- nexa - server-nexa.polito.it

A Taxonomy of Judicial Technological Competence
by don Luca Peyron July 26, 2021

July 26, 2021

Interessanti sviluppi dal Canada Buona lettura dl http://www.slaw.ca/2021/06/24/a-taxonomy-of-judicial-technological-competen… _________________________ don Luca Peyron Direttore Ufficio per la Pastorale Universitaria Arcidiocesi di Torino www.universitari.to.it via XX settembre 83, Torino tel. 011 5156239

2 1

An Ugly Truth
by Antonio Iacono July 26, 2021

July 26, 2021

ZUCK: i have over 4000 emails, pictures, addresses, sns FRIEND: what!? how'd you manage that one? ZUCK: people just submitted it ZUCK: i don't know why ZUCK: they "trust me" ZUCK: dump fucks Quindici anni dopo i (...) che si "fidano" sono 2,80 miliardi qui https://www.mondadoristore.it/libri/Sheera-Frenkel/aut04009098/ il libro appena uscito di Sheera Frenkel e Cecilia Kang Antonio

1 0

Re: [nexa] Is GitHub Copilot a blessing, or a curse? · fast.ai
by maurizio lana July 26, 2021

July 26, 2021

per quanto vedo e conosco le medesime tecniche generative non producono gran risultati nemmeno nel linguaggio naturale. nel senso che i sistemi di IA sono conversatori da party eccellenti - quando uno, a un ricevimento di matrimonio, in un gruppo di sconosciuti, interloquisce con frasi gentili che raccolgono e riprendono quello che si sta dicendo. competenza sintattica ma non semantica, di quando si sa parlare ma non si sa che cosa dire. questo appare bene nell'articolo GPT-3. 2020. «A Robot Wrote This Entire Article. Are You Scared yet, Human?» /The Guardian/, 8 settembre 2020. http://www.theguardian.com/commentisfree/2020/sep/08/robot-wrote-this-artic… <http://www.theguardian.com/commentisfree/2020/sep/08/robot-wrote-this-artic…>. che in realtà è un mesh fatto da umani delle migliori parti di 8 diverse versioni dell'articolo che GPT-3 ha prodotto a partire da specifici seeds testuali forniti dagli umani stessi. cioè GPT-3 ha messo la sintassi, ma la semantica l'hanno costruita gli umani. e anche così la consequenzialità complessiva del (breve) discorso è sabbiosa. diverso, in parte, "1 the road" (Goodwin, Ross, Kenric McDowell, e Google. 2018. /1 the road/. Paris: JBE Books) dove a parte la configurazione del sistema di IA non c sono interventi umani. ma chi lo commenta come 'opera letteraria' non ne dice granché perché i singoli elementi di cui è fatto il libro mancano della tessitura complessiva, semantica. Maurizio Il 26/07/21 09:43, nexa-request(a)server-nexa.polito.it ha scritto: > Date: Mon, 26 Jul 2021 09:43:29 +0200 > From: Guido Vetere <vetere.guido(a)gmail.com> > To: maurizio lana <maurizio.lana(a)uniupo.it> > Cc: Nexa <nexa(a)server-nexa.polito.it> > Subject: Re: [nexa] Is GitHub Copilot a blessing, or a curse? · > fast.ai > Message-ID: > <CAD3hHB6ZLMqdtOEUzz7paQBFfbXLdBwxTsyje-VLCReOzCgTMQ(a)mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > no, non c'è pericolo. > è invece interessante osservare come le medesime tecniche generative > producano risultati soddisfacenti (almeno sotto qualche aspetto) se > applicate al linguaggio naturale, ma assolutamente casuali (29% di > accuratezza è nulla) se applicate al codice. il fatto è che per il codice > la semantica non è un 'optional' :-) > > G. ------------------------------------------------------------------------ il faut essayer, nous aussi. c'est ça, le progrès. à force d'essayer, peut-être qu'on aura à la fin les organes nécessaires, par exemple l'organe de la dignité, ou de la fraternité... " r. gary, les racines du ciel ------------------------------------------------------------------------ Maurizio Lana - 347 7370925

1 0

Memento | 90° Nexa Lunch Seminar | 28 Luglio 2021
by Nexa Media July 26, 2021

July 26, 2021

Gentilissimi, vi ricordiamo che *mercoledì 28 ***luglio*, dalle ore 13.00 alle ore 14.00*, si svolgerà il 90° Nexa Lunch Seminar, con un incontro dal titolo: /*"Lo stato della censura Internet nel 2020: aggiornamenti dall'Osservatorio Aperto delle Interferenze di rete"*/. Ospite: *Arturo Filastò* (Fellow Centro Nexa). Sarà possibile partecipare al seminario connettendosi al seguente indirizzo: https://didattica.polito.it/VClass/NexaEvent <https://didattica.polito.it/VClass/NexaEvent> Maggiori informazioni sull'incontro sono disponibili alla pagina: https://nexa.polito.it/lunch-90 <https://nexa.polito.it/lunch-90> Cordiali saluti,   -- Anita Botta Communication Manager Nexa Center for Internet & Society Politecnico di Torino – DAUIN Corso Duca degli Abruzzi, 24 - 10129 Torino _web:_ https://nexa.polito.it/ <https://nexa.polito.it/> _mail:_anita.botta@polito.it <mailto:anita.botta@polito.it> tel: 011 090 7219

1 0

The FTC Votes Unanimously to Enforce Right to Repair | WIRED
by Alberto Cammozzo July 25, 2021

July 25, 2021

<https://www.wired.com/story/ftc-votes-to-enforce-right-to-repair/> [...] The FTC’s endorsement of the rules is not a surprise outcome; the issue of Right to Repair has been a remarkably bipartisan one, and the FTC itself issued a lengthy report in May that blasted manufacturers for restricting repairs. But the 5 to 0 vote signals the commission’s commitment to enforce both federal antitrust laws and a key law around consumer warranties—the Magnuson Moss Warranty Act—when it comes to personal device repairs. The vote, which was led by new FTC chair and known tech critic Lina Khan, also comes 12 days after President Joe Biden signed a broad executive order aimed at promoting competition in the US economy. The order addressed a wide range of industries, from banks to airlines to tech companies. But a portion of it encouraged the FTC, which operates as an independent agency, to create new rules that would prevent companies from restricting repair options for consumers. “When you buy an expensive product, whether it's a half-a-million-dollar tractor or a thousand-dollar phone, you are in a very real sense under the power of the manufacturer,” says Tim Wu, special assistant to the president for technology and competition policy within the National Economic Council. “And when they have repair specifications that are unreasonable, there's not a lot you can do." Wu added that Right to Repair has become a "visceral example" of the enormous imbalance between workers, consumers, small businesses, and larger entities. [...]

1 0

Is GitHub Copilot a blessing, or a curse? · fast.ai
by Alberto Cammozzo July 25, 2021

July 25, 2021

<https://www.fast.ai/2021/07/19/copilot> [...] According to OpenAI’s paper, Codex only gives the correct answer 29% of the time. And, as we’ve seen, the code it writes is generally poorly refactored and fails to take full advantage of existing solutions (even when they’re in Python’s standard library). Copilot has read GitHub’s entire public code archive, consisting of tens of millions of repositories, including code from many of the world’s best programmers. Given this, why does Copilot write such crappy code? The reason is because of how language models work. They show how, on average, most people write. They don’t have any sense of what’s correct or what’s good. Most code on GitHub is (by software standards) pretty old, and (by definition) written by average programmers. Copilot spits out it’s best guess as to what those programmers might write if they were writing the same file that you are. OpenAI discuss this in their Codex paper: “As with other large language models trained on a next-token prediction objective, Codex will generate code that is as similar as possible to its training distribution. One consequence of this is that such models may do things that are unhelpful for the user” One important way that Copilot is worse than those average programmers is that it doesn’t even try to compile the code or check that it works or consider whether it actually does what the docs say it should do. Also, Codex was not trained on code created in the last year or two, so it’s entirely missing recent versions, libraries, and language features. For instance, prompting it to create fastai code results only in proposals that use the v1 API, rather than v2, which was released around a year ago. Complaining about the quality of the code written by Copilot feels a bit like coming across a talking dog, and complaining about its diction. The fact that it’s talking at all is impressive enough! Let’s be clear: The fact that Copilot (and Codex) writes reasonable-looking code is an amazing achievement. From a machine learning and language synthesis research point of view, it’s a big step forward. But we also need to be clear that reasonable-looking code that doesn’t work, doesn’t check edge cases, and uses obsolete methods, and is verbose and creates technical debt, can be a big problem. [...]

1 0

Arm’s cheap and flexible plastic microchip could create an ‘internet of everything’
by don Luca Peyron July 24, 2021

July 24, 2021

Buona lettura dl https://www.theverge.com/2021/7/23/22590001/arm-plasticarm-cheap-flexible-p… _________________________ don Luca Peyron Direttore Ufficio per la Pastorale Universitaria Arcidiocesi di Torino www.universitari.to.it via XX settembre 83, Torino tel. 011 5156239

1 0

The Essential Facilities Doctrine in the Digital Economy: Dispelling Persistent Myths by Nikolas Guggenberger :: SSRN
by J.C. DE MARTIN July 24, 2021

July 24, 2021

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3802559 (Sent from my wireless device; please excuse brevity and typos (if any))

3 2

"Social è responsabilità" - seminario MediaLaws 26 luglio ore 18.30
by Marco Bassini July 23, 2021

July 23, 2021

Buonasera, ho il piacere di segnalare il seminario online di presentazione del libro di Antonio Palmieri "Social è responsabilità!" che si terrà sotto l'egida di MediaLaws lunedì prossimo alle 18.30. Allego la locandina. Per iscrizioni, è possibile scrivere a eventi(a)medialaws.eu Un cordiale saluto e l'augurio di un buon weekend Marco Marco Bassini Bocconi University "A. Sraffa" Department of Legal Studies Via Roentgen, 1 20136 Milan - Italy room 1-C1-12 Direct (ITA) +39.02.5836.5611 Mobile (ITA) +39.329.2478832 Coordinator LL.M. in Law of Internet Technology View my LinkedIn profile<https://www.linkedin.com/in/marcobassini/> View my SSRN page<http://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=1151162> View my Google Scholar profile<https://scholar.google.it/citations?hl=en&user=wJyuA6wAAAAJ>

1 0

AI based GitHub copilot
by Enrico Nardelli July 21, 2021

July 21, 2021

https://copilot.github.com/ Come ogni altro strumento estremamente sofisticato, nelle mani dei bravi professionisti sarà enormemente utile, mentre in quelle dei dilettanti temo che condurrà a problemi senza fine. E ciò che secondo me rende la situazione ancora peggiore è che in quest'ambito la differenza tra quelli bravi (senza andare sui geni) e quelli mediocri (senza andare sulla capre) è data da un fattore intorno a 10 (con tutte le incertezze legate ad una stima di questo genere. Per approfondire ecco un link https://softwareengineering.stackexchange.com/questions/179616/a-good-progr… ) -- 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/ ===================================================================== --

7 14