- nexa - server-nexa.polito.it

How AI Can Be Regulated Like Nuclear Energy
by Daniela Tafani Oct. 25, 2023

Oct. 25, 2023

How AI Can Be Regulated Like Nuclear Energy By Heidy Khlaaf October 24, 2023 11:48 AM Prominent AI researchers and figures have consistently dominated headlines by invoking comparisons that AI risk is on par with the existential and safety risks that were posed with the coming of the nuclear age. From statements that AI should be subject to regulation akin to nuclear energy, to declarations paralleling the risk of human extinction to that of nuclear war, the analogies drawn between AI and nuclear have been consistent. The argument for such extinction risk has hinged on the hypothetical and unproven risk of an Artificial General Intelligence (AGI) imminently arising from current Large Language Models (e.g., ChatGPT), necessitating increased caution with their creation and deployment. Sam Altman, the CEO of OpenAI, has even referenced to the well established nuclear practice of “licensing”, deemed anti-competitive by some. He has called on the creation of a federal agency that can grant licenses to create AI models above a certain threshold of capabilities. This is akin to how operators of nuclear facilities are required to be licensed by a nuclear regulator, such as the US Nuclear Regulatory Commission and the UK Office for Nuclear Regulation. As part of their licence conditions, nuclear operators must comply with regulatory safety guidelines and a legal duty to reduce risks so far as is reasonably practicable. The incessant references to nuclear safety should naturally lend urgency in applying nuclear-level regulations and assessments to the very AI systems being deliberated. Yet, there is only resistance to be met at the mention of the EU AI Act, a whittled down document compared to the intricacies and rigour of the regulations required for nuclear systems. Indeed Altman has previously noted that the EU AI Act would be “over-regulating”, specifically noting OpenAI's systems as they exist are not high-risk, and that OpenAI will cease to operate if they cannot comply with the upcoming EU AI Act. Two conflicting messages are thus being presented. One being that the future of AI capabilities present such a significant risk, that an exclusive nuclear-level regulatory licensing scheme is required, but also that existing systems do not warrant any regulation, despite well-documented harms, and deemed insignificant. And herein where the analogies to nuclear safety crumble. Safety Engineering, the discipline which assures the safety of nuclear systems, is built on the foundation that system failures or misbehaviours, no matter how small, can compound and lead to catastrophic events. Consider that all components of a nuclear system, that being uranium or plutonium, hardware, and software, down to literal concrete, are individually and rigorously regulated. These elements separately do not pose a threat of nuclear catastrophe, but in orchestration lead to the emergence of a controlled nuclear fission reaction. Conversely, failures in said orchestration can cascade into catastrophic or high-risk events. For example, a minuscule software bug known as a race condition led to the Northeast blackout of 2003. Our inability to prevent today’s AI harms, such as algorithmic discrimination and reducing the cost of disinformation or cybersecurity attacks, only entails that we are ill-prepared to trace and grasp any cascading implications and control of AI risks. And if we lack the technical and governing facilities to control or measure AI mechanisms which output these harms, then we would be mistaken to believe we have the technical foundation to resolve larger-scale avalanching risks that AI may hypothetically enable. Then why have AI labs shown resistance to following through with the nuclear-level rigour required to keep the predecessors (e.g., ChatGPT) of “extinction level” technology safe? In a blog post by the OpenAI founders, they note that “it would be important that such an [regulatory] agency focus on reducing existential risk and not issues that should be left to individual countries.” This statement attempts to not only detach the existing harms AI systems pose from hypothetical existential risks, but fails to recognise the fundamental engineering principle that harms compound. If current AI systems are indeed the foundations of a hypothetical AGI, as is often implied, then as they stand they too must be regulated to address current harms in preparation for further emergent behaviour, whether it be AGI or otherwise. A pattern emerges that invoking the analogy to nuclear war often serves to inflame specific narratives, whether it be exclusive licensing, or an attempt to divert the regulatory conversation to focus on speculative future threats instead. Both having potentially harmful impacts. For the former, AI-based systems do not possess any unique software components that warrant a generalized licensing scheme that would not heavily impede the use of software and hardware as a whole. Indeed, any implementation of such a scheme would likely result in significant overreach due to the broad definition and software components of AI systems, lending basic blocks for technological advancements to be available to a privileged few. After all, current AI systems are built using traditional hardware and software components. Even Generative AI technologies utilise Deep Neural Networks, a technique that dates back to the 1970’s. For the latter (abiding by the foundations of nuclear safety aside), current AI systems have demonstrated enough harms to warrant regulation even without an extinction threat. Nuclear-level regulation methodologies already exist for the very same hardware, software, and data components, ready to be adopted any day for AI labs to utilise for high-risk applications. Yet we have observed that the very pushback and lobbying by AI labs against the EU AI Act has proven to be fruitful, having already distracted from regulatory efforts relative to the harms posed by these AI systems today. A post by the European Commission reflected the exact wording of the extinction letter, "Mitigating the risk of extinction from AI should be a global priority”, a far cry from the risks outlined in the initial drafts of the act. If AI labs are to consistently invoke exaggerated fears through comparisons to nuclear hazards, then they must be willing to take the analogy to completion. Ironically, if they were to explore the readily available safety mechanisms for nuclear components, they would be pressed to find inflammatory language supporting the use of terminology such as “extinction” and “existential risks”. Indeed, nuclear catastrophe is supported by known scientific and geo-political capabilities that have been theoretically and meticulously studied even during the Manhattan Project. The risks identified are tethered to tractable and compounding risks. Meanwhile, there exists no scientific basis or evidence for how or when AGI will emerge (if ever), leaving us with only a hypothetical risk that has capitulated many of our regulatory efforts today. https://time.com/6327635/ai-needs-to-be-regulated-like-nuclear-weapons/

1 0

Un caso di "intercettazione telematica", in Europa (Hetzner, Germania)
by Damiano Verzulli Oct. 25, 2023

Oct. 25, 2023

Segnalo un articolo un po' tecnico che descrive uno scenario nel quale il traffico XMPP (un protocollo utilizzato per servizi a-la "chat" sincrona), seppur protetto da crittografia (TLS), è stato oggetto di intercettazione, verosimilmente da parte delle autorita' tedesche. In particolare, il provider tedesco che offriva l'infrastruttura "cloud" con la quale veniva gestito il relativo "server" (parliamo di "jabber.ru" e di "xmpp.ru"), si è adoperato per "riorganizzare" il collegamento Internet del server in esame (isolandolo dal resto dell'infrastruttura e lasciandolo connesso, in disparte dal resto) e poi "aggiungere" un livello di gestione lato TLS (ossia, mettendosi in mezzo fra i client sparsi su Internet, ed il server che era li in casa) facendo in modo che entrambi (client e server) parlassero TLS, senza rendersi conto che lui, pero', era nel mezzo (a guardare e potenzialmente alterare il relativo traffico). E' interessante osservare che il tutto è emerso alla luce del sole, solo a causa di una "distrazione" di un qualche tecnico... che si è "dimenticato" di rinnovare un certificato TLS ( che quindi è scaduto, portanto i client a.. innervosirsi... e cosi' innescando l'analisi tecnica degli admin). Che io sappia, è la prima volta che vedo "documentato" uno schema del genere. I dettagli, sono qui: https://notes.valdikss.org.ru/jabber.ru-mitm/ Ne è nato un piccolo thread (ancora piu' tecnico) sulla community ITNOG [1] dove provavo a riflettere su come "detectare" tali situazioni e, ove necessario, "impedirle". Saluti, DV [1] https://t.me/IT_NOG/99331/115800 -- Damiano Verzulli e-mail: damiano(a)verzulli.it --- possible?ok:while(!possible){open_mindedness++} --- "...I realized that free software would not generate the kind of income that was needed. Maybe in USA or Europe, you may be able to get a well paying job as a free software developer, but not here [in Africa]..." -- Guido Sohne - 1973-2008 http://ole.kenic.or.ke/pipermail/skunkworks/2008-April/005989.html

4 6

Managing AI Risks in an Era of Rapid Progress
by don Luca Peyron Oct. 24, 2023

Oct. 24, 2023

Documento interessante, buona lettura dl https://managing-ai-risks.com/ _________________________ don Luca Peyron Pastorale Universitaria - Apostolato Digitale Arcidiocesi di Torino www.universitari.to.it via XX settembre 83, Torino tel. 011 5156239

1 0

Trasparenza dell’algoritmo, la Cassazione: sì al rating reputazionale, se in linea con la privacy
by don Luca Peyron Oct. 24, 2023

Oct. 24, 2023

Buona lettura dl https://www.cybersecurity360.it/legal/privacy-dati-personali/trasparenza-de… _________________________ don Luca Peyron Pastorale Universitaria - Apostolato Digitale Arcidiocesi di Torino www.universitari.to.it via XX settembre 83, Torino tel. 011 5156239

1 0

MEMENTO | 108° Nexa Lunch Seminar | 25 ottobre 2023, ore 13.00
by Valeria Bergantino Oct. 23, 2023

Oct. 23, 2023

Gentilissime, gentilissimi, Vi ricordiamo che *mercoledì 25 ottobre*, dalle ore 13.00 alle ore 14.00, si terrà il *108° Nexa Lunch Seminar* con un incontro dal titolo: /*"Analisi e report di trasferimenti di dati personali verso domini extra-EEA"*/. Ospite dell'incontro: *Lorenzo Laudadio* (Politecnico di Torino). _L'incontro si terrà IN PRESENZA e ONLINE._ _SEDE FISICA_ dell'incontro: Centro Nexa su Internet e Società, Politecnico di Torino, Via Boggio 65/a, Torino (1° piano). Per accedere alla sala si raccomanda di suonare al citofono *Portineria* e di seguire le indicazioni lungo il percorso. QUI <https://nexa.polito.it/contatti> maggiori informazioni su come raggiungerci. _STANZA VIRTUALE _dell'incontro: https://didattica.polito.it/VClass/NexaEvent Per maggiori dettagli è possibile consultare la pagina web dell'evento: https://nexa.polito.it/lunch-108 Cordiali saluti, -- Valeria Bergantino Communication Officer Nexa Center for Internet & Society Politecnico di Torino - DAUIN Via Pier Carlo Boggio, 65/A - 10138 Torino web: https://nexa.polito.it/ mail: valeria.bergantino(a)polito.it <mailto:valeria.bergantino@polito.it> tel: 3473443585

1 0

Open letter: European Parliament must stand strong on AI Act fundamental rights protections - Access Now
by Daniela Tafani Oct. 23, 2023

Oct. 23, 2023

Open letter: European Parliament must stand strong on AI Act fundamental rights protections - Access Now For the attention of Brando Benifei MEP and Dragoş Tudorache MEP, We, the undersigned organisations, are writing to you in advance of the trilogue negotiation on the Artificial Intelligence Act on October 24 to encourage you to uphold the vital fundamental rights protections included in the Parliament’s position. It is absolutely essential that the AI Act puts people first, and regulates the use of AI in law enforcement, migration, and national security. As the true democratic representatives of the people of the European Union in these negotiations, the responsibility lies with you to resist pressure to reduce essential fundamental rights protections, and to overcome the intransigence of the co-legislators on key issues related to law enforcement, migration, transparency, and accountability. The Parliament’s position added essential protections for fundamental rights, such as bans on biometric mass surveillance in public spaces, emotion recognition, predictive policing, and biometric categorisation, as well as prohibiting the export of such systems. These are truly dystopian uses of AI that undermine the core of our rights, our freedoms, and our democracy. These uses of AI are objectively incompatible with our fundamental rights, and this has been confirmed by the European Data Protection Board and European Data Protection Supervisor in their joint opinion on the AI Act, where they called for precisely these prohibitions. In the Council’s general approach, as well as in recent compromise texts, dangerous loopholes have been presented to allow law enforcement and migration authorities to escape essential public transparency obligations. This would deprive affected people of essential, basic information about the most harmful AI systems that can be used and misused against them. A blanket exemption has also been added for systems developed or used for national security purposes, which contradicts both the treaties and the jurisprudence of the courts. The fundamental rights of people affected by AI cannot be compromised. Sacrificing essential protections for fundamental rights would render the AI Act unacceptable and unfit for purpose. We call on you to resist attempts to dilute these essential protections, as well as other key achievements of the Parliament position, such as fundamental rights impact assessments (FRIAs). If a political deal is reached at the expense of these vital issues, then two and a half years of work on this file will have been for nothing. Please stand up for the rights of the people of the European Union, defend the Parliament mandate, and ensure that the AI Act brings real protections for people. Signatories Access Now, European Digital Rights (EDRi) The Border Violence Monitoring Network (BVMN) European Center for Not-for-Profit Law (ECNL)Amnesty International Lafede.cat – Organitzacions per la Justícia Global Panoptykon Foundation Elektronisk Forpost Norge (EFN) Algorights European Network Against Racism (ENAR) AlgorithmWatch Fair Trials Media Contact If you would like to arrange an interview with one of our Access Now experts, have a question, or are interested in further information on any of our areas of focus, please contact press [at] accessnow [dot] org. Subscribe to receive media alerts. Related Artificial Intelligence Read More Contact Careers How We Use Your Data Media Usage Code of Conduct Site Terms of Use Crafted by Cornershop Creative Facebook Twitter Instagram YouTube LinkedIn TikTok https://www.accessnow.org/press-release/european-parliament-must-stand-stro…

1 0

Fwd: Oggi 21 ottobre: Global Encryption Day
by Andrea Trentini Oct. 21, 2023

Oct. 21, 2023

-------- Forwarded Message -------- Subject: Oggi 21 ottobre: Global Encryption Day Date: Sat, 21 Oct 2023 07:17:05 +0200 From: Valerio Bozzolan <gnu(a)linux.it> To: soci(a)lists.linux.it Segnalo, Oggi, 21 ottobre, e' il Global Encryption Day (#GlobalEncryptionDay) Qui materiali utili per i propri social: https://www.globalencryption.org/events/ged/2023-social-toolkit/ Buona condivisione! P.S. Se una persona traduttrice avesse tempo libero, è benvenuta nel farsi avanti, per tradurre almeno un comunicato in italiano: https://www.globalencryption.org/events/ged/campaign-in-a-box/ Nel caso, meglio avvisare qui per mail, cosi' non lo facciamo in due. Ciau :) -boz -- Valerio Bozzolan Italian Linux Society - direttore volontario https://www.ils.org/ https://www.linux.it/ E-mail sent from Evolution from a random GNU/Linux distribution, delivered from my Postfix mailserver. Have fun with software freedom! -- Mailing list info: https://lists.linux.it/listinfo/soci

1 0

How Meta and AI companies recruited striking actors to train AI
by J.C. DE MARTIN Oct. 20, 2023

Oct. 20, 2023

*How Meta and AI companies recruited striking actors to train AI* /Hollywood actors are on strike over concerns about the use of AI, but for as little as $300, Meta and a company called Realeyes hired them to make avatars appear more human./ Eileen Guoarchive October 19, 2023 https://www.technologyreview.com/2023/10/19/1081974/meta-realeyes-artificia…

1 0

Marc Andreessen e' fuori come un balcone
by Marco Fioretti Oct. 20, 2023

Oct. 20, 2023

(si dice cosi' anche fuori Roma, quando qualcuno e' completamente uscito di senno? anche se nel suo caso probabilmente meta' e' che ci crede davvero, meta' che cerca soldi per il suo fondo di investimenti) Ho visto SOLO ora sia l'esilarante (nei modi piu' sbagliati) Manifesto del Tecno-ottimista: https://a16z.com/the-techno-optimist-manifesto/ (https://a16z.com/the-techno-optimist-manifesto/) sia la critica di G. Marcus (da NON perdersi i commenti, che non sanno nemmeno da dove cominciare per sfottere Andreessen): https://garymarcus.substack.com/p/two-kinds-of-techno-optimism (https://garymarcus.substack.com/p/two-kinds-of-techno-optimism) Boh. Se posso ripetermi, Andreessen e' mezzo bambino fuori di testa, mezzo furbone, "Solaria all the way down" come scrivevo per pura coincidenza due giorni fa: https://mfioretti.substack.com/p/taking-ai-too-seriously-is-not-intelligent (https://mfioretti.substack.com/p/taking-ai-too-seriously-is-not-intelligent) e se quanto ho scritto sembra luddita, segnalo che sono pure io tecno ottimista, ma come si deve, non certo alla Andreessen: https://mfioretti.substack.com/p/how-to-tell-good-ai-from-bad-or-pointless (https://mfioretti.substack.com/p/how-to-tell-good-ai-from-bad-or-pointless) Ripeto, boh. Come fa uno cosi' ad avere ancora un minimo di credito? Marco

3 3

"Jon Stewart’s Apple TV Plus show ends, reportedly over coverage of AI and China"
by J.C. DE MARTIN Oct. 20, 2023

Oct. 20, 2023

*Jon Stewart’s Apple TV Plus show ends, reportedly over coverage of AI and China * /According to reports, The Problem With Jon Stewart has been effectively canceled over editorial disagreements with Apple. / https://www.theverge.com/2023/10/19/23924549/jon-stewart-apple-ai-china-can…

1 0