RISKS Forum mailing list archives
Risks Digest 33.86
From: RISKS List Owner <risko () csl sri com>
Date: Sat, 23 Sep 2023 21:04:32 PDT
RISKS-LIST: Risks-Forum Digest Saturday 23 September 2023 Volume 33 : Issue 86 ACM FORUM ON RISKS TO THE PUBLIC IN COMPUTERS AND RELATED SYSTEMS (comp.risks) Peter G. Neumann, founder and still moderator ***** See last item for further information, disclaimers, caveats, etc. ***** This issue is archived at <http://www.risks.org> as <http://catless.ncl.ac.uk/Risks/33.86> The current issue can also be found at <http://www.csl.sri.com/users/risko/risks.txt> Contents: Driverless Car Company Using Chatbots to Make Its Vehicles Smarter (MIT Tech Review) ChatGPT Can Now Generate Images (NYTimes) Prominent Authors Sue OpenAI Google Search first result for "Tank Man" is e fake AI image rather than actual image from China (404Media + Lauren Weinstein) Misinformation research is buckling under GOP legal attacks (WashPost) Egyptian presidential hopeful targeted by Predator spyware (WashPost) It's 2030, and digital wallets have replaced every card in our purses and pockets (ZDNET) Google accused of directing motorist to drive off collapsed bridge (BBC) Typeface trolls shaking down users of Adobe's font platform (BoingBoing) Bitcoin conspiracy theory (PGN via John Markoff) Re: Pedestrian dies after Cruise cars block ambulance (Amos Shapir, John Levine) Abridged info on RISKS (comp.risks) ---------------------------------------------------------------------- Date: Wed, 20 Sep 2023 12:08:18 -0400 (EDT) From: ACM TechNews <technews-editor () acm org> Subject: Driverless Car Company Using Chatbots to Make Its Vehicles Smarter (MIT Tech Review) Will Douglas Heaven, MIT Technology Review 14 Sep 2023, via ACM TechNews, 20 Sep 2023 U.K.-based driverless car company Wayve has tapped chatbot technology to question its vehicles about their driving decisions. The company combined its self-driving software with a large language model into the LINGO-1 hybrid model, which synchronizes video and driving data with natural-language descriptions that record the car's observations and actions. Wayve aims to know how and why its cars make certain decisions by quizzing the self-driving software at every step, helping to expose flaws faster than sifting through video playbacks or scrolling through error reports. The University of California, Berkeley's Pieter Abbeel said, ``With a system like LINGO-1, I think you get a much better idea of how well it understands driving in the world.'' [Smarter than what? A bedpost? PGN] ------------------------------ Date: Fri, 22 Sep 2023 11:35:28 -0400 (EDT) From: ACM TechNews <technews-editor () acm org> Subject: ChatGPT Can Now Generate Images (NYTimes) Cade Metz and Tiffany Hsu, *The New York Times*, 20 Sep 2032, via ACM TechNews; 22 Sep 2023 OpenAI has integrated a new version of its DALL-E image generator into its ChatGPT online chatbot. DALL-E 3 generates more detailed images than its predecessors, with notable improvements in images featuring letters, numbers, and human hands. The new version of the image generator can create images from multi-paragraph descriptions and follow detailed instructions. OpenAI's Aditya Ramesh said DALL-E 3 was given a more precise understanding of the English language. The DALL-E/ChatGPT integration means ChatGPT can generate digital images based on detailed textual descriptions provided by users or produced by the chatbot itself. OpenAI has included tools in DALL-E 3 to prevent the generation of sexually explicit images, images of public figures, and images that imitate the styles of specific artists. [It's all over now, when you can create a realistic image of almost anyone saying almost anything, however faked, ridiculous, and perhaps even irrefutable. The world as we knew it is no longer. Ground meat was always a mystery, but ground truth seems to be irrelevant. Now even what seems to be the real thing may not be real, as in 3-D printing bots cranking out not only tender synthetic steaks, spare ribs, and replaceable body parts, but also fake binding contracts, `originally' signed masterpiece paintings, altered classic movies, and almost anything else. PGN] ------------------------------ Date: Wed, 20 Sep 2023 10:48:47 -0400 From: Jan Wolitzky <jan.wolitzky () gmail com> Subject: Prominent Authors Sue OpenAI (NYTimes +) A group of prominent novelists, including John Grisham, Jonathan Franzen and Elin Hilderbrand, are joining the legal battle against OpenAI over its chatbot technology, as fears about the encroachment of artificial intelligence on creative industries continue to grow. More than a dozen authors filed a lawsuit against OpenAI on Tuesday, accusing the company, which has been backed with billions of dollars in investment from Microsoft, of infringing on their copyrights by using their books to train its popular ChatGPT chatbot. The complaint, which was filed along with the Authors Guild, said that OpenAI's chatbots can now produce *derivative works* that can mimic and summarize the authors' books, potentially harming the market for authors' work, and that the writers were neither compensated nor notified by the company. ``The success and profitability of OpenAI are predicated on mass copyright infringement without a word of permission from or a nickel of compensation to copyright owners,'' the complaint said. https://www.nytimes.com/2023/09/20/books/authors-openai-lawsuit-chatgpt-copyright.html [Ellen Ullman noted (via Dave Farber) another article on this subject: "Algorithmic destruction" and the deep algorithmic problems of AI and copyright, https://www.sfchronicle.com/tech/article/ai-artificial-intelligence-copyright-18374295.php where Chase DiFeliciantonio asks ``Could *algorithmic destruction* solve AI's copyright issues? PGN] ------------------------------ Date: Wed, 20 Sep 2023 11:17:25 -0700 From: Lauren Weinstein <lauren () vortex com> Subject: Google Search first result for "Tank Man" is e fake AI image rather than actual image from China (404MEdia) https://www.404media.co/first-google-search-result-for-tiananmen-square-tank-man-is-ai-generated-selfie/ Yes, this ranking of a fake image like that as top result IS MISINFORMATION from Google. AWFUL. -L ------------------------------ Date: Sat, 23 Sep 2023 15:53:36 -0400 From: Jan Wolitzky <jan.wolitzky () gmail com> Subject: Misinformation research is buckling under GOP legal attacks (WashPost) Academics, universities and government agencies are overhauling or ending research programs designed to counter the spread of online misinformation amid a legal campaign from conservative politicians and activists who accuse them of colluding with tech companies to censor right-wing views. The escalating campaign =E2=80=94 led by Rep. Jim Jordan (R-Ohio) and other Republicans in Congress and state government =E2=80=94 has cast a pall over programs that study not just political falsehoods but also the quality of medical information online. Facing litigation, Stanford University officials are discussing how they can continue tracking election-related misinformation through the Election Integrity Partnership (EIP), a prominent consortium that flagged social media conspiracies about voting in 2020 and 2022, several participants told The Washington Post. The coalition of disinformation researchers may shrink and also may stop communicating with X and Facebook about their findings. The National Institutes of Health froze a $150 million program intended to advance the communication of medical information, citing regulatory and legal threats. Physicians told The Post that they had planned to use the grants to fund projects on noncontroversial topics such as nutritional guidelines and not just politically charged issues such as vaccinations that have been the focus of the conservative allegations. NIH officials sent a memo in July to some employees, warning them not to flag misleading social media posts to tech companies and to limit their communication with the public to answering medical questions. https://www.washingtonpost.com/technology/2023/09/23/online-misinformation-= jim-jordan/ ------------------------------ Date: Sat, 23 Sep 2023 15:50:44 -0400 From: Jan Wolitzky <jan.wolitzky () gmail com> Subject: Egyptian presidential hopeful targeted by Predator spyware (WashPost) A prominent Egyptian opposition politician who plans to challenge President Abdel Fatah El-Sisi in elections expected early next year was targeted with a previously unknown zero-day attack in an effort to infect his phone with Predator spyware, according to new research by Google and the University of Toronto's Citizen Lab. The discovery of the valuable zero-day exploit, designed to install Predator on iPhones running even the most up-to-date operating system, prompted Apple to push a security update to users on Thursday afternoon. Citizen Lab said it had *high confidence* that the Egyptian government was responsible for the failed hacking attempt. The effort targeted journalist and former member of parliament Ahmed Eltantawy and was first reported by Mada Masr, an independent Egyptian news organization. Eltantawy had been living briefly in Lebanon but moved back to Egypt in May. https://www.washingtonpost.com/investigations/2023/09/23/predator-egypt-hack-spyware-iphone/ ------------------------------ Date: Fri, 22 Sep 2023 23:40:44 -0400 From: Gabe Goldberg <gabe () gabegold com> Subject: It's 2030, and digital wallets have replaced every card in our purses and pockets (ZDNET) OpenWallet, now joined by Microsoft, looks to the near future when digital wallets replace traditional wallets in the same way debit cards replaced checkbooks. https://www.zdnet.com/finance/its-2030-and-digital-wallets-have-replaced-every-card-in-our-purses-and-pockets/ Nice basket you have there. Be a shame if anything happened to all those eggs. [Even worse, there's another Y2K+10x problem and the date on every transaction goes back 10 years and everyone is hosed. PGN] ------------------------------ Date: Thu, 21 Sep 2023 07:30:17 -0600 From: Matthew Kruk <mkrukg () gmail com> Subject: Google accused of directing motorist to drive off collapsed bridge (BBC) https://www.bbc.com/news/world-us-canada-66873982 *The family of a U.S. man who drowned after driving off a collapsed bridge are claiming that he died because Google failed to update its maps. * Philip Paxson's family are suing the company over his death, alleging that Google negligently failed to show the bridge had fallen nine years earlier. Mr Paxson died in September 2022 after attempting to drive over the damaged bridge in Hickory, North Carolina. A spokesperson for Google said the company was reviewing the allegations. ------------------------------ Date: Thu, 21 Sep 2023 23:44:55 -0400 From: Gabe Goldberg <gabe () gabegold com> Subject: Typeface trolls shaking down users of Adobe's font platform (BoingBoing) Do you use a font through Adobe's font platform? Is it Proxima Nova? Users of the typeface report being threatened by a foundry that claims to represent its creator, and Adobe isn't taking calls. The copyright troll business model, where lawyers demand money from people who know that proving their innocence would cost even more, has come to the land of fancy fonts. https://boingboing.net/2023/09/14/typeface-trolls-shaking-down-users-of-adobes-font-platform.html ------------------------------ Date: Fri, 22 Sep 2023 21:24:36 PDT From: Peter Neumann <neumann () csl sri com> Subject: Bitcoin conspiracy theory [Thanks to John Markoff.] Nic Carter doubles down on theory Bitcoin was invented by NSA https://cointelegraph.com/news/nic-carter-supports-bitcoin-invented-by-nsa-conspiracy-theory?utm_source=artifact ------------------------------ Date: Fri, 22 Sep 2023 12:58:34 +0300 From: Amos Shapir <amos083 () gmail com> Subject: Re: Pedestrian dies after Cruise cars block ambulance (RISKS-33.85) It seems that autonomous cars have stumbled upon what every beginner driver should realize: That the hardest part of driving is not operating a vehicle, but dealing with other drivers; and that the hardest part of that, is trying to decipher their intentions and acting accordingly. Robotic systems are notoriously bad at recognizing human intentions -- a lot of context is required, which is often implicit, implied, or even virtual. ------------------------------ Date: 20 Sep 2023 15:21:10 -0700 From: "John Levine" <johnl () iecc com> Subject: Re: Pedestrian dies after Cruise cars block ambulance (RISKS-33.85) I've seen a lot of arguments about whether self-driving cars are more or less safe than human drivers, based on accidents or deaths per mile driven. They probably are, although robocars haven't get driven enough for the numbers to be meaningful. But we're only now coming to grips with the fact that when they fail, they often do so in ways very unlike the ways that human drivers fail. They may be very good at stopping for red lights, but they are a lot less good at going through a red light to get out of the way of an ambulance. ------------------------------ Date: Sat, 1 Jul 2023 11:11:11 -0800 From: RISKS-request () csl sri com Subject: Abridged info on RISKS (comp.risks) The ACM RISKS Forum is a MODERATED digest. Its Usenet manifestation is comp.risks, the feed for which is donated by panix.com as of June 2011. => SUBSCRIPTIONS: The mailman Web interface can be used directly to subscribe and unsubscribe: http://mls.csl.sri.com/mailman/listinfo/risks => SUBMISSIONS: to risks () CSL sri com with meaningful SUBJECT: line that includes the string `notsp'. Otherwise your message may not be read. *** This attention-string has never changed, but might if spammers use it. => SPAM challenge-responses will not be honored. Instead, use an alternative address from which you never send mail where the address becomes public! => The complete INFO file (submissions, default disclaimers, archive sites, copyright policy, etc.) has moved to the ftp.sri.com site: <risksinfo.html>. *** Contributors are assumed to have read the full info file for guidelines! => OFFICIAL ARCHIVES: http://www.risks.org takes you to Lindsay Marshall's searchable html archive at newcastle: http://catless.ncl.ac.uk/Risks/VL.IS --> VoLume, ISsue. Also, ftp://ftp.sri.com/risks for the current volume/previous directories or ftp://ftp.sri.com/VL/risks-VL.IS for previous VoLume If none of those work for you, the most recent issue is always at http://www.csl.sri.com/users/risko/risks.txt, and index at /risks-33.00 ALTERNATIVE ARCHIVES: http://seclists.org/risks/ (only since mid-2001) *** NOTE: If a cited URL fails, we do not try to update them. Try browsing on the keywords in the subject line or cited article leads. Apologies for what Office365 and SafeLinks may have done to URLs. ==> Special Offer to Join ACM for readers of the ACM RISKS Forum: <http://www.acm.org/joinacm1> ------------------------------ End of RISKS-FORUM Digest 33.86 ************************
Current thread:
- Risks Digest 33.86 RISKS List Owner (Sep 23)