March 2018 - nexa - server-nexa.polito.it

France leads the open source market in Europe
by Marco Ciurcina May 10, 2018

May 10, 2018

France leads the European open source market, with penetration in IT services and software higher than that of other countries, says the French company Pierre Audoin Consultants (PAC) in a survey published during the Paris Open Source Summit (POSS 2017). In France, the open source market reached EUR 4.46 billion in 2017, with annual growth of 8.3%. Open source represents 9.9% of the total French IT market, compared to 6.4% in Germany and 6.5% in the United Kingdom. continua qui https://joinup.ec.europa.eu/news/market qualcuno conosce la percentuale dell'Italia? m.c.

4 8

Raccomandazioni della Commissione europea sulla robotica
by Diego Giorio March 31, 2018

March 31, 2018

Con l'occasione, buona Pasqua a tutti D. http://www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT+REPORT+A8-… RELAZIONE recante raccomandazioni alla Commissione ...<http://www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT+REPORT+A8-…> www.europarl.europa.eu proposta di risoluzione del parlamento europeo allegato alla proposta di risoluzione:raccomandazioni particolareggiate concernenti il contenuto della proposta richiesta

2 1

Re: [nexa] Fwd: Publication du rapport Villani sur l'intelligence artificielle
by Giuseppe Attardi March 30, 2018

March 30, 2018

Sottolineo dei punti su cui avevamo discusso con lui a Bologna: Une voie européenne à trouver pour contrer l'offensive des Gafa. Parmi les pistes envisagées figure la création d'un "cloud souverain européen" qui pourrait s'appuyer sur le groupe français OVH et la création d'une agence publique sur le modèle de la DARPA américaine, qui fut à l'origine d'internet. Les pays qui seront les leaders dans le domaine de l’intelligence artificielle seront amenés à capter une grande partie de la valeur des systèmes qu’ils transforment, mais également à contrôler ces mêmes sys tèmes, mettant en cause l’indépendance des autres pays.

2 1

Car Wars
by Giacomo Tesio March 30, 2018

March 30, 2018

Salve, vi propongo una lettura interessante sulle auto a guida autonoma, fantascientifica ma non troppo, dalla Deakin University (Australia) con sondaggio opzionale incluso. http://this.deakin.edu.au/lifestyle/car-wars Ad un certo punto, un bambino deve fare una ricerca sulla Siria Then came the car thing. Just like that one in Australia, except this > wasn’t random terrorists killing anyone they could get their hands on – > this was a government. We all watched the live streams as the > molotov-chucking *terrorists or revolutionaries or whatever* in the > streets of Damascus* were chased through the streets by the cars that the > government had taken over*, some of them – most of them! – *with > horrified people trapped inside*, pounding on the emergency brakes as > their cars ran down the people in the street [...] > Fa riflettere. Da un lato, la proprietà comunitaria dei mezzi di trasporto (autonomo o meno) sarà una direzione quasi obbligata da un punto di vista ecologico. Ma le implicazioni politiche del rimuovere le coscienze umane dalla guida dei veicoli (o dal controllo degli automi), potrebbero essere più complesse di quanto immaginato. E' molto più facile per una persona controllare 1000 automi che controllare 10 persone. Probabilmente, come proponeva Raffaele Meo in un'altro scambio, sarebbe più saggio limitare la circolazione dei futuri veicoli autonomi a strade dedicate sotterranee, in una specie di evoluzione della metropolitana. Si tratterebbe di un problema *molto *più semplice dal punto di vista computazionale. Un altro passaggio, che diventerà una delle mie citazioni preferite: Enough to satisfy their honor, before capitulating to *the unstoppable > force of non-technical certitude about deeply technical subjects*. The > conviction that once the lawyers had come up with the answer, it was the > engineers’ job to implement it, not trouble them with tedious technical > wheedles about what was and wasn’t possible. Fino a quando programmare non sarà parte della nostra cultura condivisa, come leggere, scrivere e contare, l'onere di questa conoscenza sarà quella di impedirne un uso sconsiderato. Giacomo

2 1

La Poste rode un abonnement illimité pour la livraison de colis
by Luigi Scorca March 30, 2018

March 30, 2018

Amazon Prime, La Poste Second.. TBC Interessante notare come anche nel mercato delle consegne si vada nella stessa direzione della telefonia.. https://www-lesechos-fr.cdn.ampproject.org/c/s/www.lesechos.fr/amp/74/21641…

1 0

Facebook VP wrote site's actions were 'de facto good' – even if they led to deaths | Technology | The Guardian
by Alberto Cammozzo March 30, 2018

March 30, 2018

<https://www.theguardian.com/technology/2018/mar/29/facebook-memo-mission-an…> [] A top Facebook executive warned in a leaked memo that the platform could lead to deaths and could help terrorists plan attacks, but argued that the negative outcomes were a reasonable byproduct of the company’s broader “growth tactics” and mission to “connect” people. The public disclosure of the 2016 memo, in which a vice-president, Andrew “Boz” Bosworth, wrote “anything that allows us to connect more people more often is *de facto* good”, prompted the CEO, Mark Zuckerberg, to defend his company’s mission in a hastily released statement on Thursday. Zuckerberg said Facebook “never believed the ends justify the means” and said he strongly disagreed with the memo circulated by Bosworth, whom he described as a talented but provocative force in the company. []

1 0

Fwd: Machine-Learned Automated Decision Making Report and Panel in Brussels
by Enrico Nardelli March 29, 2018

March 29, 2018

Vi segnalo i seguente rapporto When Computers Decide: European Recommendations on Machine-Learned Automated Decision Making <http://www.informatics-europe.org/component/phocadownload/category/10-repor…> che abbiamo prodotto congiuntamente tra Informatics Europe e ACM Europe Council C'è anche una pressrelease sul nostro sito web news article on our website. <http://www.informatics-europe.org/news/435-ethics_adm.html> Ciao, Enrico -- 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/ ===================================================================== --

8 18

WIRED: The Next Cold War Is Here, and It's All About Data
by Alberto Cammozzo March 29, 2018

March 29, 2018

GDPR diventerà uno standard globale? <https://www.wired.com/story/opinion-new-data-cold-war> [] Combatants in the new Cold War are fighting over the currency of the modern age: personal information. The battles are over who controls data. Vying against each other are those societies that believe that individuals have an absolute right to control their personal data—to exercise the same kind of dominion over data that they do over their bodies or their personal property—and those that believe that personal data is a good to be traded on the open market and thus subject to the same market forces at play elsewhere. May the most innovative, efficient company win. The EU stands firmly for the interests of the individual. The regulatory language of the GDPR cogently expresses its view, harmonizing data protection rules throughout the EU and requiring that any company, anywhere, must respect the data rights of EU citizens, or face stiff penalties. Europeans must provide positive consent for the ways their data is used, and they have the right to access and erase that data, as well as the “right to be forgotten.” In the opposite corner sits the United States and the giant US corporations that trade in personal data for profit, and whose practices have expanded largely unchecked. One ideology puts the control of personal data in the hands of the individual, the other cedes that control to the corporation. (A third approach is state control of data, which is emerging as China’s social credit system, though that remains as yet an internal policy.) But these differing views about data protection cannot jostle for dominance for much longer. As trade grows increasingly global, it’s becoming clear that personal data crosses borders far too easily for contrasting models to co-exist. Now, weeks before the GDPR goes into effect, the evidence is mounting that the EU’s approach will dominate. One good measure of this dominance is the speed with which countries around the world are recognizing the supremacy of the EU data protection standard by adopting models that align closely with the GDPR. The number of countries that have attained official EU recognition of the “adequacy” of their standards grow steadily; South Korea and Japan will join the list soon. For their part, EU leaders are clear in their intent. “We want to set the global standard,” Věra Jourová, the European commissioner for justice, told POLITICO last year. “Privacy is a high priority for us.” And so it must also be for those who wish to trade with this powerful economic bloc. Even in the United States, there are signs of movement toward embracing higher standards for data protection. The United States offers a program, called Privacy Shield, that enables American companies to certify that their data protection practices meet EU standards (though this program is questioned by privacy purists in the EU). And some of the most trusted US corporations go to great lengths to respect the data rights of people in other countries (see the recent Microsoft case before the Supreme Court). Likewise, individual states establish GDPR-like laws for their citizens (see New York’s recent cybersecurity regulation). The sheer volume of companies that are willingly modifying their data protection practices, at great cost, to become “GDPR compliant” should be evidence enough that there is appetite in the US business community for the certainty of a unified data protection regime. []

2 1

Re: [nexa] Responsabilità legale e ML
by Giacomo Tesio March 29, 2018

March 29, 2018

Il giorno 29 marzo 2018 10:10, Pier Giorgio Chiara <pg.chiara94(a)gmail.com> ha scritto: > il fatto che sia estremamente costoso, come giustamente fai notare, non > può essere considerato ostacolo decisivo all'attribuzione di una sorta di > responsabilità oggettiva? > No. A meno che il denaro non valga più della vita umana. Si chiama *proprietà privata*. Se non ti puoi permettere una Ferrari, non la compri. Non puoi rubare i soldi degli altri (ovvero scaricare i costi sugli altri). > in fondo si richiederebbe uno sforzo probante che solo le aziende più > grande -forse- sarebbero in grado di sostenere. > Francamente con la tecnologia attuale, non conosco nessuna azienda in grado di sostenere questo sforzo per una rete neurale discretamente complessa. NESSUNA. Le parole chiavi sono "tecnologia attuale" e "complessità". Su una rete neurale semplice, invece è possibile. E' una questione di equilibrio. Nota però che io non sto ancora parlando dell'aspetto probante. Il costo di cui stiamo parlando va assorbito PRIMA di immettere un prodotto pericoloso sul mercato. > Ciò non scoraggerebbe la ricerca? > AL CONTRARIO! La necessità di AI affidabili, spiegabili e prevedibili sta già lanciando grandi finanziamenti alla ricerca in questo senso. Le AI spiegabili (la cui spiegabilità al momento è ancora molto deludente) è un tema caldissimo della ricerca, letteralmente pieno di soldi. Perché le aziende che puntano sulla AI sono le prime a rendersi conto del castello di sabbia su cui stanno costruendo... E d'altro canto, ci sono SEMPRE altre soluzioni (magari hardware, magari software) per evitare incidenti. La spiegabilità della AI non può essere un dito dietro cui nascondersi per risparmiare denaro. Giacomo PS: aggiungo in copia la lista, perché la tua domanda è di nuovo chiara e pertinente (che è probabilmente il miglior complimento che si possa ricevere da un hacker! :-D).

1 0

Fwd: Publication du rapport Villani sur l'intelligence artificielle
by J.C. DE MARTIN March 29, 2018

March 29, 2018

Tra meno di un'ora a Parigi. jc -------- Forwarded Message -------- Subject: Publication du rapport Villani sur l'intelligence artificielle Resent-Date: Wed, 28 Mar 2018 18:45:58 +0200 Resent-From: d004620(a)polito.it Date: Wed, 28 Mar 2018 16:45:56 +0000 From: Mission Villani <missionvillani(a)cnnumerique.fr> Reply-To: equipe.missionvillani(a)cnnumerique.fr To: demartin(a)polito.it Publication du rapport Villani sur l'intelligence artificielle Voir la version en ligne <http://g5nx.mjt.lu/lnk/AMoAAALo5rwAAVGER5gAAGbvarsAARqUvlwAGpz2AAct2wBau8bE…> *DONNER UN SENS ÀL’INTELLIGENCE ARTIFICIELLE* Pour une stratégie nationale et européenne Le sommet #AIForHumanity se tiendra demain jeudi 29 mars à partir de 8h45 au Collège de France. À cette occasion, le député et mathématicien Cédric Villani présentera les conclusions de son rapport “Donner un sens à l’intelligence artificielle”. La matinée sera rythmée par des débats entre personnalités politiques et experts mondiaux du domaine. Le Président de la République, Emmanuel Macron, clôturera le sommet en annonçant la stratégie française en matière d’IA. Découvrez le rapport et suivez le live de l'évènement sur le site https://www.aiforhumanity.fr/ ACCÉDER AU SITE <http://g5nx.mjt.lu/lnk/AMoAAALo5rwAAVGER5gAAGbvarsAARqUvlwAGpz2AAct2wBau8bE…> Bien cordialement, L'équipe de la mission Contact presse *Candice Foehrenbach* Attachée parlementaire de Cédric Villani candice.foehrenbach(a)clb-an.fr *Yann Bonnet* Secrétaire général du Conseil national du numérique et membre de la mission presse(a)cnnumerique.fr | +33 (0)1 44 97 25 08

1 0