675 shaares
"This RFC proposes a moderation policy: it does not mandate technical choices or decisions, but it tries to provide a common way for people to collaborate together in a shared space, which is the most important aspect of the GNOME project.
The goal of this RFC is not to catch violations; instead, the goal is to remove plausible deniability, and to force a choice between following the policy and intentionally violating it."
The goal of this RFC is not to catch violations; instead, the goal is to remove plausible deniability, and to force a choice between following the policy and intentionally violating it."
"En résumé, le choix atypique d’Anthropic d’entrer en Bourse dès maintenant révèle un changement profond de la perception du rapport rendement-risque. L’objectif est de faire reposer un ensemble de risques convergents sur les plans d’épargne-retraite des citoyens ordinaires, sur les investisseurs institutionnels, sur les Etats et la société dans son ensemble, avant que l’illusion collective autour de l’IA ne vole en éclats."
"Companies like Scale AI first outsourced data work to wherever labor was cheapest. Now, they are recruiting physicians, lawyers, and Ph.D.s to generate and annotate data, rank model answers, and verify outputs. Higher credentials can raise the advertised rate, but the platform still controls payment and working conditions."
Kristen Maevyn wrote an essay named The Decoherence Trap which talks about a problem where the individual parts of a system may look fine in isolation, but don’t make sense as part of a larger system.
"Decoherence: the accumulation of locally valid but mutually incompatible or redundant models within a system, such that the cost of change is dominated by translating among those models and preventing unintended side-effects."
"Decoherence: the accumulation of locally valid but mutually incompatible or redundant models within a system, such that the cost of change is dominated by translating among those models and preventing unintended side-effects."
Cf. Accelerando de Charlie Stross !
“Just as industrial agriculture replaced the diversity of the living soil with chemical fertilizers and crop monocultures — productive in the short term, devastating over time — the current AI moment is producing an epistemic monoculture.”
As prominent conversational AI providers like OpenAI adopt advertising-based business models, traditional web and mobile tracking practices are expanding into conversational AI services [15].
However, despite their growing adoption, the tracking, data-sharing, and monetization practices of conversational AI services remain largely opaque and have received comparatively limited scrutiny from researchers, regulators, and the public.
However, despite their growing adoption, the tracking, data-sharing, and monetization practices of conversational AI services remain largely opaque and have received comparatively limited scrutiny from researchers, regulators, and the public.
As Artificial Intelligence (AI) is increasingly promoted and used in qualitative research, it also raises profound methodological issues. This position paper critically interrogates the role of generative AI (genAI) in the context of qualitative coding methodologies. Despite widespread hype and claims of efficiency, we propose that genAI is not methodologically valid within qualitative inquiries, and its use risks undermining the robustness and trustworthiness of qualitative research. The lack of meaningful documentation, commercial opacity, and the inherent tendencies of genAI systems to produce incorrect outputs all contribute to weakening methodological rigor. Overall, the balance between risk and benefits does not support the use of genAI in qualitative research, and our position paper cautions researchers to put sound methodology before technological novelty.
Thierry Ménissier
"While AI coding tools boost security development speed by 17%, false positives from scanning tools increase workloads by 30%. Critically, 3 in 5 teams now ship code faster than security reviews can keep up, and 82% of practitioners warn that overreliance on AI tools is weakening developers’ security skills. Yet, preparedness is possible: teams establishing dedicated AI security roles and formal policies report significantly higher readiness across emerging threats."
"LLMs are the perfect technology for manipulating the medical literature to say nearly anything you want via systematic review methods"