• ‘Thousands of Dollars for Something I Didn’t Do’ - The New York Times
    https://www.nytimes.com/2023/03/31/technology/facial-recognition-false-arrests.html

    “I’m locked up for something I have no clue about,” Mr. Reid, 29, said.

    His parents made phone calls, hired lawyers and spent thousands of dollars to figure out why the police thought he was responsible for the crime, eventually discovering it was because Mr. Reid bore a resemblance to a suspect who had been recorded by a surveillance camera. The case eventually fell apart and the warrants were recalled, but only after Mr. Reid spent six days in jail and missed a week of work.

    Mr. Reid’s wrongful arrest appears to be the result of a cascade of technologies — beginning with a bad facial recognition match — that are intended to make policing more effective and efficient but can also make it far too easy to apprehend the wrong person for a crime. None of the technologies are mentioned in official documents, and Mr. Reid was not told exactly why he had been arrested, a typical but troubling practice, according to legal experts and public defenders.

    “In a democratic society, we should know what tools are being used to police us,” said Jennifer Granick, a lawyer at the American Civil Liberties Union.

  • Google C.E.O. Sundar Pichai on the A.I. Moment: ‘You Will See Us Be Bold’ - The New York Times
    https://www.nytimes.com/2023/03/31/technology/google-pichai-ai.html

    Sundar Pichai has been trying to start an A.I. revolution for a very long time.

    In 2016, shortly after being named Google’s chief executive, Mr. Pichai declared that Google was an “A.I.-first” company. He spent lavishly to assemble an all-star team of A.I. researchers, whose breakthroughs powered changes to products like Google Translate and Google Photos. He even predicted that A.I.’s impact would be bigger than “electricity or fire.”

    So it had to sting when A.I.’s big moment finally arrived, and Google wasn’t involved.

    Instead, OpenAI — a scrappy A.I. start-up backed by Microsoft — stole the spotlight in November by releasing ChatGPT, a poem-writing, code-generating, homework-finishing marvel. ChatGPT became an overnight sensation, attracting millions of users and kicking off a Silicon Valley frenzy. It made Google look sluggish and vulnerable for the first time in years. (It didn’t help when Microsoft relaunched its Bing search engine with OpenAI’s technology inside, instantly ending Bing’s decade-long run as a punchline.)

    In an interview with The Times’s “Hard Fork” podcast on Thursday, his first extended interview since ChatGPT’s launch, Mr. Pichai said he was glad that A.I. was having a moment, even if Google wasn’t the driving force.

    #Intelligence_artificielle #Google

  • The Open Letter to Stop ‘Dangerous’ AI Race Is a Huge Mess | Chloe Xiang
    https://www.vice.com/en/article/qjvppm/the-open-letter-to-stop-dangerous-ai-race-is-a-huge-mess

    The letter was penned by the Future of Life Institute, a nonprofit organization with the stated mission to “reduce global catastrophic and existential risk from powerful technologies.” It is also host to some of the biggest proponents of longtermism, a kind of secular religion boosted by many members of the Silicon Valley tech elite since it preaches seeking massive wealth to direct towards problems facing humans in the far future. One notable recent adherent to this idea is disgraced FTX CEO Sam Bankman-Fried. Source: Motherboard

    • Gary Marcus a signé la lettre, il est très loin de la « AI Hype », et a un point de vue beaucoup plus pondéré.

      I am not afraid of robots. I am afraid of people.
      https://garymarcus.substack.com/p/i-am-not-afraid-of-robots-i-am-afraid

      For now, all the technolibertarians are probably cackling; if they had wanted to sabotage the “develop AI with care” crowd, they couldn’t have found a better way to divide and conquer.

      In truth, over 50,000 people signed the letter, including a lot of people who have nothing to do with the long term risk movement that the FLI itself is associated with. These include, for example, Yoshua Bengio (the most cited AI researcher in recent years), Stuart Russell (a well-known AI researcher at Berkeley), Pattie Maes (a prominent AI researcher at MIT), John Hopfield (a physicist whose original work on machine learning has been massively influential), Victoria Krakovna (a leading researcher at DeepMind working on how to get machines to do what we want them to do), and Grady Booch (a pioneering software architect who has been speaking out about the unreliability of current techniques as an approach to software engineering).

      But a few loud voices have overshadowed the 50,000 who have signed.

    • Un aspect qui me chagrine un peu, c’est que même chez Gary Marcus, ça se focalise sur des travers que seraient des utilisations frauduleuses de l’IA : désinformation et fishing essentiellement. (Et tout le monde nous fait un peu chier avec ces histoires de désinformation, comme si Trump, QAnon, les climatosceptiques et les covidiots, les gouvernements qui mentent, avaient besoin de la moindre IA pour générer et rendre « crédibles » leurs foutaises délirantes.)

      Pourtant il y a toutes les utilisations qui sont soit déjà légales, soit prochainement légales, et qui sont totalement épouvantables : « aide » à la justice (lui est noir et pauvre, il ira en prison parce que l’IA super-finaude a trouvé qu’il avait une tête à récidiver), « aide » aux contrôles des aides sociales (elle selon l’IA, elle a un profil à picoler sont argent de la CAAF, alors on va lui couper les vivres), pourquoi pas l’orientation des gamins avec des algorithmes qui font flipper tout le monde (je sais, Parcoursup est loin de l’IA, mais je n’ai aucun doute que c’est la prochaine étape), aide aux flics (celui-là, l’IA a décidé de te me le ficher S illico, vu qu’il est abonné au flux RSS de rezo.net et qu’il lit Bastamag…), automatisation complète de la médecine (au lieu d’une aide au diagnostic, on remplacera carrément le médecin avec une IA), etc.

      Automatisation des accès aux droits (immigration, solidarités, logement, éducation…), et incompétence organisées des personnels. Et renforcement de ce principe d’autorité (« le logiciel se trompe moins que les humains ») que déjà beaucoup de personnels ne sont plus en position de prendre la responsabilité d’aller à l’encontre d’une décision prise par un algorithme.

    • Ouais enfin quand tu t’impliques dans un débat, tu es censé te renseigner un peu sur ce qui s’est passé avant dans le champs.

    • Il faut que tu soies plus explicite.

      Ça fait un moment que je suis Gary Marcus, parce qu’il est justement opposé à la « AI Hype », qu’il a déjà publié plusieurs textes expliquant qu’il ne croit pas à l’avénement intelligence générale avec les outils actuels (ce n’est pas un gourou qui annonce la singularité grâce à Bitcoin et ChatGPT, ni un adepte du longtermisme). Et que dans le même temps, il avait déjà publié des textes de méfiance envers les outils actuels, avant de signer l’appel en question (dont il reconnaît explicitement des limites et des problèmes dans le texte qu’il a publié cette nuit – et il y évoque explicitement le texte de Timnit Gebru que tu as posté ci-dessus).

    • Je suppose que « se renseigner » fait référence au paragraphe 6.2 du document On the Dangers of Stochastic Parrots : Can Language Models Be Too Big ? (mars 2021)
      https://dl.acm.org/doi/pdf/10.1145/3442188.3445922

      6.2 Risks and Harms
      The ersatz fluency and coherence of LMs raises several risks, precisely because humans are prepared to interpret strings belonging to languages they speak as meaningful and corresponding to the communicative intent of some individual or group of individuals who have accountability for what is said. We now turn to examples, laying out the potential follow-on harms.

      Là où Gary Marcus a tendance à insister sur des usages plus volontairement nuisibles (« bad actors ») :
      https://www.theatlantic.com/technology/archive/2023/03/ai-chatbots-large-language-model-misinformation/673376

      Et quand ça passe au grand public, ça devient particulièrement éthéré. L’édito d’Ezra Klein dans le NY Times (il y a 15 jours) a peut-être influencé l’émergence de l’appel, et c’est très très flou sur les risques liés à l’AI (grosso modo : « c’est tellement puissant qu’on ne comprend pas vraiment », pas loin de la Hype AI) :
      https://www.nytimes.com/2023/03/12/opinion/chatbots-artificial-intelligence-future-weirdness.html%20https://www.nytimes.com/2023/03/12/opinion/chatbots-artificial-intelligence-future-weirdness.html

    • Je ne sais pas comment faire plus explicite. Une pétition sur l’IA cosignée par Melon Musk et pas par M. Mitchell ou T. Gebru, quand tu connais un tout petit peu le domaine, tu devrais juste te méfier avant d’engager ton nom. Mais bon… you do you, comme on dit.

  • Ukraine Says It Will Use Legal Means to Evict Orthodox Pro-Russia Monks - The New York Times
    https://www.nytimes.com/2023/03/29/world/europe/ukraine-russia-orthodox-monks.html

    Many Ukrainians argue that the church has not clearly stated its position on the conflict and is therefore compromised. The Ukrainian security services have gone further, describing the Russian-aligned church as an incubator of pro-Russia sentiment and infiltrated by priests and monks who have directly aided Moscow in the war.
    Image

    Dozens of priests and monks from the Moscow Patriarchate have been arrested in recent months, accused of spying for the Kremlin and even helping to direct Russian airstrikes. President Volodymyr Zelensky of Ukraine has said the expulsions — which affect roughly 200 monks and 300 trainees — are needed for “spiritual independence.”

  • An Untold Story Behind Jimmy Carter’s Presidential Defeat - The New York Times
    https://www.nytimes.com/2023/03/18/us/politics/jimmy-carter-october-surprise-iran-hostages.html?te=1&nl=from-the-times&emc

    Comment un des président les plus progressistes des Etats-unis s’est fait voler sa ré-élection par des Républicains qui indiquent à l’Iran qu’il faut garder les otages jusqu’à l’élection de Reagan... et le néolibéralisme fut.

    Peter Baker

    By Peter Baker

    March 18, 2023

    WASHINGTON — It has been more than four decades, but Ben Barnes said he remembers it vividly. His longtime political mentor invited him on a mission to the Middle East. What Mr. Barnes said he did not realize until later was the real purpose of the mission: to sabotage the re-election campaign of the president of the United States.

    It was 1980 and Jimmy Carter was in the White House, bedeviled by a hostage crisis in Iran that had paralyzed his presidency and hampered his effort to win a second term. Mr. Carter’s best chance for victory was to free the 52 Americans held captive before Election Day. That was something that Mr. Barnes said his mentor was determined to prevent.

    His mentor was John B. Connally Jr., a titan of American politics and former Texas governor who had served three presidents and just lost his own bid for the White House. A former Democrat, Mr. Connally had sought the Republican nomination in 1980 only to be swamped by former Gov. Ronald Reagan of California. Now Mr. Connally resolved to help Mr. Reagan beat Mr. Carter and in the process, Mr. Barnes said, make his own case for becoming secretary of state or defense in a new administration.
    ImageMr. Carter sitting at a desk in the Oval Office, looking down.
    The hostage crisis in Iran hampered Mr. Carter’s effort to win a second term.Credit...Associated Press
    Mr. Carter sitting at a desk in the Oval Office, looking down.

    What happened next Mr. Barnes has largely kept secret for nearly 43 years. Mr. Connally, he said, took him to one Middle Eastern capital after another that summer, meeting with a host of regional leaders to deliver a blunt message to be passed to Iran: Don’t release the hostages before the election. Mr. Reagan will win and give you a better deal.

    Then shortly after returning home, Mr. Barnes said, Mr. Connally reported to William J. Casey, the chairman of Mr. Reagan’s campaign and later director of the Central Intelligence Agency, briefing him about the trip in an airport lounge.

    Mr. Carter’s camp has long suspected that Mr. Casey or someone else in Mr. Reagan’s orbit sought to secretly torpedo efforts to liberate the hostages before the election, and books have been written on what came to be called the October surprise. But congressional investigations debunked previous theories of what happened.

    #Jimmy_Carter #Ronald_reagan #Magouille

  • A ‘New Cold War’ Looms in Africa as U.S. Pushes Against Russian Gains - The New York Times
    https://www.nytimes.com/2023/03/19/world/africa/chad-russia-wagner.html

    Les #états-unis au secours de la #Françafrique

    In Africa, the more forceful American approach aims partly to shore up the crumbling position of #France, which in recent years has ceded ground to Russia in former colonies like Mali and the Central African Republic.

  • Credit Suisse to Borrow as Much as $54 Billion From Swiss Central Bank - The New York Times
    https://www.nytimes.com/2023/03/15/business/economy/credit-suisse-swiss-bank-54-billion-loan.html

    The Swiss lender had ended Wednesday fighting for its life, with shares at a record low and the cost to insure against a default the highest ever. The loan is an effort to avert further damage.

    Plutôt que de renflouer des canards boiteux, ne devrait-on pas plutôt aider l’Ukraine ? 54 milliards, c’est quasiment le budget annuel de l’armée Russe après tout. Et c’est quasiment ce qu’a reçu l’Ukraine en aide financière militaire.

  • Pentagon Blocks Sharing Evidence of Possible Russian War Crimes With Hague Court
    https://www.nytimes.com/2023/03/08/us/politics/pentagon-war-crimes-hague.html

    The Pentagon is blocking the Biden administration from sharing evidence with the International Criminal Court in The Hague gathered by American intelligence agencies about Russian atrocities in Ukraine, according to current and former officials briefed on the matter.

    American military leaders oppose helping the court investigate Russians because they fear setting a precedent that might help pave the way for it to prosecute Americans. The rest of the administration, including intelligence agencies and the State and Justice Departments, favors giving the evidence to the court, the officials said.

    President Biden has yet to resolve the impasse, officials said.

    The evidence is said to include details relevant to an investigation the chief prosecutor of the International Criminal Court, Karim Khan, began after Russia’s invasion of Ukraine a year ago. The information reportedly includes material about decisions by Russian officials to deliberately target civilian infrastructure and to abduct thousands of Ukrainian children from occupied territory.

    In December, Congress modified longstanding legal restrictions on American help to the court, allowing the United States to assist with its investigations and eventual prosecutions related to the war in Ukraine. But inside the Biden administration, a policy dispute over whether to do so continues to play out behind closed doors.

    [...] In late December, lawmakers enacted two laws aimed at increasing the chances that Russians would be held accountable for war crimes in Ukraine.

    One was a stand-alone bill expanding the jurisdiction of American prosecutors to charge foreigners for war crimes committed abroad. The other, a provision about the International Criminal Court embedded in the large appropriations bill Congress passed in late December, received little attention at the time.

    But that provision was significant. While the U.S. government remains prohibited from providing funding and certain other aid to the court, Congress created an exception that allows it to assist with “investigations and prosecutions of foreign nationals related to the situation in #Ukraine, including to support victims and witnesses.”

    Despite that legal change and Congress’s signal of support, the Pentagon has stood firm that the United States should not help the International Criminal Court investigate Russians for their actions in Ukraine since Russia is not a party to the treaty that established the court.

    That resistance has attracted criticism both inside and outside the executive branch. Some legal specialists contend that there is scant benefit to hewing to that position because the rest of the world essentially rejects that interpretation.

    #cour_pénale_internationale

  • Intelligence Suggests Pro-Ukrainian Group Sabotaged Pipelines, U.S. Officials Say - The New York Times
    https://www.nytimes.com/2023/03/07/us/politics/nord-stream-pipeline-sabotage-ukraine.html


    The Nord Stream 2 gas pipeline in Lubmin, Germany, last year.
    Credit. Krisztian Bocsi/Bloomberg

    New intelligence reporting amounts to the first significant known lead about who was responsible for the attack on the Nord Stream pipelines that carried natural gas from Russia to Europe.
    […]
    Officials who have reviewed the intelligence said they believed the saboteurs were most likely Ukrainian or Russian nationals, or some combination of the two. U.S. officials said no American or British nationals were involved.

    The pipelines were ripped apart by deep sea explosions in September, in what U.S. officials described at the time as an act of sabotage. European officials have publicly said they believe the operation that targeted Nord Stream was probably state sponsored, possibly because of the sophistication with which the perpetrators planted and detonated the explosives on the floor of the Baltic Sea without being detected. U.S. officials have not stated publicly that they believe the operation was sponsored by a state.

    The explosives were most likely planted with the help of experienced divers who did not appear to be working for military or intelligence services, U.S. officials who have reviewed the new intelligence said. But it is possible that the perpetrators received specialized government training in the past.

    Officials said there were still enormous gaps in what U.S. spy agencies and their European partners knew about what transpired. But officials said it might constitute the first significant lead to emerge from several closely guarded investigations, the conclusions of which could have profound implications for the coalition supporting Ukraine.

    Any suggestion of Ukrainian involvement, whether direct or indirect, could upset the delicate relationship between Ukraine and Germany, souring support among a German public that has swallowed high energy prices in the name of solidarity.

    • Enquêtes Nord Stream : des traces mènent à l’Ukraine
      https://www.zeit.de/politik/ausland/2023-03/nordstream-2-ukraine-anschlag

      Concrètement, selon les informations du studio ARD Capital, les enquêteurs sont contrastés, SWR et ZEIT ont réussi à identifier le bateau qui aurait été utilisé pour l’opération secrète. Il s’agirait d’un yacht loué à une société basée en Pologne, appartenant apparemment à deux Ukrainiens. Selon l’enquête, l’opération secrète en mer a été menée par une équipe de six personnes. On dit qu’il s’agissait de cinq hommes et d’une femme. En conséquence, le groupe était composé d’un capitaine, de deux plongeurs, de deux assistants de plongée et d’un médecin, qui auraient transporté les explosifs sur les lieux du crime et les y auraient placés. La nationalité des auteurs n’est apparemment pas claire. Les assassins ont utilisé des passeports falsifiés par des professionnels, qui auraient été utilisés, entre autres, pour louer le bateau.

      Selon l’enquête, le commandement a quitté Rostock le 6 septembre 2022. L’équipement pour l’opération secrète était auparavant transporté au port dans un camion de livraison, dit-on. Selon les recherches, les enquêteurs ont ensuite réussi à localiser à nouveau le bateau le lendemain à Wieck (Darß) et plus tard sur l’île danoise de Christiansø, au nord-est de Bornholm. Le yacht a ensuite été rendu au propriétaire dans un état non nettoyé. Selon les recherches, les enquêteurs ont trouvé des traces d’explosifs sur la table dans la cabine. Selon les informations du studio capital ARD, Kontraste, SWR et ZEIT, un service secret occidental aurait envoyé à l’automne, c’est-à-dire peu après la destruction, un tuyau aux services partenaires européens, selon lequel un commando ukrainien serait responsable de la destruction. Après cela, il y aurait eu d’autres indications de renseignement qu’un groupe pro-ukrainien pourrait être responsable.

      traduction bancale de gg...

    • Unique source : « US officials ». La seule chose qu’on peut en tirer, c’est essayer de se faire des nœuds au cerveau de ce que signifie une telle publication dans le NYTimes. Par contre rigoureusement rien sur la crédibilité de l’affirmation.

    • C’est sûr que toute cette « enquête » pue la barbouzerie. On peut juste remarquer qu’ils n’ont pas osé faire dire aux « journalistes » que les services n’avaient rien vu venir et qu’ils avaient été les premiers surpris.
      le début de l’article du Spiegel :

      phrase 1

      Les autorités allemandes d’enquête ont réalisé une percée dans l’enquête sur l’attaque des gazoducs NS1 et NS2.

      phrase suivante

      Après une enquête (Recherche) commune du studio berlinois d’ARD, du magazine politique d’ARD Kontraste, de la SWR et du Zeit, il a pu être reconstitué, à la suite des investigations, comment et quand l’attaque à l’explosif a été préparée.

      Normalement, les « autorités » (Behörden) ça ne peut pas être des médias…

      Et le début du passage que traduit @Lyco réalise une audacieuse synthèse des deux sources :

      En pratique, d’après les informations de ARD-Berlin, Kontraste, SWR et le Zeit, les enquêteurs ont réussi à identifier le bateau soupçonné d’avoir été utilisé pour l’opération secrète.

      La femme mentionnée était « le » médecin :- )

      Plus loin, et, effectivement, l’original allemand doit être lu avec un soin bien plus grand que gg:translate

      Les enquêteurs de l’enquête commune (Recherche, le terme renvoie précisément à l’enquête menée par les médias) ont pu identifier des traces d’explosif sur la table du yacht.

      D’après des informations de (toujours) ces (mêmes) enquêteurs des médias, un service secret occidental a dû, peu de temps après l’attaque, transmettre à certains de ces partenaires européens un avis selon lequel ce serait un commando ukrainien qui serait responsable de la destruction.

      dans l’article allemand, aucune référence à des US officials, que de vaillants et fiers journalistes qui ont pu tranquillement remonter la filière… et, avec leur petit bras et leur farouche volonté de découvrir la vérité, ainsi trouver tout cela. Où il subsiste tout de même beaucoup de conditionnels.

    • sur l’article du NYT cf. ces commentaires, via @dedefensa
      https://seenthis.net/messages/993465

      (je dois avouer que je n’ai pas lu l’article du New York Times, je me suis focalisé sur celui du Zeit, où il y avait déjà pas mal de quoi faire…)

      (oui, je sais Zeit c’est féminin, et à chaque fois ça me fait bizarre, alors je précise : c’est une ellipse

      le Zeit

      pour

      l’hebdomadaire Die Zeit

      peut-être que ça me soulagera…