Ohio State University: ChatGPT often won’t defend its answers – even when it is right

Ohio State University: ChatGPT often won’t defend its answers – even when it is right. “A team at The Ohio State University challenged large language models (LLMs) like ChatGPT to a variety of debate-like conversations in which a user pushed back when the chatbot presented a correct answer. Through experimenting with a broad range of reasoning puzzles including math, common sense and logic, the study found that when presented with a challenge, the model was often unable to defend its correct beliefs, and instead blindly believed invalid arguments made by the user.”

TechCrunch: Early impressions of Google’s Gemini aren’t great

TechCrunch: Early impressions of Google’s Gemini aren’t great. “This week, Google took the wraps off of Gemini, its new flagship generative AI model meant to power a range of products and services including Bard, Google’s ChatGPT competitor. In blog posts and press materials, Google touted Gemini’s superior architecture and capabilities, claiming that the model meets or exceeds the performance of other leading gen AI models like OpenAI’s GPT-4. But the anecdotal evidence suggests otherwise.”

NiemanLab: The press adopts a new level of transparency around images

NiemanLab: The press adopts a new level of transparency around images. “The press has often been light on contextual information and details about the images they use. Typically, publications only provide the reader with a tiny gray caption, perhaps with a name and maybe some context related to its use or production method or where it was found, such as ‘illustration,’ ‘archival photo,’ ‘photo,’ or ‘social media.’ … A newfound level of transparency around images could be vital in educating the press and the public about images and their credibility.”

Business Insider: Elon Musk’s biggest fans on X love Community Notes — until it comes for them

Business Insider: Elon Musk’s biggest fans on X love Community Notes — until it comes for them. “Some may argue that the fact that no X user, including Musk himself, is immune to a correction from Community Notes is a testament to the feature’s capability, but critics have repeatedly pointed out the flaws of the system behind the feature.”

The Conversation: Disinformation is rampant on social media – a social psychologist explains the tactics used against you

The Conversation: Disinformation is rampant on social media – a social psychologist explains the tactics used against you. “Foreign governments, internet trolls, domestic and international extremists, opportunistic profiteers and even paid disinformation agencies exploit the internet to spread questionable content. Periods of civil unrest, natural disasters, health crises and wars trigger anxiety and the hunt for information, which disinformation agents take advantage of. Certainly it’s worth watching for the warning signs for misinformation and dangerous speech, but there are additional tactics disinformation agents employ.”

Deepfake row: Govt advisory for social media companies in two days (New Indian Express)

The New Indian Express: Deepfake row: Govt advisory for social media companies in two days. “The government will issue advisories to social media intermediaries such as Meta and X (formerly known as Twitter) in the next two days to ensure compliance on addressing deepfake issues, Minister of State for Electronics and IT Rajeev Chandrasekhar said on Tuesday.”

Platformer: Amazon’s Q has ‘severe hallucinations’ and leaks confidential data in public preview, employees warn

Platformer: Amazon’s Q has ‘severe hallucinations’ and leaks confidential data in public preview, employees warn. “Three days after Amazon announced its AI chatbot Q, some employees are sounding alarms about accuracy and privacy issues. Q is ‘experiencing severe hallucinations and leaking confidential data,’ including the location of AWS data centers, internal discount programs, and unreleased features, according to leaked documents obtained by Platformer. “

Business Insider: An image of Israel Kamakawiwoʻole shows Google search still can’t tell AI-generated pictures apart from genuine ones

Business Insider: An image of Israel Kamakawiwoʻole shows Google search still can’t tell AI-generated pictures apart from genuine ones. “If you’re struggling to differentiate AI-generated images from real ones, you’re not alone. An AI-generated image of the late Hawaiian singer Israel Kamakawiwoʻole is currently showing up as the top search result on Google when you search his name. Ethan Mollick, a professor at the University of Pennsylvania’s Wharton School, spotted the change and shared a screenshot of it on X on Sunday.”

New Indian Express: Fake letters on social media keep leaders on tenterhooks

New Indian Express: Fake letters on social media keep leaders on tenterhooks. “Fake news, in the form of letters, being spread on social media platforms is not only creating confusion in political circles but also causing a big headache for the leaders.”

University of Arkansas Little Rock: UA Little Rock Receives $5 Million From U.S. Army To Combat Adversarial Information Campaigns

University of Arkansas Little Rock: UA Little Rock Receives $5 Million From U.S. Army To Combat Adversarial Information Campaigns. “The project, set to run through 2025, aims to identify research gaps in deviant socio-technical behaviors, shape an agenda focused on developing strategies that can counter emerging threats, and create tools for near real-time analysis of such threats.”

The political economy of social media: A new eBook (Centre for Economic Policy Research)

Centre for Economic Policy Research: The political economy of social media: A new eBook. “The emergence of social media has reshaped the way humans communicate, interact and coordinate with each other. Assessing the impact of this transformation on politics has been one of the great social science questions of the last or decade or so, and will continue to occupy researchers for a long time to come. A new CEPR eBook provides a snapshot of how economists have been trying to answer this question.”

‘Sombre occasion’: X reprimanded after Voice referendum (Sydney Morning Herald)

Sydney Morning Herald: ‘Sombre occasion’: X reprimanded after Voice referendum. “X has had its status as a signature to the Australian code of practice on disinformation and misinformation revoked following a complaint that it didn’t allow the reporting of misinformation during the Voice to parliament referendum.”

Hard Drives, YouTube, and Murder: India’s Dark History of Digital Hate (WIRED)

WIRED: Hard Drives, YouTube, and Murder: India’s Dark History of Digital Hate. “Today more than half the population of India—759 million people—are online. The country has 467 million active YouTube users—the most in the world. The users are no longer predominantly urban. Nobody has tapped into this proliferation better than right-wing groups dedicated to fostering communal disharmony, moving from hard disks filled with videos and laptops in temples to the vast reach of YouTube and WhatsApp.”

Nature: ChatGPT generates fake data set to support scientific hypothesis

Nature: ChatGPT generates fake data set to support scientific hypothesis. “In a paper published in JAMA Ophthalmology on 9 November1, the authors used GPT-4 — the latest version of the large language model on which ChatGPT runs — paired with Advanced Data Analysis (ADA), a model that incorporates the programming language Python and can perform statistical analysis and create data visualizations. The AI-generated data compared the outcomes of two surgical procedures and indicated — wrongly — that one treatment is better than the other.”

Cornell Chronicle: Crowdsourced fact-checking fights misinformation in Taiwan

Cornell Chronicle: Crowdsourced fact-checking fights misinformation in Taiwan. “In a new study, Andy Zhao, a doctoral candidate in information science based at Cornell Tech, compared professional fact-checking articles to posts on Cofacts, a community-sourced fact-checking platform in Taiwan. He found that the crowdsourced site often responded to queries more rapidly than professionals and handled a different range of issues across platforms.”