Skip to content

Piero Bosio Social Web Site Personale Logo Fediverso

Social Forum federato con il resto del mondo. Non contano le istanze, contano le persone
  1. Home
  2. Categories
  3. Uncategorized
  4. lol, "if only someone had warned us about this sort of thing?!"

lol, "if only someone had warned us about this sort of thing?!"

Uncategorized
42 31 3
Log in to reply

Feed RSS
lol, "if only someone had warned us about this sort of thing?!"

Gli ultimi otto messaggi ricevuti dalla Federazione
@pierobosio@soc.bosio.info
Post suggeriti
  • wikidata@wikis.worldundefined

    ⏰ Reminder!

    Uncategorized wikidata factchecking openknowledge llm genai workshop
    1
    1
    0 Votes
    1 Posts
    8 Views
    wikidata@wikis.worldundefined
    ⏰ Reminder! Fact-checking with Wikidata workshop📆 20 January 2026 | 🕟 16:30–18:00 (UTC+1)Philippe Saadé (Wikimedia Germany) and @DataTalks.Club host a hands-on workshop using the Wikidata Model Context Protocol (MCP) for fact-checking beyond generative AI.✔ Wikidata intro✔ Fact-checking with MCP, Large Language Models (LLM) & semantic search✔ Build a small fact-checking pipeline👉 Register now: https://luma.com/7fs5v7os#Wikidata #FactChecking #OpenKnowledge #LLM #GenAI #Workshop
  • gert@poliversity.itundefined

    Language models cannot reliably distinguish belief from knowledge and fact

    Uncategorized llms epistemology knowledge
    1
    0 Votes
    1 Posts
    8 Views
    gert@poliversity.itundefined
    Language models cannot reliably distinguish belief from knowledge and factAbstract-----------«As language models (LMs) increasingly infiltrate into high-stakes domains such as law, medicine, journalism and science, their ability to distinguish belief from knowledge, and fact from fiction, becomes imperative. Failure to make such distinctions can mislead diagnoses, distort judicial judgments and amplify misinformation. Here we evaluate 24 cutting-edge LMs using a new KaBLE benchmark of 13,000 questions across 13 epistemic tasks. Our findings reveal crucial limitations. In particular, all models tested systematically fail to acknowledge first-person false beliefs, with GPT-4o dropping from 98.2% to 64.4% accuracy and DeepSeek R1 plummeting from over 90% to 14.4%. Further, models process third-person false beliefs with substantially higher accuracy (95% for newer models; 79% for older ones) than first-person false beliefs (62.6% for newer; 52.5% for older), revealing a troubling attribution bias. We also find that, while recent models show competence in recursive knowledge tasks, they still rely on inconsistent reasoning strategies, suggesting superficial pattern matching rather than robust epistemic understanding. Most models lack a robust understanding of the factive nature of knowledge, that knowledge inherently requires truth. These limitations necessitate urgent improvements before deploying LMs in high-stakes domains where epistemic distinctions are crucial.»#ai #LLMs #epistemology #knowledgehttps://www.nature.com/articles/s42256-025-01113-8
  • remixtures@tldr.nettime.orgundefined

    "By comparison, at least, the way the Chinese government speaks about AI is more modest.

    Uncategorized usa china aihype aibubble
    1
    0 Votes
    1 Posts
    13 Views
    remixtures@tldr.nettime.orgundefined
    "By comparison, at least, the way the Chinese government speaks about AI is more modest. Yes, China’s economic leadership views AI as a priority and has boldly claimed it seeks to lead the world by 2030. Yet the rhetoric lacks the eschatological tone common in Silicon Valley. Chinese economic planners appear more interested in AI as a tool for industrial processes than as a means of creating a superintelligence that will reach the singularity. The State Council’s 2025 “AI+” initiative is focused entirely on efficiency-enhancing applications rather than intelligence explosions.There is another important difference. China is banking far more heavily on simpler, lower-cost open-source AI models. In the US, most of the leading “frontier” AI models are secret and proprietary, in part as a business model and in part due to the apocryphal fears that the wrong actors could trigger human extinction. The smaller, lower-cost Chinese models may be seeking, in that sense, to be the more nimble 1970s Toyota rivals to the giant American cars produced by General Motors.More importantly, China is hedging its bets by investing heavily in a wide range of other technologies that might reasonably be described as “the future”. In 2024, the country invested an estimated $940bn in clean-energy capex, broadly defined as renewables, electricity grids and energy storage (batteries), dwarfing its AI investments. In these sectors, AI is meant to be a complement — the glue rather than the structure.While China’s overall economy remains weaker than it was in the 2010s, elements of this broader strategy seem to be bearing fruit."https://www.ft.com/content/12581344-6e37-45a0-a9d5-e3d6a9f8d9ba#USA #China #AI #AIHype #AIBubble
  • ricardo@mastodon.bsd.cafeundefined

    Remembering the Olivetti Programma 101, 25/10/1965

    Uncategorized history technology computers
    1
    1
    0 Votes
    1 Posts
    10 Views
    ricardo@mastodon.bsd.cafeundefined
    Remembering the Olivetti Programma 101, 25/10/1965 #history #technology #computers