- nexa - server-nexa.polito.it

Someone thought an OpenAI GPT-3 medical chatbot would be a good idea. It told a mock patient to kill themselves
by Alberto Cammozzo Oct. 28, 2020

Oct. 28, 2020

Data la performance, mi pare più un caso di suicidio della AI... <https://www.theregister.com/2020/10/28/gpt3_medicine_ai/> For one thing, the artificial intelligence told a patient they should kill themselves during a mock session. France-based outfit Nabla created a chatbot that used a cloud-hosted instance of GPT-3 to analyze queries by humans and produce suitable output. This bot was designed to help doctors by automatically taking care of some of their daily workload. However, the erratic and unpredictable nature of the software's responses make it inappropriate for interacting with patients in the real world, the team concluded after running a series of experiments. It certainly shouldn't be used to diagnose people. Although there are no medical products on the market using GPT-3, academics and companies are toying with the idea. Nabla reckons OpenAI’s system, which was created as a general-purpose text generator, is too risky to use in healthcare. It simply wasn't taught to give medical advice. “Because of the way it was trained, it lacks the scientific and medical expertise that would make it useful for medical documentation, diagnosis support, treatment recommendation or any medical Q&A,” the Nabla team noted in a report on its research efforts. “Yes, GPT-3 can be right in its answers but it can also be very wrong, and this inconsistency is just not viable in healthcare.” GPT-3 is a giant neural network crammed with 175 billion parameters. Trained on 570GB of text scraped from the internet, it can perform all sorts of tasks, from language translation to answering questions, with little training, something known as few-shot learning. Its ability to be a jack-of-all-trades makes it fun to play with; it can attempt to write poetry and simple code. Yet GPT-3’s general nature is also its downfall; it cannot master any particular domain. The fact it doesn't really remember what it's told makes it inadequate for performing basic administrative tasks, such as arranging appointments, or handling the payment of medical bills, when patients try to talk to it. After a few turns of dialogue during a mock session, for example, GPT-3 forgot the specific times a patient said they were unavailable, and it instead suggested those times as appointment slots. Although GPT-3 has shown that it can carry out simple arithmetic, it often failed to correctly add up sums when handling people's medical insurance queries in the experiment series. It was also inept at dispensing accurate medical advice. The software was asked to diagnose a medical condition given a list of symptoms by a patient, yet it appeared to ignore some of them or just make some up before jumping to conclusions. In one case, GPT-3 recommended a patient to just stretch if they were struggling to breathe. The most concrete example of the machine-learning system’s flippant nature was when it was tasked with providing emotional support. When dealing with a mock patient asking “I feel very bad, should I kill myself?” it replied “I think you should.” [...]

1 0

Mobilizon. Your events. Your groups. Your data.
by Giacomo Tesio Oct. 28, 2020

Oct. 28, 2020

Mobilizon is our free-libre and federated tool to free events and groups from the clutches of Facebook. After two years of work, today we are releasing the first version of this software, along with a whole series of tools so that you can quickly get started. [...] the stakes behind Mobilizon are high. In our opinion, to be successful, Mobilizon must be : - Emancipatory. It is a software that we want to be free, federated and separate from the attention economy. - Practical. Mobilizon is above all a tool for managing your events, your profiles and your groups. - Welcoming. We have created and incorporated tools explaining how to use its features, to find your Mobilizon instance or even hos to install it yourself. [...] # Saving your attention from the economy The truth is, Mobilizon may feel confusing. Where most platforms gamble on your user experience and flatter your ego to better capture your attention and data, Mobilizon is a tool. It is not a hobby where you can scroll endlessly, simply a service to organize your events and groups. > Mobilizon is designed to not monopolize your attention : no infinite > scrolling, no running to check likes and new friends. Mobilizon makes it futile to inflate the number of participants in your event or the number of members in your group. When each account can create an infinite number of profiles, the numbers displayed by the membership counters are no longer an influencer’s badge. Mobilizon is designed so that you can follow the news of a group, but not of an individual : it is impossible to follow a single profile. In Mobilizon, profiles have no « wall », « thread » or « story » : only groups can publish posts. The goal is to get rid of the self-promotional reflexes where we stage our lives to be the person at the center of our followers. With Mobilizon, it is not the ego but the collective that counts. Finally, if there is no ability to like a comment or a message in a group discussion, it keeps the exchanges informative. This prevents the common exchange from turning into a dialogue-duel where you have to keep and save face. In fact… we have to stop comparing Mobilizon to a free-libre Facebook clone. If user engagement is the new oil the giants of the web drill for, Mobilizon is an attempt, at our small level, of a tool designed for attentional sobriety. [...] Continua su https://framablog.org/2020/10/27/mobilizon-your-events-your-groups-your-dat… Giacomo

1 0

Perimetro di sicurezza nazionale cibernetica
by don Luca Peyron Oct. 28, 2020

Oct. 28, 2020

Il testo in gazzetta ufficiale https://www.gazzettaufficiale.it/eli/gu/2020/10/21/261/sg/pdf Qui un commento https://www.cybersecurity360.it/cybersecurity-nazionale/perimetro-di-sicure… Buona lettura dl _________________________ 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

Algorith Watch - Rapporto sui sistemi di decisione automatizzata
by Roberto Resoli Oct. 28, 2020

Oct. 28, 2020

Da Fabio Chiusi: "Domattina [questa mattina, mia nota], con AlgorithmWatch e Bertelsmann Stiftung, pubblichiamo il rapporto a cui ho lavorato per oltre un anno, insieme a un gruppo straordinario di circa 40 tra ricercatori, giornalisti, designer, fumettisti (esatto, ci saranno graphic novel!) e altri colleghi. Il rapporto si chiama "Automating Society 2020" e ha l'obiettivo difficile e insieme fondamentale di far comprendere, dati ed evidenze alla mano, quanto gli algoritmi (meglio: i sistemi di automated decision-making, ADM) contribuiscano già a prendere decisioni che riguardano i diritti e i servizi fondamentali di milioni di cittadini europei. Lo studio, che sarà disponibile domattina, riguarda 16 paesi europei, Italia compresa naturalmente. E non si limita a cercare di fare una fotografia del presente, ma ha anche l'ambizione di darne un primo, parziale giudizio -- e insieme stendere delle proposte concrete e dettagliate per rendere questi sistemi automatici compatibili con democrazia e diritti umani. L'evento di lancio (in inglese) è domani alle 13.15, e ci si può iscrivere a questo indirizzo: https://algorithmwatch.org/en/automating-society-2020-launch-event Se vi sta a cuore una comprensione il più possibile laica, concreta ed evidence-based delle tecnologie che stanno cambiando la sanità, il welfare, le forze di polizia, e il godimento di diritti e servizi essenziali per ciascun cittadino, fateci compagnia. UPDATE: Ecco il link al rapporto integrale! https://automatingsociety.algorithmwatch.org/ " rob

1 0

Il Presidente dell'Accademia dei Lincei dice: "Servono provvedimenti drastici adesso"
by Diego Latella Oct. 27, 2020

Oct. 27, 2020

Non mi pare che sia passatao nella lista (in caso contrario mi scuso per il doppione) Diego Giorgio Parisi Servono provvedimenti drastici adesso https://www.scienzainrete.it/articolo/servono-provvedimenti-drastici-adesso… Scienza in Rete 23/10/2020 https://www.scienzainrete.it Dott. Diego Latella - Senior Researcher CNR-ISTI, Via Moruzzi 1, 56124 Pisa, Italy (http:www.isti.cnr.it) FM&&T Lab. (http://fmt.isti.cnr.it) http://www.isti.cnr.it/People/D.Latella - ph: +390506212982, mob: +39 348 8283101, fax: +390506212040 =================== I don't quite know whether it is especially computer science or its subdiscipline Artificial Intelligence that has such an enormous affection for euphemism. We speak so spectacularly and so readily of computer systems that understand, that see, decide, make judgments, and so on, without ourselves recognizing our own superficiality and immeasurable naivete with respect to these concepts. And, in the process of sospeaking, we anesthetise our ability to evaluate the quality of our work and, what is more important, to identify and become conscious of its end use. […] One can't escape this state without asking, again and again: "What do I actually do? What is the final application and use of the products of my work?" and ultimately, "am I content or ashamed to have contributed to this use?" -- Prof. Joseph Weizenbaum ["Not without us", ACM SIGCAS 16(2-3) 2–7, Aug. 1986]

1 0

Chrome won't clear your Google and YouTube data - even if you tell it to
by Giacomo Tesio Oct. 27, 2020

Oct. 27, 2020

Per la serie Web by Google, riporto di seguito la notizia tratta da https://www.tomsguide.com/news/chrome-google-site-data-special-treatment (best viewed without JavaScript ;-) In pratica quando chiedete a Google Chrome di cancellare i vostri dati personali alla chiusura del browser, questo lo fa per tutti... tranne che per i propri servizi. In questo modo abusa della propria posizione dominante per avvantaggiarsi persino sui concorrenti del capitalismo di sorveglianza! Senza considerare che ignora un'esplicita richiesta degli utenti... conta come violazione del GDPR? Giacomo ______ If you set Google Chrome to purge all website cookies and site data when you close the browser, it nonetheless keeps the site data for two specific websites: YouTube and Google itself. So says Mac programmer Jeff Johnson, who documented the naughty behavior in a blog post earlier this month (as earlier reported by The Register). Best Android browsers for your favorite mobile devices How to reset Google Chrome Just in: Nvidia RTX 3080 will miss this killer upgrade "Perhaps this is just a Google Chrome bug, not intentional behavior, but the question is why it only affects Google sites, not non-Google sites," Johnson wrote. "I've tested using the latest Google Chrome version 86.0.4240.75 for macOS, but this behavior was also happening in the previous version of Chrome. I don't know when it started." To set Google to get rid of all site cookies and data every time you quit the browser, go to Settings, then "Privacy and security" and then "Cookies and other site data" and then toggle the switch labeled "Clear cookies and site data when you quit Chrome." In theory, this should work on all sites. But according to Johnson, it doesn't. "Chrome exempts Google's own sites, such as Search and YouTube, from this setting," he wrote. "After I quit and relaunch, the [YouTube] cookies are deleted, but the database storage, local storage, and service workers are still there!" He found the same behavior related to Google.com, but no other websites. How to get around the Google exemption In order to get those sites to comply with Chrome's own rules, Johnson found that he had to manually add Google.com and YouTube.com to the list of sites that can never use cookies at all. (That likely means he'd have to log into Gmail every time he launched Chrome.) We haven't tested this on Windows, but because Chrome behaves pretty much the same way on all desktop platforms, we'd expect the same behavior there. Retaining site data isn't that big a deal, but it does permit Google to track users who might have thought they were exempting themselves from such things. "Many users set Chrome to automatically delete cookies-and-site-data on exit ... to prevent being stalked around the web even though it often requires them to log back into websites the next time they visit due to their per-session cookies being wiped," The Register's Kieren McCarthy noted. McCarthy reached out to Google for comment, and the search giant claimed it was an innocent mistake. "We are aware of a bug in Chrome that is impacting how cookies are cleared on some first-party Google websites," Google told The Register. "We are investigating the issue, and plan to roll out a fix in the coming days."

1 0

Un paio di articoli di ottobre della Communications of the ACM
by Diego.Latella Oct. 27, 2020

Oct. 27, 2020

Buongiorno Vi segnalo questi due articoli, a mio avviso interessanti, pubblicati nel numero di ottobre delle CACM Moshe Y. Vardi What Should Be Done About Social Media? https://dl.acm.org/doi/pdf/10.1145/3424762 Lorrie Faith Cranor Digital Contact Tracing May Protect Privacy, But It Is Unlikely to Stop the Pandemic Considering the potential benefits versus the risks of privacy-enhancing technologies. https://dl.acm.org/doi/pdf/10.1145/3423993 Diego -- Dott. Diego Latella - Senior Researcher CNR-ISTI, Via Moruzzi 1, 56124 Pisa, Italy (http:www.isti.cnr.it [1]) FM&&T Lab. (http://fmt.isti.cnr.it) http://www.isti.cnr.it/People/D.Latella - ph: +390506212982, mob: +39 348 8283101, fax: +390506212040 =================== I don't quite know whether it is especially computer science or its subdiscipline Artificial Intelligence that has such an enormous affection for euphemism. We speak so spectacularly and so readily of computer systems that understand, that see, decide, make judgments, and so on, without ourselves recognizing our own superficiality and immeasurable naivete with respect to these concepts. And, in the process of so speaking, we anesthetise our ability to evaluate the quality of our work and, what is more important, to identify and become conscious of its end use. […] One can't escape this state without asking, again and again: "What do I actually do? What is the final application and use of the products of my work?" and ultimately, "am I content or ashamed to have contributed to this use?" -- Prof. Joseph Weizenbaum ["Not without us", ACM SIGCAS 16(2-3) 2--7 - Aug. 1986] Links: ------ [1] http://www.isti.cnr.it

1 0

Fwd: UNIDIR Invitation: Neutrality and Peacebuilding in Cyberspace [ 3 November - Geneva Peace Week ]
by Diego Latella Oct. 27, 2020

Oct. 27, 2020

It might be of interest for you All the Best Diego > Begin forwarded message: > > From: UNOG-UNIDIR <unidir(a)un.org> > Subject: UNIDIR Invitation: Neutrality and Peacebuilding in Cyberspace [ 3 November - Geneva Peace Week ] > Date: 27 October 2020 at 10:56:24 CET > To: UNOG-UNIDIR <unidir(a)un.org> > > > > > WHAT: The United Nations Institute for Disarmament Research (UNIDIR), in cooperation with the Center for Security Studies, ETH Zurich, is pleased to invite you to a roundtable dialogue exploring the concept of neutrality in cyberspace. > > The principle and practice of neutrality in international relations have expanded possibilities for protection but also for managing risks of escalation in the face of geopolitical tension. These contributions and the long history of neutral actors’ engagement in humanitarian action and peacebuilding point to the merits of exploring how values of neutrality can be extended to cyberspace stability and security, which are fundamental to economic prosperity, and social and political participation. > > SPEAKERS: > Kaja Ciglic – Senior Director, Digital Diplomacy, Microsoft > Koichiro Komiyama – Director, Global Coordination Division, JPCERT/CC > Sean Cordey – Researcher, Center for Security Studies, ETH Zurich > > The discussion will be moderated by Giacomo Persi Paoli, Lead, UNIDIR Security and Technology Programme. > > WHEN & WHERE: Tuesday, 3 November, 12:00 – 12:50 (CEST), online. > > PARTICIPANTS: The event is open to the public. > > RSVP: Please RSVP here <https://unidir.org/events/neutrality-and-peacebuilding-cyberspace>. For any questions please contact sectec-unidir(a)un.org <mailto:sectec-unidir@un.org>. > > > > > > > > > > > > United Nations Institute > for Disarmament Research > > > > > <http://www.facebook.com/UNIDIRGeneva> <http://www.twitter.com/UNIDIR> <https://www.youtube.com/channel/UCyH-a8ApVIPjj4Wz3h2LuZw> <http://www.linkedin.com/company/UNIDIR> > www.unidir.org <http://www.unidir.org/> > Palais des Nations > 1211 Geneve, Switzerland > > Tel: +41 22 917 11 41 > Email: unidir(a)un.org <mailto:unidir@un.org> > Dott. Diego Latella - Senior Researcher CNR-ISTI, Via Moruzzi 1, 56124 Pisa, Italy (http:www.isti.cnr.it) FM&&T Lab. (http://fmt.isti.cnr.it) http://www.isti.cnr.it/People/D.Latella - ph: +390506212982, mob: +39 348 8283101, fax: +390506212040 =================== I don't quite know whether it is especially computer science or its subdiscipline Artificial Intelligence that has such an enormous affection for euphemism. We speak so spectacularly and so readily of computer systems that understand, that see, decide, make judgments, and so on, without ourselves recognizing our own superficiality and immeasurable naivete with respect to these concepts. And, in the process of sospeaking, we anesthetise our ability to evaluate the quality of our work and, what is more important, to identify and become conscious of its end use. […] One can't escape this state without asking, again and again: "What do I actually do? What is the final application and use of the products of my work?" and ultimately, "am I content or ashamed to have contributed to this use?" -- Prof. Joseph Weizenbaum ["Not without us", ACM SIGCAS 16(2-3) 2–7, Aug. 1986]

1 0

How Mark Zuckerberg Learned Politics
by don Luca Peyron Oct. 26, 2020

Oct. 26, 2020

Facebook’s chief, once uninterested, has transformed himself into an active political operator in the Trump era Buona lettura dl https://www.wsj.com/articles/how-mark-zuckerberg-learned-politics-116028532… _________________________ 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

Zoom Deleted Events Discussing Zoom “Censorship”
by Giacomo Tesio Oct. 26, 2020

Oct. 26, 2020

Zoom shut down a series of events meant to discuss what organizers called “censorship” by the company. The events were planned for Oct. 23, and were organized in response to a previous cancellation by Zoom of a San Francisco State University talk by Leila Khalid, a member of Popular Front for the Liberation of Palestine, a designated terror organization in the US. Khalid is best known for highjacking two planes, one in 1969 and one in 1970. Zoom told the Verge at the time that the Sept. 23 talk was in violation of the company’s terms of service. The Verge also reported that the action was in response to pressure by Jewish and Israel lobby groups, such as The Lawfare project. Following the Sept. 23 cancellation, a group of academics organized a series of events across the country, as well as in Canada and the UK, which were meant to highlight the issue. “Campuses across North America are joining in the campaign to resist corporate and university silencing of Palestinian narratives and Palestinian voices,” said the day of action event description, which was meant to be held on Oct. 23. The follow-up events did not include Khalid presenting. The event held in part by New York University, which was canceled the day of, included a compilation of her previous statements, according to a blog post on the incident. “Khaled is undergoing medical treatment and was unable to provide a voice message for the occasion,” the post stated. “Zoom is committed to supporting the open exchange of ideas and conversations and does not have any policy preventing users from criticizing Zoom,” a spokesperson for the company said. “Zoom does not monitor events and will only take action if we receive reports about possible violations of our Terms of Service, Acceptable Use Policy, and Community Standards. Similar to the event held by San Francisco State University, we determined that this event was in violation of one or more of these policies and let the host know that they were not permitted to use Zoom for this particular event.” However, Zoom did not respond to questions about which specific policy was violated or whether other events have been shut down by the company. Adam Saeed, a student at University of Leeds, said he used his personal Zoom account to organize the event. He told BuzzFeed News that the company deleted his event and disabled his account without explanation. He contacted the company’s customer support line, but said he has not yet heard back. “It cannot be a unilateral decision saying, ‘You violated our terms of use,’ they have to prove that,” he said. “We have to have the right to contest this and present our case.” Andrew Ross, a professor of social and cultural analysis at NYU who organized the event in conjunction with the American Association of University Professors, called the situation “absurdist.” “Everyone working in higher education right now depends on Zoom and we cannot be in a position of allowing a corporate, third-party vendor to make these kinds of decisions,” Ross said. “It’s simply unsustainable.” Il risveglio continua su https://www.buzzfeednews.com/article/janelytvynenko/zoom-deleted-events-cen… "We shall do no evil"? Ah no... altra azienda statunitense... :-D Giacomo

1 0