It’s a profound, powerful relief.
The Truman shopping show
What is the worst thing you can imagine if you put together surveillance, facial recognition and geolocalisation? I tell you. The real-world cookie: the same way online retailers track online visitors behaviour, facial-recognition technology could be used for marketing, helping stores track customers in real life. Marketing and security divisions are being merged.
Retailers are a huge market for facial recognition vendors. Stores, from the small corner one to shopping malls, already have most of the technology in place to start tracking customers: not just the security cameras, but also the cameras inside digital signs and kiosks, which track whether shoppers are paying attention to ads. The basic hardware is there: the next wave is upgrading the software.
“The software often comes with a database of criminals or known shoplifters, which comes from combining the shoplifter registries of participating stores, said Clare Garvie, who studies the technology and its privacy implications at Georgetown Law. It’s unclear exactly what it takes to be put in these databases, let alone how to get your name removed”.
Already last year, Facebook was lobbying hard to limit the regulation of facial recognition. Incidentally, from a survey emerges that Amazon is the second most trusted institution in the United States, behind the military, and followed by Google. Ciao ciao regulations, welcome omnipresent gods of shopping.
Sidewalks of Toronto
Alphabet presented it as the ultimate smart city evolution led by its branch Sidewalk Labs, but it was clear from the beginning that Sidewalk Toronto, the project to create a new kind of mixed-use, complete community on Toronto’s Eastern Waterfront, would have raised major surveillance problems.
The first big coup to the credibility of the project arrived last week, when Ann Cavoukian, a leading privacy expert (she’s behind the Privacy by design approach), stepped down from her advisory role. Her no-go concerns two increasingly evident problems of the data economy: the governance of centralised databases and the grey area constituted by third-parties exploitation of those databases. As she puts it:
“If personally identifiable data are not de-identified at source, we will be creating another central database of personal information (controlled by whom?), that may be used without data subjects’ consent, that will be exposed to the risks of hacking and unauthorized access”.
Cavoukian’s move was a reaction to the presentation of SideWalk Toronto digital governance proposal, which was intended to reassure privacy skeptics by detailing how data collected would be managed by an independent civic data trust, and not owned or controlled by Google. The proposal is a compendium of best practices: the data trust, privacy by design, a Responsible Data Impact Assessment (RDIA) to be filled by third-parties. Surprisingly - suspiciously - for a Google product, nothing seems to be automatised in the process. Sidewalk Labs is operating in a space quite new for regulators:
Existing laws on urban data do not address ownership. And urban data is only regulated when it contains personally identifiable information. Even then, these rules are often not followed in the public realm. We seek to build on them.
Of course, as a private enormous company would. So the question here is: are good practices reassuring enough when the implementing body is this kind of private company?
Almighty threads
“If you wanted to watch 14 episodes of the anime The Melancholy of Haruhi Suzumiya in every possible order, what’s the shortest string of episodes you’d need to watch?” By answering this question, a 4chan user accidentally provided the most elegant solution to part of a mathematical problem involving superpermutations, unsolved since 1993. Internet 1 - Scientific authority 0.
Robbie Barat, the author of the algorithm utilised by the Portrait of Edmond Belamy (the first AI-generated artwork to go to an official sale at Christie’s, see the previous issue) got really upset on Twitter. A worth-browsing thread on copyright, open source, injustice, and the art market. #underdogs
Robbie Barrat
left: the "AI generated" portrait Christie's is auctioning off right now

right: outputs from a neural network I trained and put online *over a year ago*.

Does anyone else care about this? Am I crazy for thinking that they really just used my network and are selling the results?
That’s all for this round! Hit reply to meet in Lisbon and/or send feedback, hit forward to include new people in the loop and make me happy.
Did you enjoy this issue?
If you don't want these updates anymore, please unsubscribe here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue