@xuv writes

Tag: AI

  • Quoting Ellsworth Toohey

    AI systems learned to code by devouring Stack Overflow’s archive of human-solved problems. Now that developers ask ChatGPT instead of posting publicly, solutions exist only in private chat histories. They’re not indexed, not peer-reviewed, not available to the next person with the same bug. We’re trading a searchable knowledge commons for ephemeral conversations with a…

  • Quoting Max Riesenhuber

    The problem is that, in LLMs, words (symbols) are not grounded in experiences of the real world, so any meaning of words needs to be inferred from the relations of words to each other, floating in some abstract space and being prone to misinterpretation, let alone hallucinations. Max Riesenhuber is the co-director of Georgetown’s Center…

  • Citation de Grégory Chatonsky

    Qu’est-ce qu’un espace latent ? Techniquement, il s’agit d’une représentation mathématique intermédiaire, située entre les couches d’entrée et de sortie d’un réseau de neurones, où les données sont encodées sous forme de vecteurs statistiques dans un espace de haute dimensionnalité, des centaines, des milliers, parfois des millions de dimensions. Contrairement aux espaces de représentation traditionnels (pixels…

  • Quoting Cory Doctorow

    What the gadget does is nowhere near as important as who the gadget does it for and who it does it to. Your car can use a cutting-edge computer vision system to alert you when you’re drifting out of your lane – or it can use that same system to narc you out to your insurer so they…

  • Citation de Juan Sebastián Carbonell

    Une tech­no­lo­gie ne s’impose pas iné­luc­ta­ble­ment parce qu’elle est plus “effi­ciente”, mais parce que des acteurs qui contrô­lent les res­sour­ces économiques, consi­dé­rant qu’elle cor­res­pond mieux à leurs inté­rêts, déci­dent de lui donner réa­lité. — Juan Sebastián Carbonell Via Rezo.net qui reprend un article de Mediapart sur le dernier livre, intitulé “Taylorisme augmenté”, par l’auteur.

  • Citando Marta Peirano

    Los modelos viven encerrados en un mundo de palabras, un universo wittgensteniano compuesto exclusivamente de signos, jerarquías y estadística. No pueden mentir, alucinar o equivocarse porque todas las palabras son verdaderas cuando dejan de describir el mundo. No habrá inteligencia artificial general

  • Quoting Janelle Shane

    I wish more people understood that any “apology” or “self-reflection” from chatbots are meaningless – they’re just continuing with your improv session. ChatGPT will apologize for anything – by @janellecshane

  • Quoting Nick McGreivy

    To be clear, AI can drive scientific breakthroughs. My concern is about their magnitude and frequency. Has AI really shown enough potential to justify such a massive shift in talent, training, time, and money away from existing research directions and towards a single paradigm?

  • Quoting Benedict Evans

    It might also be that the chatbot as chatbot is the right UX only for some people and some use-cases, and most people will experience this technology as features and capabilities wrapped inside other things.

  • Quoting Dave Karpf

    Inside every Silicon Valley company are two businesses. One is a business that attempts to sell products for more than it costs to make them. The other is a speculative stock, trafficking in vibes. So long as the latter doesn’t overshadow the former, that can be just fine. But (again ell oh ell), selling actual…

  • Paper surgery

    This early 2000 work by Stephen J Shanabrook, in collaboration with Veronika Georgieva, feels like GenAI artifacts reproduced in analog form.

  • Link dump from MDRAI

    Tonight I attended an event at Elisava organized by the Master in Design for Responsible AI. I gathered a few links from the presentations by Caroline Sinders and Leandro Ucciferri. I’m just going to dump them below to explore later.