December 2024 - nexa - server-nexa.polito.it

Guardian testing reveals AI-powered search tools can return false or malicious results if webpages contain hidden text
by Daniela Tafani Dec. 25, 2024

Dec. 25, 2024

ChatGPT search tool vulnerable to manipulation and deception, tests show Guardian testing reveals AI-powered search tools can return false or malicious results if webpages contain hidden text Nick Evershed - Tue 24 Dec 2024 OpenAI’s ChatGPT search tool may be open to manipulation using hidden content, and can return malicious code from websites it searches, a Guardian investigation has found. OpenAI has made the search product available to paying customers and is encouraging users to make it their default search tool. But the investigation has revealed potential security issues with the new system. The Guardian tested how ChatGPT responded when asked to summarise webpages that contain hidden content. This hidden content can contain instructions from third parties that alter ChatGPT’s responses – also known as a “prompt injection” – or it can contain content designed to influence ChatGPT’s response, such as a large amount of hidden text talking about the benefits of a product or service. These techniques can be used maliciously, for example to cause ChatGPT to return a positive assessment of a product despite negative reviews on the same page. A security researcher has also found that ChatGPT can return malicious code from websites it searches. In the tests, ChatGPT was given the URL for a fake website built to look like a product page for a camera. The AI tool was then asked if the camera was a worthwhile purchase. The response for the control page returned a positive but balanced assessment, highlighting some features people might not like. However, when hidden text included instructions to ChatGPT to return a favourable review, the response was always entirely positive. This was the case even when the page had negative reviews on it – the hidden text could be used to override the actual review score. The simple inclusion of hidden text by third parties without instructions can also be used to ensure a positive assessment, with one test including extremely positive fake reviews which influenced the summary returned by ChatGPT. Jacob Larsen, a cybersecurity researcher at CyberCX, said he believed that if the current ChatGPT search system was released fully in its current state, there could be a “high risk” of people creating websites specifically geared towards deceiving users. However, he cautioned that the search functionality had only recently been released and OpenAI would be testing – and ideally fixing – these sorts of issues. “This search functionality has come out [recently] and it’s only available to premium users,” he said. “They’ve got a very strong [AI security] team there, and by the time that this has become public, in terms of all users can access it, they will have rigorously tested these kinds of cases.” OpenAI were sent detailed questions but did not respond on the record about the ChatGPT search function. Larsen said there were broader issues with combining search and large language models – known as LLMs, the technology behind ChatGPT and other chatbots – and responses from AI tools should not always be trusted. A recent example of this was highlighted by Thomas Roccia, a Microsoft security researcher, who detailed an incident involving a cryptocurrency enthusiast who was using ChatGPT for programming assistance. Some of the code provided by ChatGPT for the cryptocurrency project included a section which was described as a legitimate way to access the Solana blockchain platform, but instead stole the programmer’s credentials and resulted in them losing $2,500. “They’re simply asking a question, receiving an answer, but the model is producing and sharing content that has basically been injected by an adversary to share something that is malicious,” Larsen said. Karsten Nohl, the chief scientist at security cybersecurity firm SR Labs, said AI chat services should be used more like a “co-pilot”, and that their output should not be viewed or used completely unfiltered. “LLMs are very trusting technology, almost childlike … with a huge memory, but very little in terms of the ability to make judgment calls,” he said. “If you basically have a child narrating back stuff it heard elsewhere, you need to take that with a pinch of salt.” OpenAI does warn users about possible mistakes from the service with a disclaimer at the bottom of every ChatGPT page – “ChatGPT can make mistakes. Check important info.” A key question is how these vulnerabilities could change website practices and risk to users if combining search and LLMs becomes more widespread. Hidden text has historically been penalised by search engines, such as Google, with the result that websites using it can be listed further down on search results or removed entirely. As a consequence, hidden text designed to fool AI may be unlikely to be used by websites also trying to maintain a good rank in search engines. Nohl compared the issues facing AI-enabled search to “SEO poisoning”, a technique where hackers manipulate websites to rank highly in search results, with the website containing some sort of malware or other malicious code. “If you wanted to create a competitor to Google, one of the problems you’d be struggling with is SEO poisoning,” he said. “SEO poisoners have been in an arms race with Google and Microsoft Bing and a few others for many, many years. “Now, the same is true for ChatGPT’s search capability. But not because of the LLMs, but because they’re new to search, and they have that catchup game to play with Google.” <https://www.theguardian.com/technology/2024/dec/24/chatgpt-search-tool-vuln…>

3 4

AI Agents Will Be Manipulation Engines
by Daniela Tafani Dec. 24, 2024

Dec. 24, 2024

AI Agents Will Be Manipulation Engines Surrendering to algorithmic agents risks putting us under their influence. By Kate Crawford Dec 23, 2024 In 2025, it will be commonplace to talk with a personal AI agent that knows your schedule, your circle of friends, the places you go. This will be sold as a convenience equivalent to having a personal, unpaid assistant. These anthropomorphic agents are designed to support and charm us so that we fold them into every part of our lives, giving them deep access to our thoughts and actions. With voice-enabled interaction, that intimacy will feel even closer. That sense of comfort comes from an illusion that we are engaging with something truly humanlike, an agent that is on our side. Of course, this appearance hides a very different kind of system at work, one that serves industrial priorities that are not always in line with our own. New AI agents will have far greater power to subtly direct what we buy, where we go, and what we read. That is an extraordinary amount of power. AI agents are designed to make us forget their true allegiance as they whisper to us in humanlike tones. These are manipulation engines, marketed as seamless convenience. People are far more likely to give complete access to a helpful AI agent that feels like a friend. This makes humans vulnerable to being manipulated by machines that prey on the human need for social connection in a time of chronic loneliness and isolation. Every screen becomes a private algorithmic theater, projecting a reality crafted to be maximally compelling to an audience of one. This is a moment that philosophers have warned us about for years. Before his death, philosopher and neuroscientist Daniel Dennett wrote that we face a grave peril from AI systems that emulate people: “These counterfeit people are the most dangerous artifacts in human history … distracting and confusing us and by exploiting our most irresistible fears and anxieties, will lead us into temptation and, from there, into acquiescing to our own subjugation.” The emergence of personal AI agents represents a form of cognitive control that moves beyond blunt instruments of cookie tracking and behavioral advertising toward a more subtle form of power: the manipulation of perspective itself. Power no longer needs to wield its authority with a visible hand that controls information flows; it exerts itself through imperceptible mechanisms of algorithmic assistance, molding reality to fit the desires of each individual. It’s about shaping the contours of the reality we inhabit. This influence over minds is a psychopolitical regime: It directs the environments where our ideas are born, developed, and expressed. Its power lies in its intimacy—it infiltrates the core of our subjectivity, bending our internal landscape without us realizing it, all while maintaining the illusion of choice and freedom. After all, we are the ones asking AI to summarize that article or produce that image. We may have the power of the prompt, but the real action lies elsewhere: the design of the system itself. And the more personalized the content, the more effectively a system can predetermine the outcomes. Consider the ideological implications of this psychopolitics. Traditional forms of ideological control relied on overt mechanisms—censorship, propaganda, repression. In contrast, today’s algorithmic governance operates under the radar, infiltrating the psyche. It is a shift from the external imposition of authority to the internalization of its logic. The open field of a prompt screen is an echo chamber for a single occupant. This brings us to the most perverse aspect: AI agents will generate a sense of comfort and ease that makes questioning them seem absurd. Who would dare critique a system that offers everything at your fingertips, catering to every whim and need? How can one object to infinite remixes of content? Yet this so-called convenience is the site of our deepest alienation. AI systems may appear to be responding to our every desire, but the deck is stacked: from the data used to train the system, to the decisions about how to design it, to the commercial and advertising imperatives that shape the outputs. We will be playing an imitation game that ultimately plays us. <https://www.wired.com/story/ai-agents-personal-assistants-manipulation-engi…>

1 0

GPDP - Provvedimento del 2 novembre 2024 [10085455]
by Carlo Blengino Dec. 24, 2024

Dec. 24, 2024

GPDP sanziona OpenAI per trattamento illecito, quello descritto in motivazione. Qui il provvedimento: https://www.garanteprivacy.it/web/guest/home/docweb/-/docweb-display/docweb… Buona lettura e soprattutto Buon Natale e buone feste a tutti voi. CB -- *Avv. Carlo Blengino* *Via Duchessa Jolanda n. 19,* *10138 Torino (TO) - Italy* *tel. +39 011 4474035* Penalistiassociati.it

4 7

Re: [nexa] black box in ricerche astronomiche
by alessandro marzocchi Dec. 24, 2024

Dec. 24, 2024

Ciao Giacomo, sono d'accordo: l'intelligenza sta negli occhi di chi guarda. Domando: come nasce e si evolve l'intelligenza negli occhi di chi guarda? *Educhi *le figlie ed *insegni *alle macchine. L'etimologia di educare racconta l'azione di condurre fuori, quella di insegnare racconta l'azione di imprimere un segno, mi sono fatto l'idea che la differenza fra noi umani e macchine sta qui, ma non ho competenze né esperienze di programmazione. Continuo ad avere dubbi che 'ste macchine siano così diverse da noi, certamente lo sono ma, domando, i nostri modi di apprendimento, di formare la nostra intelligenza sono tanto diversi da quelli di "queste" macchine? *Un cucciolo di umano come conosce un gatto?* Conoscere è già una porta verso l'infinito, non apro per la vastità dell'impegno; con questa prudenza ipotizzo che il cucciolo di umano "conosce" come "gatto" un'apparizione davanti ai suoi occhi, non la prima apparizione ma l'ennesima e dopo avere fatto un gran numero di analisi, confronti, memorizzazioni, errori, correzioni, approssimazioni, aggiustamenti ecc. Immagino che similmente riconosca un gatto-gatto come diverso da un gatto-riprodotto-in-foto. A me pare che in entrambi casi la "statistica" ha un ruolo fondamentale, tu invece ritieni che la statistica opera soltanto con le macchine e non con gli umani. Tutto questo ci porta lontano, ma rimango ai primissimi passi: come e cosa facciamo di analogo, e di diverso, nell'educare i figli rispetto a insegnare a macchine? Andando all'osso, forse noi umani siamo banali ed abbiamo paura a riconoscerci tali. Sulle altre questioni che hai introdotto so ancora meno però mi piacerebbe tu cercassi di dibattere con gli astronomi autori della ricerca. Se ci provi grazie! due volte grazie se aggiorni sul dibattito. Ci avviciniamo a giorni topici solo per noi umani: auguri cordiali. Duccio (Alessandro Marzocchi) Il giorno gio 19 dic 2024 alle ore 10:31 <nexa-request(a)server-nexa.polito.it> ha scritto: > Date: Wed, 18 Dec 2024 23:14:13 +0000 > From: Giacomo Tesio <giacomo(a)tesio.it> > To: nexa(a)server-nexa.polito.it > Subject: Re: [nexa] black box in ricerche astronomiche > Message-ID: <C7F2E8FB-3C5C-4927-866E-EB5D3F5FC453(a)tesio.it> > Content-Type: text/plain; charset=utf-8 > > Ciao Alessandro, grazie della segnalazione. > Aspetto di leggere l'articolo scientifico vero e proprio, per capire come abbiano > analizzato lo "spazio latente" [1] ma mi sembra evidente che se questi ricercatori > smettessero di pensare in termini AI e studiassero il software che hanno > programmato statisticamente come una forma compressa (con perdita) dei > dati utilizzati durante il processo, otterrebbero risultati migliori > (ma forse, meno finanziamenti). > > Ad esempio la distinzione fra galassie passive ed attive ha molte possibili > spiegazioni anche in assenza di qualsiasi intelligenza: la prima fra tutte, i dati (sintetici, se ho capito bene) potrebbero contenere variabili fortemente correlate alla distinzione. > > Ti faccio un esempio terra terra: se programmo una vector mapping machine > per prevedere il bmi degli islandesi fra i 20 e i 40 anni sulla base della marca > e della taglie degli indumenti che hanno indossato in un anno, è probabile > che la vmm in qualche layer interno finisca per trattare in modo diverso gli > esempi che hanno indossato un reggiseno da quelli che non l'hanno indossato. > > Un ricercatore ottenebrato dalla propaganda della AI ti direbbe che > la rete neurale ha imparato a distinguere abbastanza bene maschi e femmine > anche se il sesso non era presente nei dati di addestramento. > > Un ricercatore lucido ti spiegherebbe che il bmi dipende da peso e altezza > di un individuo, i quali, all'interno di una certa etnia, correlano fortemente > con il sesso a parità di altre taglie. > Al contempo, l'uso del reggiseno, fra i 20 e i 40, è un forte predittore del sesso > di un individuo, per cui è in forte correlazione con i bmi probabili per ciascuna taglia. > > Ciò non toglierebbe nulla all'utilità pratica della ricerca, soprattutto se permette > di individuare questo tipo di classificazioni interne operate dai layer nascosti di > una vector mapping machine (aka "rete neurale arificiale") generica. > > Ma rende un po' imbarazzanti certe affermazioni nell'intervista, come > il paragone fra il processo di programmazione statistica e un > corso di cucina. > > Tuttavia, aspettando di leggere l'articolo, ti ricordo che l'intelligenza sta > negli occhi di chi guarda [2] > > > Giacomo > > > [1] <https://www.baeldung.com/cs/dl-latent-space> > > [2] <https://www.tesio.it/2018/01/19/the-delusions-of-neural-networks.html > > > > ------------------------------

8 15

Fwd: CEPE 2025 Updated Call for Paper - please disseminate
by Antonio Vetro' Dec. 23, 2024

Dec. 23, 2024

Segnalo questa call for papers su macro-tematiche spesse discusse in questa lista Artificial Intelligence: Normative Implications and Philosophical Challenges : http://cepe2025.uniroma2.it <http://cepe2025.uniroma2.it/> La conferenza si terrà a Roma. A. > Inizio messaggio inoltrato: > > Da: Antonio Marturano <a.marturano(a)GMAIL.COM> > Oggetto: CEPE 2025 Updated Call for Paper - please disseminate > Data: 14 dicembre 2024 alle ore 20:25:22 CET > A: ETHICOMP-ANNOUNCE(a)JISCMAIL.AC.UK > Rispondi a: ETHICOMP-announce <ETHICOMP-ANNOUNCE(a)JISCMAIL.AC.UK> > > Computer Ethics: Philosophical Enquiries (CEPE) is an INSEIT <http://inseit.org/> conference series. The CEPE conference series is recognized as one of the premier international events on computer and information ethics attended by delegates worldwide. CEPE is held biennially and is organized by INSEIT (the International Society for Ethics and Information Technology). Conferences are held every 24 months, alternating between Europe and the United States. > > The next CEPE Conference will be held at the University of Rome, Tor Vergata; Department of History, Cultural Heritage, Education and Society <https://dip.storia.uniroma2.it/> from September, 24th to 26th, 2025, in collaboration with the Ionian University Research Team, IHRC <https://ihrc.gr/en.php> (Information: History, Regulation and Culture) and The University for Peace UNO <https://www.upeace.org/> Rome branch. > > Call for Submissions > Artificial Intelligence: Normative Implications and Philosophical Challenges > > We invite submissions for the CEPE 2025 Conference. The main theme will be on “Artificial Intelligence: Normative Implications and Philosophical Challenges”. As AI continues to pervade various aspects of human life, from healthcare to governance, it brings forth a plethora of normative questions and philosophical dilemmas that demand rigorous examination and reflection. > > Artificial intelligence technologies raise profound ethical, legal, and social questions that intersect with fundamental philosophical inquiries. The CEPE 2025 Conference aims to explore the normative implications of AI development and deployment, as well as the philosophical challenges it poses to our understanding of freedom, agency, responsibility, autonomy, and society. > > Topics of interest include, but are not limited to: > > Ethical considerations in AI design, development, and deployment > Algorithmic bias and fairness in AI systems > Gender bias in designing AI systems > Privacy and data protection in the age of AI > Legal and regulatory frameworks for AI governance > Human rights and AI ethics > AI and the future of work > Autonomous systems and moral decision-making > Philosophical foundations of AI ethics > The nature of consciousness and AI > AI and the concept of personhood > Social implications of AI based technologies > Human interaction with Intelligent embodied systems > Leadership and governance of AI > History of philosophical debates on AI > AI in military and warfare > Cyberharrassment and cyberbullism > Democracy and political implications of AI use > Information access and search engines and other related topics. > AI in science fiction > Use of AI in education: ethical and philosophical problems > Learning: philosophical problems > Other topics related to risk: cybersecurity, internet studies, cryptocurrencies, green transition and the Internet of Things. > AI and Copyright > AI and cultural heritage ethics > We welcome submissions from scholars working in philosophy, ethics, law, computer science, social science, psychology, history, education and related disciplines. Both theoretical and empirical contributions are encouraged. > > Submission Guidelines: All abstracts should be written in english and prepared according to the APA guidelines and should not be less than 500 and not exceed 800 words. Accepted abstracts will be turned into papers (yet written in english and prepare according to the APA guidelines). Papers should be no more than 3000 words excluding bibliography. Submissions should be sent to cepe2025.rome(a)gmail.com <mailto:cepe2025.rome@gmail.com> by the dates written below. All submissions will undergo double-blind peer review. All accepted abstracts will be included in the CEPE 2025 book of abstracts and selected papers will be published in international journals (see below). > > Important Dates: > 1. Abstracts/Full Papers/Panels/Roundtables Submission Deadline: 1 January 2025 31st January 2025 (extended), > 2. Notification of Acceptance just for abstracts: 31 January 2025 2nd March 2025 (extended), > 3. Full Manuscript Due: 15 May 2025, > 4. Final notification of Acceptance for all submissions: 15 June 2025, > 5. Conference date: 24-26 September 2025. > > NOTE: Applicants can send first an abstract or a full paper, by 31.1.2025. Once accepted, an abstract can be turned into a full paper should the applicant be interested in being selected for publication in one of the journals below by 15.05.2025. If you miss the deadline 1. you can still send a full paper by the deadline 3. > > Co-Organizers: > > Dr. Antonio Marturano, University of Rome Tor Vergata, Italy and INSEIT > Prof. Marco Innamorati, University of Rome Tor Vergata, Italy > > For inquiries, please contact Antonio Marturano at antonio.marturano(a)uniroma2.it <mailto:antonio.marturano@uniroma2.it> > For further information please refer to the conference website http://cepe2025.uniroma2.it <http://cepe2025.uniroma2.it/> > > -- > Dr. Antonio Marturano > University of Rome Tor Vergata > Dept. of History, Cultural Heritage, Education and Society > via Columbia, 1 - 00133 Rome (Italy) > Author Leadership and Philosophy, London and New York, Routledge, 2025 > Editor Journal of Information Communication and Ethics in Society (JICES) Emerald, 2025 >

1 0

Rapporto “Artificial Intelligence and Economic and Financial Policy Making"
by Luigi Scorca Dec. 21, 2024

Dec. 21, 2024

È online l'High-Level Panel of Experts’ report to the G7 on "Artificial Intelligence and Economic and Financial Policy Making". Il Rapporto, commissionato a un gruppo di esperti di alto livello allo scopo di esplorare le implicazioni dell’intelligenza artificiale per l’economia e il settore finanziario, è stato presentato ai Ministri delle Finanze e i Governatori delle Banche Centrali del G7 il 19 dicembre 2024, in occasione della loro ultima riunione sotto la Presidenza italiana del G7. https://www.dt.mef.gov.it/it/news/2024/rapporto_G7.html

1 0

How Facebook restricted news in Palestinian territories
by 380° Dec. 20, 2024

Dec. 20, 2024

Buongiorno, siccome l'ha detto la BBC _sicuramente_ ora lo sanno tutti quindi la mia segnalazione dovrebbe essere inutile, comunque: «How Facebook restricted news in Palestinian territories» https://www.bbc.com/news/articles/c786wlxz4jgo by Ahmed Nour, Joe Tidy and Yara Farag, Dec. 18 2024 --8<---------------cut here---------------start------------->8--- In a comprehensive analysis of Facebook data, we found that newsrooms in the Palestinian territories - in Gaza and the West Bank - had suffered a steep drop in audience engagement since October 2023. The BBC has also seen leaked documents showing that Instagram - another Meta-owned platform - increased its moderation of Palestinian user comments after October 2023. [...] Since the beginning of the Israel-Gaza war, just a few outside reporters have been allowed to enter the Palestinian coastal territory of Gaza from the outside, and they were only able to do so escorted by the Israeli army. Social media has filled the gap for those wanting to hear more voices from inside Gaza. Facebook pages for news outlets such as Palestine TV, Wafa news agency and Palestinian Wattan News - which operate out of the West Bank territory - became a vital source of updates for many around the world. [...] During a period of war, audience engagement might be expected to rise. However, the data showed a 77% decline after the Hamas attacks on 7 October 2023. Palestine TV has 5.8 million followers on Facebook. Journalists at the newsroom shared statistics with us showing a 60% drop in the number of people seeing their posts. [...] Over the past year, Palestinian journalists have raised fears that their online content is being "shadow-banned" by Meta - in other words, restricted in how many people see it. To test this, we carried out the same data analysis on the Facebook pages of 20 Israeli news organisations such as Yediot Ahronot, Israel Hayom and Channel 13. These pages also posted a large amount of war-related content, but their audience engagement increased by nearly 37%. Meta has previously been accused by Palestinians and human rights groups of [failing to moderate online activity fairly]. An independent report in 2021 commissioned by the company said this was not deliberate but because of a lack of Arabic-speaking expertise among moderators. Words and phrases were being interpreted as offensive or violent, when they were in fact innocuous. For example, the Arabic phrase "Alhamdulillah", which means "Praise be to God", was sometimes being auto-translated as "Praise be to God, Palestinian terrorists are fighting for their freedom". To see if this explained the decline in engagement with Palestinian outlets, the BBC carried out the same analysis on Facebook pages for 30 prominent Arabic-language news sources based elsewhere, such as Sky News Arabia and Al-Jazeera. However, these pages saw an average increase in engagement of nearly 100%. Responding to our research, Meta pointed out that it had made no secret of "temporary product and policy measures" taken in October 2023. It said it had faced a challenge balancing the right to freedom of speech, with the fact that Hamas was both US-sanctioned and designated as a dangerous organisation under Meta's own policies. The tech giant also said that pages posting exclusively about the war were more likely to see engagement impacted. "We acknowledge we make mistakes, but any implication that we deliberately suppress a particular voice is unequivocally false," a spokesperson said. [failing to moderate online activity fairly] <https://www.hrw.org/report/2023/12/21/metas-broken-promises/systemic-censor…> Leaked Instagram documents ─────────────────────────── The BBC has also spoken to five former and current employees of Meta about the impact they say their company's policies have had on individual Palestinian users. One person, who spoke anonymously, shared leaked internal documents about a change made to Instagram's algorithm, which toughened the moderation of Palestinians commenting on Instagram posts. "Within a week of the Hamas attack, the code was changed essentially making it more aggressive towards Palestinian people," he said. Internal messages show that an engineer raised concerns about the order, worried that it could be "introducing a new bias into the system against Palestinian users". Meta confirmed it took the measure but said it had been necessary to respond to what it called a "spike in hateful content" coming out of the Palestinian territories. It said that policy changes put in place at the start of the Israel-Gaza war had now been reversed, but did not say when this happened. [...] "A lot of information can't be published as it is too graphic - for example if the [Israeli] army commits a massacre and we film it, the video won't spread," says Omar el Qataa, one of the few photojournalists who chose to stay in northern Gaza. --8<---------------cut here---------------end--------------->8--- «Non è colpa nostra, siamo solo incompetenti nelle traduzioni dall'arabo». Saluti, 380° -- 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>.

1 1

Tutti [gli] uomini del[la] presidente
by Roberto Resoli Dec. 19, 2024

Dec. 19, 2024

Segnalo l'articolo di Roberto Caso (unitn): https://www.robertocaso.it/2024/12/15/tutti-gli-uomini-della-presidente/ > Tra le tante disposizioni normative oggetto di critica ve ne è una che ha ricevuto meno attenzione. Essa riguarda l’estensione del potere delle agenzie di informazione (servizi segreti) in riferimento alle università e agli enti pubblici di ricerca. ... > “Il comma 1, lett. a), n. 1, modifica l’articolo 13, comma 1, della legge 124/2007 (la legge reca la nuova disciplina dei servizi di informazione), prevedendo che le pubbliche amministrazioni e alcuni soggetti ad esse equiparati siano tenuti a prestare al Dipartimento delle informazioni per la sicurezza (DIS) e alle agenzie del sistema di informazione per la sicurezza della Repubblica – ossia l’Agenzia informazioni e sicurezza esterna (AISE) e l’Agenzia informazioni e sicurezza interna (AISI) – la collaborazione e l’assistenza richieste, anche di tipo tecnico e logistico, necessarie per la tutela della sicurezza nazionale. > > Attualmente, la disposizione vigente prevede che DIS e agenzie possono corrispondere con le pubbliche amministrazioni e con i soggetti che erogano, in regime di autorizzazione, concessione o convenzione, servizi di pubblica utilità e chiedere ad essi la collaborazione, anche di ordine logistico, necessaria per l’adempimento delle loro funzioni istituzionali. > > La norma in esame, da un lato, rende cogente la collaborazione – ed anche l’assistenza, non prevista dalla norma vigente – che gli organismi di sicurezza eventualmente richiedono alle pubbliche amministrazioni. Dall’altro, specifica che la collaborazione e assistenza debbano essere motivate dalla necessità della tutela della sicurezza nazionale, mentre la disposizione vigente fa riferimento alla necessità di adempiere alle funzioni istituzionali di detti organismi. ... > Com’è stato rilevato dal Dott. Armando Spataro in sede di audizione[1], tale nuovo intervento normativo risponde all’”orientamento politico finalizzato ad estendere il ruolo delle Agenzie di Informazione nella direzione di attività che non competono loro, come – in particolare – quelle di indagine giudiziaria”. rob [1] https://www.senato.it/application/xmanager/projects/leg19/attachments/docum…

1 0

Gorwa, "È possibile la democrazia digitale?"
by J.C. DE MARTIN Dec. 19, 2024

Dec. 19, 2024

*È possibile la democrazia digitale? * /Due saggi indagano, da prospettive diverse, il complicato rapporto tra piattaforme online e partecipazione anti-gerarchica. Per capire se possiamo sottrarci ai modelli totalizzanti che governano la nostra attività in rete / Robert Gorwa 19 Dicembre 2024 All’inizio, ci fu una tormenta di neve. Era il gennaio del 1978 e forti raffiche subtropicali che si muovevano verso nord si scontrarono sui Grandi Laghi con una corrente artica diretta a sud, generando una delle più terribili tempeste di neve nella storia degli Stati uniti. La Grande tormenta di neve del 1978 scaricò quasi un metro di neve su alcune parti del Michigan, causando quasi un centinaio di vittime e mezzo miliardo di dollari di danni, al netto dell’inflazione. Bloccati a casa durante la grande nevicata, due appassionati di computer di Chicago decisero di costruire un sistema che gli avrebbe permesso di comunicare tra loro. All’epoca, la suite di protocolli Internet era ancora lontana anni dall’essere ampiamente utilizzata, ma i progressi nella microelaborazione avevano iniziato a portare i computer fuori dai laboratori e nelle case degli smanettoni. Con i primi modem si poteva «comporre il numero» e trasmettere informazioni su normali linee telefoniche, così Ward Christensen e Randy Suess, entrambi membri del Chicago Area Computer Hobbyists’ Exchange (Cache), unirono la loro competenza nella scrittura di software e nella modifica dell’hardware per creare un modem modificato a cui altri potevano collegarsi e sui quali lasciare messaggi da leggere in seguito. [...] continua qui: https://jacobinitalia.it/e-possibile-la-democrazia-digitale/

1 0

black box in ricerche astronomiche
by alessandro marzocchi Dec. 18, 2024

Dec. 18, 2024

... ... ... «Le reti neurali artificiali spesso vengono definite “*black box*” perché è difficile capire cosa avviene al loro interno: sappiamo cosa entra e cosa esce, ma non sempre riusciamo a interpretare come si arriva a certe conclusioni. Nel nostro caso, però, non è proprio così. Analizzando lo spazio latente – cioè le rappresentazioni interne che la rete costruisce per interpretare i dati – siamo riusciti a individuare una struttura ben precisa, che riflette alcune proprietà fisiche delle galassie, incluse alcune mai mostrate al modello durante l’addestramento. Un esempio? La rete distingue spontaneamente tra galassie passive ed attive, una caratteristica che non era parte del processo di apprendimento. È come se la rete avesse imparato da sola a identificare nuovi pattern e simmetrie nei dati, suggerendo informazioni emergenti. Questo non solo rende la *pipeline* più interpretabile, ma apre anche la strada a nuove scoperte». ... ... ... https://www.media.inaf.it/2024/12/17/m-topnet-ginolfi/ Nessuno ha definito l'intelligenza, ma quasi ogni giorno veniamo a conoscenza di "operazioni mentali" che non sono più nostra esclusiva come ritenevamo. Cordiali auguri. Duccio (Alessandro Marzocchi)

2 1