May 2019 - nexa - server-nexa.polito.it

accordo Poste Microsoft

by armeo＠mail.nexacenter.org

carissimi, nessuno degli amici di Nexa ha reagito a questo mio mail che qui ricopio: "" Poste non è una pubblica amministrazione, come giustamente rileva Stefano, ma è una società per azioni posseduta - non sono sicuro - dallo Stato nella misura del 60%. A prescindere dal fatto che penso che la privatizzazione di quel servizio sia stata una follia (ma so che questa mia opinione non è condivisa dalle persone importanti), gli uomini dello Stato presenti nel Consiglio di amministrazione si sono posti il quesito dei costi e benefici per il nostro Paese dell'accordo con la Microsoft? Ad esempio (ma questa è solo una parte del problema), quanto costerà al nostro Paese, che vive un momento di tragica condizione economica, l'esborso verso un'aziende straniera per un lavoro che poteve essere svolto benissimo da operatori pubblici e privati del nostro paese?"" Se avessi scritto una mail con una proposta sulle "fake news" (che sono una realtà imperante su tutte le comunità di ominidi "sapientes" da molti millenni) molti sarebbero intervenuti. Ma su una notizia così grave, non so se per condivisione dell'iniziativa o per rassegnazione, nessuno interviene neppure su Nexa. Formulo una proposta precisa: affidare a una struttura pubblica del Paese (da torinese sto pensando a CSP, CSI, TOPIX, ma anche a CINECA e a molte strutture di ricerca oltre ai dipartimenti di informatica delle università del nostro Paese) lo sviluppo di software libero per attuare le funzionalità dei servizi messi a disposizione da Microsoft e alle stesse Poste la gestione dei relativi servizi. Analoghe considerazioni si applicano all'accordo Confindustria-IBM. Raf

May 27, 2019

Who to Sue When a Robot Loses Your Fortune - Bloomberg

by Alberto Cammozzo

<https://www.bloomberg.com/news/articles/2019-05-06/who-to-sue-when-a-robot-…> Robots are getting more humanoid every day, but they still can’t be sued. So a Hong Kong tycoon is doing the next best thing. He’s going after the salesman who persuaded him to entrust a chunk of his fortune to the supercomputer whose trades cost him more than $20 million. The case pits Samathur Li Kin-kan, whose father is a major investor in Shaftesbury Plc, which owns much of London’s Chinatown, Covent Garden and Carnaby Street, against Raffaele Costa, who has spent much of his career selling investment funds for the likes of Man Group Plc and GLG Partners Inc. It’s the first-known instance of humans going to court over investment losses triggered by autonomous machines and throws the spotlight on the “black box” problem: If people don’t know how the computer is making decisions, who’s responsible when things go wrong? [...]

May 26, 2019

"Perhaps the best/only way to explain neural networks to a five-year-old"

by J.C. DE MARTIN

HT all'account twitter MIT CSAIL, che ha anche coniato il subject di questo email. jc https://www.reddit.com/r/explainlikeimfive/comments/7buzbs/eli5_what_are_ne… "The little league team you coach just won the big game, and you ask them if they want to go out for pizza or for burgers. Each kid starts screaming their preference, and you go with whatever was the loudest. This is basically how a neural net works but on multiple levels. The top-level nodes get some input, each detects a certain property and screams when it sees it...the more intense the property, the louder they scream. Now you have a bunch of nodes screaming "it's dark!", "it's has red!", "it's roundish!" as various volumes. The next level listens and based on what they hear they start screaming about more complex features. "It has a face!", "It has fur", until finally get to a level where it is screaming "It's a kitty!". The magic part is no one tells them when to scream, it is based on feedback. Your little league team went for burgers, and some of them got sick. Next week, they might not scream for burgers, or might not scream as loudly. They have collectively learned that burgers might not have been a great choice, and are more likely to lean away from the option. A neural net gets training in much the same way. You feed it a bunch of kitty and non-kitty pictures. If the net gets it right, the nodes are reinforced so they are more likely to do the same thing in similar situations. If it is wrong, they get disincentivized. Initially, its results will be near random, but if you have designed it correctly, it will get better and better as the nodes adjust. You often have neural nets that work without any human understanding exactly how."

May 25, 2019

Re: [nexa] Platform capitalism and the future of international education | Discover Society

by Enrico Nardelli

Molto interessante anche questo passaggio "platform capitalism has significant internal fissures that inevitably lead to a disembodying of student and teacher and a tendency for the university to move away from their role as nurturers of student learning and growth and guardians of knowledge, towards a position where students and teachers are seen within the data gathered around them as crops to be harvested. Education systems in this comparative environment have come under increasing pressure to homogenise, with PISA and other data driven policy drivers suggesting a ‘best way’ of doing education. This in turn has put limits on countries’ abilities to have local solutions, often to controversial effect. This trend has been mirrored in the university sector with unprofitable courses being shut down and pressures increasing to get students through courses in order to maintain student numbers. This education marketplace looks only set to expand." Il 23/05/2019 09:18, J.C. DE MARTIN ha scritto: > “Governments are beginning to realise that platforms cannot be allowed to grow exponentially unwatched and unregulated. Rather they should insist that workers’ rights are protected and people’s data safeguarded” > > > https://discoversociety.org/2019/05/01/platform-capitalism-and-the-future-o… > > > > (Sent from my wireless device; please excuse brevity and typos (if any)) > > _______________________________________________ > nexa mailing list > nexa(a)server-nexa.polito.it > https://server-nexa.polito.it/cgi-bin/mailman/listinfo/nexa > -- EN ===================================================================== Prof. Enrico Nardelli Dipartimento di Matematica - Universita' di Roma "Tor Vergata" Via della Ricerca Scientifica snc - 00133 Roma tel: +39 06 7259.4204 fax: +39 06 7259.4699 mobile: +39 335 590.2331 e-mail: nardelli(a)mat.uniroma2.it home page: http://www.mat.uniroma2.it/~nardelli blog: http://www.ilfattoquotidiano.it/blog/enardelli/ http://link-and-think.blogspot.it/ ===================================================================== --

May 25, 2019

Data Violence and How Bad Engineering Choices Can Damage Society

by J.C. DE MARTIN

Data Violence and How Bad Engineering Choices Can Damage Society Cultural harms can go well beyond search results — which can be bad news for vulnerable communities Anna Lauren Hoffmann https://medium.com/s/story/data-violence-and-how-bad-engineering-choices-ca… (Sent from my wireless device; please excuse brevity and typos (if any))

May 25, 2019

Re: [nexa] Who or what is to blame when you’re harmed by an algorithm

by Stefano Quintarelli

> not the first involving the liability of algorithms. In March of 2018 ricordo che il Prof Maiocchi (informatica all'Unimi), quando feci l'universita', aveva un libro con N esempi di liability degli algoritmi (ovvero dei produttori/venditori) ciao, s. On 24/05/2019 17:05, Enrico Poli wrote: > > Non mi sembra di averlo visto passare il lista, e credo risuoni con > alcune discussioni che ho letto qui: > > https://go.technologyreview.com/who-or-what-is-to-blame-when-youre-harmed-b… > > > «A Hong Kong tycoon lost more than $20 million dollars after entrusting > part of his fortune to an automated platform. Without a legal framework > to sue the technology, he placed the blame on the nearest human instead: > the man who sold it to him. > It’s the first known case over automated investment losses, but not the > first involving the liability of algorithms. In March of 2018, a > self-driving Uberstruck and killed > <https://www.theverge.com/2018/3/28/17174636/uber-self-driving-crash-fatal-a…>a > pedestrian in Tempe, Arizona, sending another case to court. A year > later, Uber wasexonerated > <https://qz.com/1566048/uber-not-criminally-liable-in-tempe-self-driving-car…>from > all criminal liability, but the safety driver could face charges of > vehicular manslaughter instead. > > […] > > In other case studies, Elish found the same pattern to hold true: even > in a highly automated system where humans have limited control of its > behavior, they still bear the brunt of its failures. Elish calls this > phenomenon a “moral crumple zone.” “While the crumple zone in a car is > meant to protect the human driver,” she writes in her paper, “the moral > crumple zone protects the integrity of the technological system, at the > expense of the nearest human operator.” Humans act like a “liability > sponge,” she says, absorbing all legal and moral responsibility in > algorithmic accidents no matter how little or unintentionally they are > involved. > > This pattern offers important insight into the troubling way we speak > about the liability of modern AI systems. In the immediate aftermath of > the Uber accident, headlines pointed fingers at Uber, but less than a > few days later, the narrative shifted > <https://www.theverge.com/2018/3/28/17174636/uber-self-driving-crash-fatal-a…> to > focus on the distraction of the driver. > > “We need to start asking who bears the risk of [tech companies’] > technological experiments,” says Elish. Safety drivers and other human > operators often have little power or influence over the design of the > technology platforms they interface with. Yet, in the current regulatory > vacuum, they will continue to pay the steepest cost.» > > > > > > _______________________________________________ > nexa mailing list > nexa(a)server-nexa.polito.it > https://server-nexa.polito.it/cgi-bin/mailman/listinfo/nexa > -- reserve your meeting with me at http://cal.quintarelli.it

May 24, 2019

Yes but who are the persons who micro-work to make France's AI possible?

by Antonio Casilli

Dear all, In February I shared with you our preliminary study estimating the number of persons micro-working in France, based on the first results of our research projet DiPLab. Today, we published the complete results, featured in the report Micro-work in France. Behind Automation, New Forms of Precarious Labour? ( Le Micro-Travail en France. Derrière l’automatisation, de nouvelles précarités au travail ? , 73 pages, available in French here [ http://diplab.eu/ | http://diplab.eu/ ] , with an executive summary in English). Micro-work is accessible in France through at least 23 digital platforms which sport 15,000 'very active' workers and more than 250,000 'casual' ones. While not all of them micro-work every day, they constitute a large reservoir from which the industry can draw according to its needs. Their activity consists in highly fragmented online or mobile-based micro-tasks, generally paid by the piece. Most often, these tasks require low qualification and are equally poorly paid (sometimes just a few cents. The socio-demographic profile of micro-workers in France DiPLab researchers analyzed native web data, administered an online survey to nearly 1000 micro-workers, and conducted more than 90 interviews with workers, business and platform owners. According to the report: ➡️56% of the respondents identify as women; ➡️63% are 25-44 years old; ➡️43.5% of micro-workers hold at least a bachelor’s degree. The geographical distribution of micro-work follows the distribution of the working population and national income. While micro-work is neither exclusively urban nor concentrated in Paris, it does not seem to constitute an opportunity to access the labor market and to increase the purchasing power for rural populations or residents in low-income regions. The average monthly income of micro-work in France, is very unevenly distributed and bears on average 21€euros per month. This income is sometimes complementary to a main one. Yet it is very often a necessary source of financial support, especially for over 22% of the micro-workers who live below the poverty threshold. Micro-work plays a key role in the development of French (and European) AI Micro-work plays a major role in the development of artificial intelligence (AI) by French companies. By performing repetitive tasks, micro-workers annotate, label, enrich the data used for the deployment of algorithms, connected speakers, virtual assistants. DiPLab respondent perform three main types of micro-tasks: ➡️AI preparation: micro-workers generate or annotate large training datasets used to calibrate machine learning solutions; ➡️AI verification: micro-workers oversee existing AI technologies, by correcting and checking their results; ➡️AI impersonation: micro-workers performe human tasks which are successively repackaged and branded as “artificial” intelligence. Micro-work has a huge impact on present day's technology and economy, as it modifies both employment relationships and market structures. This is why it is paramount to assess the societal consequences of this new form of work. How can we regulate this new labor force and strengthen its social protection? Above all, given the predominant place of micro-work in the development and commercialization of artificial intelligence solutions, how can we ensure that micro-workers' contributions to technological innovation are recognized at their actual value? Hope you find this interesting. Please feel free to share. Antonio -- Antonio A. Casilli Associate Professor (HDR), Telecom ParisTech Member, Interdisciplinary Institute for Innovation (i3 UMR 9217 CNRS) Associate Member, LACI-IIAC (EHESS) Faculty Fellow, Nexa Center for Internet & Society

May 24, 2019

accordo IBM-Confindustria

by armeo＠mail.nexacenter.org

A differenza del caso Poste-Microsoft, la scelta di Confindustria a favore del cosidetto "ibrido" IBM mi pare giuridicamente lecita. Ma anche in questo caso, l'analisi costi benefici per il nostro Paese mi pare disastrosa. Raf

May 24, 2019

Who or what is to blame when you’re harmed by an algorithm

by Enrico Poli

Non mi sembra di averlo visto passare il lista, e credo risuoni con alcune discussioni che ho letto qui: https://go.technologyreview.com/who-or-what-is-to-blame-when-youre-harmed-b… «A Hong Kong tycoon lost more than $20 million dollars after entrusting part of his fortune to an automated platform. Without a legal framework to sue the technology, he placed the blame on the nearest human instead: the man who sold it to him. It’s the first known case over automated investment losses, but not the first involving the liability of algorithms. In March of 2018, a self-driving Uber struck and killed <https://www.theverge.com/2018/3/28/17174636/uber-self-driving-crash-fatal-a…> a pedestrian in Tempe, Arizona, sending another case to court. A year later, Uber was exonerated <https://qz.com/1566048/uber-not-criminally-liable-in-tempe-self-driving-car…> from all criminal liability, but the safety driver could face charges of vehicular manslaughter instead. […] In other case studies, Elish found the same pattern to hold true: even in a highly automated system where humans have limited control of its behavior, they still bear the brunt of its failures. Elish calls this phenomenon a “moral crumple zone.” “While the crumple zone in a car is meant to protect the human driver,” she writes in her paper, “the moral crumple zone protects the integrity of the technological system, at the expense of the nearest human operator.” Humans act like a “liability sponge,” she says, absorbing all legal and moral responsibility in algorithmic accidents no matter how little or unintentionally they are involved. This pattern offers important insight into the troubling way we speak about the liability of modern AI systems. In the immediate aftermath of the Uber accident, headlines pointed fingers at Uber, but less than a few days later, the narrative shifted <https://www.theverge.com/2018/3/28/17174636/uber-self-driving-crash-fatal-a…> to focus on the distraction of the driver. “We need to start asking who bears the risk of [tech companies’] technological experiments,” says Elish. Safety drivers and other human operators often have little power or influence over the design of the technology platforms they interface with. Yet, in the current regulatory vacuum, they will continue to pay the steepest cost.»

May 24, 2019

Re: [nexa] accordo Poste -Microsoft

by armeo＠mail.nexacenter.org

Poste non è una pubblica amministrazione, come giustamente rileva Stefano, ma è una società per azioni posseduta - non sono sicuro - dallo Stato nella misura del 60%. A prescindere dal fatto che penso che la privatizzazione di quel servizio sia stata una follia (ma so che questa mia opinione non è condivisa dalle persone importanti), gli uomini dello Stato presenti nel Consiglio di amministrazione si sono posti il quesito dei costi e benefici per il nostro Paese dell'accordo con la Microsoft? Ad esempio (ma questa è solo una parte del problema), quanto costerà al nostro Paese, che vive un momento di tragica condizione economica, l'esborso verso un'aziende straniera per un lavoro che poteve essere svolto benissimo da operatori pubblici e privati del nostro paese? Raf Il 24/05/2019 10:54, Stefano Quintarelli ha scritto: > Sono cose diverse > > Uno è l'obbligo di fare una valutazione prestazioni/ prezzo e > scegliere l'alternativa migliore. Questa è cosa normale per ogni > acquisto della p.a. e l'enforcement si basa sul danno erariale > eventualmente causato. un cittadino potrebbe fare un esposto alla > Corte dei conti, in caso di spesa ingiustificata. > > Un altro è l'obbligo, che abbiamo messo nel CAD, di dare il codice > sorgente di ogni software sviluppato ad ogni soggetto economico sia > interessato (e quindi non più solo ad altre amministrazioni, ma anche > ad altri sviluppatori o software house o analisti o whatever). La > versione precedente, ovvero il riuso da parte dell'amministrazione non > aveva funzionato, perché ognuna ha bisogno di cose particolari. Questo > è un cambiamento significativo. Potete scrivere al comune e chiedere i > sorgenti, che so, del sistema di whistleblowing, e il comune ve li > deve dare. > > Poi, ricordo che poste non è una pubblica amministrazione. > > Ciao! > > In data 24 maggio 2019 10:32:55 AM "armeo(a)mail.nexacenter.org" > <meo(a)polito.it> ha scritto: > > Non è che le norme italiane non hanno funzionato a dovere. Più > semplicemente sono state ignorate. > > Vedi, ad esempio, la recente notizia apparsa su Key4iz: > > > Poste italiane, accordo con Microsoft per la trasformazione digitale > del Gruppo > > Non c'è difesa contro il potere. > > Raf > > > Il 24/05/2019 10:23, J.C. DE MARTIN ha scritto: > *Pa, al via l’obbligo di pubblicare in open source tutto il proprio > codice** > * > –di Luca Tremolada > > 23 maggio 2019 > > Con la pubblicazione in Gazzetta ufficiale è al via l'obbligo per le > Pubbliche amministrazioni di pubblicare in open source tutto il > proprio codice e di valutare software già esistente prima di > realizzarne di nuovo. Le norme italiane sono tra le più avanzate > d'Europa in materia di open source. Finora però non hanno funzionato a > dovere. > > https://www.ilsole24ore.com/art/tecnologie/2019-05-20/pa-via-l-obbligo-pubb… > > <https://www.ilsole24ore.com/art/tecnologie/2019-05-20/pa-via-l-obbligo-pubb…> > > > > > _______________________________________________ > nexa mailing list > nexa(a)server-nexa.polito.it > https://server-nexa.polito.it/cgi-bin/mailman/listinfo/nexa > > > > ---------- > _______________________________________________ > nexa mailing list > nexa(a)server-nexa.polito.it > https://server-nexa.polito.it/cgi-bin/mailman/listinfo/nexa > > >

May 24, 2019