Microsoft's AI image creation tool and news aggregation service, MSN.com, have faced controversy and backlash due to the inappropriate and damaging content generated by their AI systems. In one instance, Microsoft's Bing AI's Image Creator produced images that didn't make much sense, leading to concerns about the effectiveness and reliability of the tool [1edebfc2]. In another incident, Microsoft's AI-generated poll published on MSN.com caused significant reputational damage to The Guardian. The poll speculated on the cause of a woman's death and received backlash from readers, prompting accusations against Microsoft for damaging The Guardian's reputation [ab30102d]. The incidents raise questions about the responsibility and accountability of AI systems in content creation and news aggregation, as well as the need for proper vetting and oversight to prevent similar incidents in the future. The discussions surrounding these incidents also highlight the challenges and risks associated with AI-generated content and the importance of responsible deployment and regulation.
A survey conducted by Newsworks found that the spread of misinformation and fake news is the top concern with AI, followed by AI's lack of human creativity and judgment, and the potential loss of jobs [a2ca2964]. Three-quarters of MPs agree that trusted journalism is critical in minimizing the risk of misinformation. Newsbrand editors surveyed also agreed that the risk to the public from AI-generated misinformation is greater than ever before. The public values content produced by humans and favors guidelines or regulations for AI-generated content on the web. The need for regulation around clearly declaring when and how AI is used to create online content was also highlighted. Newsworks CEO Jo Allan emphasized the important role of journalists and newsbrands in our democratic society. The article also mentions the importance of protecting publishers' content from being exploited by AI companies.
These incidents and public concerns shed light on the consequences of AI-generated content on reputation and the potential for spreading misinformation. They underscore the need for responsible deployment, oversight, and regulation of AI systems in content creation and news aggregation to ensure accuracy, ethics, and accountability. The discussions around these issues also emphasize the important role of trusted journalism and the value of human-generated content in combating misinformation and maintaining public trust.
Researchers at Cornell University have developed an algorithm called SneakyPrompt that can bypass generative AI censorship and generate prohibited images. The algorithm discovered that certain sequences of characters, which mean nothing, are interpreted clearly by AI and can be used to push the AI to generate explicit images. The researchers emphasize that these discoveries should be used to alert the creators of these AI systems about the flaw in their systems, rather than to circumvent security measures. The algorithm is unlikely to be released to the general public [49a9c60e].
This article discusses a dedicated Instagram page called @hardaipics that shares AI-generated images. The page has gained a large following in a short period of time. The article highlights the creativity and advancements of AI in visualizations. It also emphasizes the collaboration between humans and AI, showcasing how AI can assist in various fields such as design, music, law enforcement, and healthcare. The article addresses the limitations of AI compared to human creativity and explores the potential risks of AI. It concludes by encouraging readers to appreciate and engage with AI-generated content [ebed7581].
Researchers have developed an AI algorithm that can create realistic fake faces that are difficult to distinguish from real ones. The algorithm generates random faces with blurred backgrounds and serious expressions, tricking the human brain into perceiving them as genuine portraits. The study found that people had difficulty differentiating between AI-generated faces and real ones, with a particular correlation between attractiveness and the ability to detect fakes. This blurring of the line between real and fake on the internet is causing concern among internet users. In other news, several mobile phones with impressive features and affordable prices are being highlighted, including a phone with a 100 MP camera, 256 GB of memory, and 5G connectivity for 229 euros, and another phone with 12 GB of RAM and 256 GB of memory for 350 euros. Additionally, a mid-range phone with 16 GB of RAM is praised for its excellent performance and affordability. Lastly, a popular air fryer with a 6-liter capacity is being sold for less than 70 euros. [55f88616]
Artists and computer scientists are using a tool called Nightshade to protect their copyrighted images from being replicated by AI models. Nightshade alters images in small ways that are imperceptible to humans but dramatically different to AI platforms. By poisoning the AI models with visions of cats, Nightshade tricks them into seeing something completely different. The tool is free to use and aims to render some narrower applications unusable enough to force companies to pay up when scraping artists' work. While Nightshade represents a revolt against AI models, experts believe that AI developers will likely patch their programs to defend against such countermeasures [363f6a4a].
A new report reveals that popular AI image-generators contain thousands of images of child sexual abuse. The report highlights that these images are being used to train AI systems, resulting in the production of explicit imagery of fake children and the transformation of social media photos of real teens into nudes. The Stanford Internet Observatory found over 3,200 images of suspected child sexual abuse in the AI database LAION, which is used to train leading AI image-makers. LAION has temporarily removed its datasets in response to the report. The report calls for companies to take action to address this harmful flaw in AI technology [336bd195].
Nightshade, an AI poisoning tool created by researchers at the University of Chicago, received 250,000 downloads in the first five days of its release. The tool is designed to disrupt AI models scraping and training on artworks without consent. The leader of the project, Ben Zhao, expressed surprise at the high enthusiasm and overwhelming response. Nightshade works by altering artworks on a pixel level to appear as different content to machine learning algorithms. The creators of Nightshade also developed Glaze, a tool that prevents AI models from learning an artist's signature style. Glaze has received 2.2 million downloads since its release. The Glaze/Nightshade team plans to release a combined version of the tools in the future. An open-source version of Nightshade may also be developed. The creators have not heard directly from model makers behind AI image generating technology. The downloads of Nightshade come from all over the globe, indicating a broad user base [7272edcb].