December 2022 - nexa - server-nexa.polito.it

AI’s fairness problem: understanding wrongful discrimination in the context of automated decision‑making

by Antonio Vetro'

Segnalo per gli iscritti in lista che si interessassero della tematica: Cossette-Lefebvre, H., Maclure, J. AI’s fairness problem: understanding wrongful discrimination in the context of automated decision-making. AI Ethics (2022). https://doi.org/10.1007/s43681-022-00233-w (accesso aperto) Uno dei migliori articoli che ho letto finora sul tema, almeno per quanto riguarda la parte analitica e per il c.d. “framing” della questione. La parte propositiva conclusiva (3 linee guida per l’uso del ML nei processi decisionali) è un po’ meno dettagliata, ma l’ho trovata valida e molto ragionevole. Nota di merito per gli autori che hanno scritto con un linguaggio accessibile a tutti, e non solo agli addetti ai lavori. Un caro saluto A.

Dec. 9, 2022

Re: [nexa] Podcast di Walter Vannini "DataKnightmare: L'algoritmico è politico"

by Marco A. Calamari

On ven, 2022-12-09 at 14:59 +0100, J.C. DE MARTIN wrote: > Conoscete? > https://www.spreaker.com/show/dataknightmare Si, assolutamente da non perdere. L'unico podcast che seguo con regolarità, e mi sono ascoltato anche quasi tutte le stagioni precedenti. Come Cassandra, ma molto più arrabbiato e legato all'attualità Dimenticavo, per chi non conosce Cassandra, le sue profezie e le sue nostalgie ... https://calamarim.medium.com/ https://cassandracrossing.org/ Restando sul politico, anche Das Bose Buro, talvolta è illeggibile, talvolta offre spunti interessanti https://keinpfusch.net/ > jc > _______________________________________________ > nexa mailing list > nexa(a)server-nexa.polito.it > https://server-nexa.polito.it/cgi-bin/mailman/listinfo/nexa

Dec. 9, 2022

Podcast di Walter Vannini "DataKnightmare: L'algoritmico è politico"

by J.C. DE MARTIN

Conoscete? https://www.spreaker.com/show/dataknightmare jc

Dec. 9, 2022

"AI weapons investors get an undisclosed advertorial in the Atlantic"

by J.C. DE MARTIN

*AI weapons investors get an undisclosed advertorial in the Atlantic * /The magazine failed to spell out that Robert Work and Eric Schmidt have financial stakes in the Great Power competition they’re hawking./ December 8, 2022 Written by Eli Clifton <https://responsiblestatecraft.org/author/eclifton/> <https://www.twitter.com/EliClifton> The Atlantic “Ideas” article <https://www.theatlantic.com/ideas/archive/2022/12/us-china-military-rivalry…> had all the trappings of an insightful think piece co authored by one of the most successful former CEOs in Silicon Valley, Eric Schmidt. Schmidt, who headed up Google from 2001 to 2011, writing alongside Robert O. Work, described by The Atlantic as “the 32nd U.S. deputy secretary of defense,” were given over 2000 words to lay out “Offset-X,” a strategy “for the U.S. to restore the technological superiority of its military over all potential adversaries.” At the heart of their strategy is pivot by the Department of Defense to great power competition and the rapid development and deployment of artificial intelligence technologies. But there’s a huge potential conflict of interest that the Atlantic failed to disclose to readers on Monday: Schmidt’s venture capital firm Innovation Endeavorsis an enthusiastic investor <https://www.protocol.com/enterprise/eric-schmidt-ai-china> in AI products for the military. Work, for his part, is chairman of the board for Sparkcognition Government Systems, a company thatdescribes itself <https://sparkgov.ai/> as the “first full-spectrum artificial intelligence company that leverages proven commercial technologies to meet the needs of the most pressing national security missions.” In other words, the policies advocated by the two writers in the Atlantic could provide them with direct financial benefits. [...] continua qui: https://responsiblestatecraft.org/2022/12/08/ai-weapons-investors-get-an-un…

Dec. 9, 2022

Re: [nexa] noyb win: Personalized Ads on Facebook, Instagram and WhatsApp declared illegal

by Antonio

> Il problema di questa notizia e' che manca un qualche riferimento ufficiale. Sto monitorando attentamente svariate fonti ma... finora, nessuna traccia. Da TechCrunch [1] "In line with Art. 65 (5) GDPR, we cannot comment on the content of the decisions until after the Irish DPC has notified the controller of its final decisions," she told TechCrunch. "As indicated in our press release [2], the EDPB looked into whether or not the processing of personal data for the performance of a contract is a suitable legal basis for behavioural advertising, but at this point in time we cannot confirm what the EDPB’s decision in this matter was". The DPC also declined to comment on the newspaper’s report — but deputy commissioner Graham Doyle confirmed to us that it will announce binding decisions on these complaints in early January. We’ve also reached out to Meta for a response to the development. A Meta spokesperson sent this statement: "This is not the final decision and it is too early to speculate. GDPR allows for a range of legal bases under which data can be processed, beyond consent or performance of a contract. Under the GDPR there is no hierarchy between these legal bases, and none should be considered better than any other. We’ve engaged fully with the DPC on their inquiries and will continue to engage with them as they finalise their decision." The company was recently spotted in a filing setting aside €3 billion for data protection fines in 2022 and 2023 — a large chunk of which has yet to land. A. [1] https://techcrunch.com/2022/12/06/meta-gdpr-forced-consent-edpb-decisions/ [2] https://edpb.europa.eu/news/news/2022/edpb-adopts-art-65-dispute-resolution…

Dec. 8, 2022

Re: [nexa] noyb win: Personalized Ads on Facebook, Instagram and WhatsApp declared illegal

by Damiano Verzulli

Il 07/12/22 18:42, Enrico Nardelli ha scritto: > > As reported by The Wall Street Journal > (https://www.wsj.com/articles/metas-targeted-ad-model-faces-restrictions-in-…) > the EDPB has decided that Meta cannot force users to agree to > personalized ads. [...] Now, 4.5 years later, the European Data > Protection Board (EDPB) found Meta's alleged "bypass" of the GDPR > illegal. The EDPB also rejected the view of the Irish Data Protection > Commission (DPC) who previously sided with Meta, after taking four > years to investigate the case. > > https://noyb.eu/en/noyb-win-personalized-ads-facebook-instagram-and-whatsap… > Il problema di questa notizia e' che manca un qualche riferimento ufficiale. Sto monitorando attentamente svariate fonti ma... finora, nessuna traccia. Nell'articolo su noyb.eu leggo: "The EDPB decision itself was not published, but will be published together with the final decision of the DPC in January 2023." Per l'ufficialità, insomma... c'e' da aspettare un po'. Bye, DV -- Damiano Verzulli e-mail: damiano(a)verzulli.it --- possible?ok:while(!possible){open_mindedness++} --- "...I realized that free software would not generate the kind of income that was needed. Maybe in USA or Europe, you may be able to get a well paying job as a free software developer, but not here [in Africa]..." -- Guido Sohne - 1973-2008 http://ole.kenic.or.ke/pipermail/skunkworks/2008-April/005989.html

Dec. 7, 2022

noyb win: Personalized Ads on Facebook, Instagram and WhatsApp declared illegal

by Enrico Nardelli

As reported by The Wall Street Journal (https://www.wsj.com/articles/metas-targeted-ad-model-faces-restrictions-in-…) the EDPB has decided that Meta cannot force users to agree to personalized ads. In May 2018, when the GDPR came into force in the EU, Meta Ireland Ltd. believed it could "bypass" the requirement to get opt-in consent from users, by simply adding a provision in the terms and conditions. On 25 May 2018, the digital rights organization noyb filed complaints with the relevant Data Protection Authorities (DPAs). Now, 4.5 years later, the European Data Protection Board (EDPB) found Meta's alleged "bypass" of the GDPR illegal. The EDPB also rejected the view of the Irish Data Protection Commission (DPC) who previously sided with Meta, after taking four years to investigate the case. https://noyb.eu/en/noyb-win-personalized-ads-facebook-instagram-and-whatsap… -- -- EN https://www.hoepli.it/libro/la-rivoluzione-informatica/9788896069516.html ====================================================== Prof. Enrico Nardelli Presidente di "Informatics Europe" Direttore del Laboratorio Nazionale "Informatica e Scuola" del CINI Dipartimento di Matematica - Università di Roma "Tor Vergata" Via della Ricerca Scientifica snc - 00133 Roma home page: https://www.mat.uniroma2.it/~nardelli blog: https://link-and-think.blogspot.it/ tel: +39 06 7259.4204 fax: +39 06 7259.4699 mobile: +39 335 590.2331 e-mail: nardelli(a)mat.uniroma2.it online meeting: https://blue.meet.garr.it/b/enr-y7f-t0q-ont ====================================================== --

Dec. 7, 2022

Sarco Suicide Pod aims to take assisted dying out of doctors' hands with AI

by 380°

Buonasera, la notizia è piuttosto vecchiotta, l'articolo è del 9 Dic. 2021 mentre il "Sarco pod" [1] è stato inventato nel 2017. https://www.euronews.com/next/2021/12/08/the-sarco-suicide-pod-aims-to-take… «The Sarco suicide pod aims to take assisted dying out of doctors' hands with AI and 3D printing» --8<---------------cut here---------------start------------->8--- A 3D-printed "suicide capsule" has passed a legal review in Switzerland, potentially clearing the way for the technology to be put into use in the country's legal assisted suicide clinics. The Sarco capsule's creator, Dr. Philip Nitschke, told Euronews Next that his aim was to allow anyone to download the design and print it themselves. In the future, an AI screening process will allow Nitschke's assisted dying advocacy organisation, Exit International, to "demedicalise" the dying process by removing the need for medical professionals to be involved, he said. --8<---------------cut here---------------end--------------->8--- La notizia è rilanciata anche dalla stessa azienda nel suo media book: https://www.exitinternational.net/sarco/sarco-aims-to-take-assisted-dying-o… Anche Gizmondo ne parla https://gizmodo.com/controversial-assisted-suicide-pod-cleared-for-use-in-s… --8<---------------cut here---------------start------------->8--- “Our aim is to develop an artificial intelligence screening system to establish the person’s mental capacity,” Nitschke told the site. “Naturally there is a lot of skepticism, especially on the part of psychiatrists. But our original conceptual idea is that the person would do an online test and receive a code to access the Sarco.” --8<---------------cut here---------------end--------------->8--- L'azienda nelle sue FAQ scrive: https://www.exitinternational.net/sarco/faqs/ --8<---------------cut here---------------start------------->8--- The Sarco is not and never will be for sale. The plans will be published in The Peaceful Pill eHandbook. The cost of Sarco will depend on what is charged by the 3D print shop. While the price is dropping all the time, it will never be very cheap. In 2022, the cost to print a Sarco is around €17,000. In the early stages of development it was thought that Sarco could double as one's coffin. If the Sarco is printed from suitable biodegradable materials, this is still true. But the high printing cost of Sarco intimates that it may be better to have a Sarco that can be used multiple times over. Better and cheaper for the user. More friendly to the planet. [...] Exit believes that the right to a good death at a time of one's choosing should be a fundamental human right afforded to all adults of sound mind. What type of gift is life if you cannot give it away? Because Sarco uses no drugs, no doctors need be involved (except for an assessment of mental capacity). Sarco is a game-changer because it will allow a person of sound mind control over their own death. No longer will they have to beg a panel of doctors to grant their final wish (by prescribing lethal drugs). --8<---------------cut here---------------end--------------->8--- A quanto pare l'idea dell'inventore di sostituire la commissione medica con un "mental capacity assessment" svolto da non meglio precisata intelligenza artificiale è seria. Saluti, 380° [1] https://en.wikipedia.org/wiki/Sarco_pod -- 380° (Giovanni Biscuolo public alter ego) «Noi, incompetenti come siamo, non abbiamo alcun titolo per suggerire alcunché» Disinformation flourishes because many people care deeply about injustice but very few check the facts. Ask me about <https://stallmansupport.org>.

Dec. 7, 2022

Re: [nexa] Ericsson Mobility report: nel 2028 i video occuperanno l'80% del traffico mobile.

by Antonio

> ... La buonanima di Godard e quel santo di Tarkovskij ci resterebbero male a sentire che non ci piace il cinema/il video Godard, Tarkovskij ... Kubrick, Bergman, Fellini, Renoir, Gilliam, Hitchcock, Monicelli, Rossellini (e chissà quanti ne dimentico) registi geniali. Cinema come settima arte, da cinefilo non può non piacermi ma il video, usa e getta, il "tuovideo", è un'altra cosa. Sartori, nel suo "homo videns" si scagliava contro la Televisione, io, nel mio piccolo, contro l'"homo digitalis" prosumer di video, anzi, di "prodotti visivi". Non ci sono bastati decenni di TV spazzatura, ci mancava la videoteca-net spazzatura. Certo, magari uno/una su un milione diventerà la Chiara Ferragni del futuro, o verrà chiamata/o da Fiorello perché ha 3,5 milioni di follower su Tik Tok, ma quante macerie lasceranno gli altri 999998? Su youtube ci sono partite di calcio giocate ... alla playstation con milioni di visualizzazioni, si può essere caduti così in basso? Mi verrebbe da chiudere questa gente in un cineclub a sorbirsi a ripetizione la "corazzata Potemkin" di Eisenstein come nel "Secondo tragico Fantozzi" del grande Paolo Villaggio. A.

Dec. 6, 2022

Re: [nexa] Ericsson Mobility report: nel 2028 i video occuperanno l'80% del traffico mobile.

by M. Fioretti

On Tue, Dec 06, 2022 12:54:42 PM +0100, Andrea Bolioli via nexa wrote: > Il fatto che adesso abbondino i video scemi su TikTok non vuol dire > che di per sè il video sia un linguaggio scemo. no, ci mancherebbe altro. Ma... > La formazione di una persona in modo che diventi libera e capace di > autodeterminarsi non passa solo dalla lettura di libri primo, il problema e' tutto nel "solo". Magari fra 2/3 secoli sara' diverso, ma oggi e per almeno qualche altra generazione la stragrande maggioranza della cultura e conoscenza umana che e' importante conoscere, almeno da Iliade e Bibbia in poi, e' testo. Chi non legge perche' preferisce il video si condanna a capirci poco o niente di dove, come e perche' vive, o a dover fare molta piu' fatica per arrivarci. secondo, sempre nel prevedibile futuro, quanto sopra e' dovuto anche al fatto che il testo e' anche infinitamente piu' riutilizzabile, "mixabile" e analizzabile del video, anche automaticamente, per aumentare e condividere conoscenza terzo, il problema vero forse e' la versione "next level" dell'usare fogli elettronici a sproposito quando invece servirebbe un qualsiasi database: ovvero l'infame abitudine di usare video INVECE di testo o immagini (*), quando i secondi sarebbero infinitamente piu' efficaci, vedi: https://stop.zona-m.net/2020/07/video-tutorials-considered-harmful/ Marco (*) idem per whatsapp audio invece che scritti, ovviamente... -- Help me write my NEXT MILLION WORDS for digital awareness: https://stop.zona-m.net/2021/10/funding-2021-2022/

Dec. 6, 2022