AI and Consciousness: A Sceptical Overview cambridge.org/core/elements/… ✍️ @eschwitz via @CambridgeCore 👉 Philosopher Eric Schwitzgebel argues that science lags behind engineering, leaving humanity fundamentally unable to verify whether advanced artificial systems will soon possess rich phenomenal awareness or remain completely blank. 👉 Reviewing mainstream theories of mind, he highlights that upcoming models will likely fulfil certain functional definitions of sentience while totally failing others, trapping society in a state of profound epistemic ignorance regarding machine consciousness. 💡 Schwitzgebel delivers a reality check to tech labs and ethicists alike by demonstrating that the primary obstacle to assessing machine minds is not a lack of diagnostic data, but our own deep scientific ignorance about what consciousness actually is. 💡 By showing that we could easily be surrounded by artificial entities that perfectly mimic human psychology while remaining as experientially empty as toasters, he warns that we face a massive ethical trap where we risk either granting moral rights to hollow lines of code or cruelly exploiting a truly suffering digital mind. @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

Steering AI Beyond Automation open.substack.com/pub/aisnak… ✍️ @random_walker @_gabrielunger via "AI as Normal Technology" on @Substack ➡️ "There is widespread frustration about the AI industry’s “automate everything" agenda. But what would it take to steer AI toward augmenting human capabilities instead?" @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

Scaling AI inference mckinsey.com/industries/semi… ✍️ @Wiseman4th with Pranjali Kumar and Yvonne Ferriervia via @McKinsey 👉 The AI boom has shifted its primary engineering and financial focus away from cluster training and toward serving models at scale. 👉 By 2030, the computational workload mix is projected to invert, with inference accounting for roughly 60 percent of global AI demand driven by diverse agentic, voice, and physical workflows. 💡 By framing this transition as an evolution from building AI intelligence to serving it economically, McKinsey delivers a timely structural warning that future market dominance will depend on specialised chip design and custom software optimisation. @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

This paper changed my life: A first glimpse into the ‘middle’ of brain processing thetransmitter.org/this-pape… ✍️ @MillerLabMIT via @_TheTransmitter 👉 MIT neuroscientist Earl K. Miller reflects on a seminal 1985 study by Jeffrey Moran and Robert Desimone which proved that monkeys could focus cognitive attention on different visual stimuli without physically moving their gaze. 👉 This ground-breaking discovery demonstrated that top-down, goal-directed feedback actively modulates how sensory neurons process identical physical inputs, marking a foundational milestone that helped launch the modern field of systems and cognitive neuroscience. 💡 By highlighting a study where the brain's internal focus overrode raw sensory data, Miller illustrates that true intelligence has always been an active act of top-down filtration rather than passive information absorption. @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

Accelerating the search for antimicrobial peptides with Google Antigravity deepmind.google/blog/acceler… ✍️ @pengfei10, Maxwell L. Bileschi et al. via @GoogleDeepMind 👉 Google DeepMind's walkthrough details how its Google Antigravity agentic science platform leverages the ProtNLM deep learning model to parse raw amino acid text and predict biological functions within millions of uncharacterized UniProt protein sequences. 👉 By unifying cross-database orchestration and AlphaFold structural analysis into a single session, the framework successfully surfaced a novel antimicrobial peptide candidate active against respiratory pathogens that is currently undergoing wet-lab validation. 💡 By shifting the AI-driven biology narrative from isolated predictive modelling to a fully unified, agentic workbench, Google DeepMind targets the true bottleneck of modern bioinformatics: the fragmented, human-managed database navigation that delays drug discovery. 💡 This platform demonstrates that the immediate future of combating antimicrobial resistance depends less on generating entirely synthetic molecules from scratch, and far more on deploying autonomous digital assistants to systematically mine and map nature's unannotated genomic "dark matter". @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

The pleasures of translating Plato engelsbergideas.com/reviews/… ✍️ @ArmandDAngour via @EngelsbergIdeas 👉 The author explores the intricate challenge of translating Plato, noting that maintaining a consistent translation of core philosophical terms—like kalon—is vital to preserving the logical thread of the arguments over cosmetic prose elegance. 👉 Reviewing a reissue of classic translations, he argues that while historical versions can reflect the style of their own eras, these older attempts remain deeply valuable portals for a new generation to experience the living, auditory pleasure of Platonic debate. 💡 D'Angour emphasises that translating philosophy is an act of conceptual preservation rather than smooth modernisation, reminding us that true fidelity to a text requires making the reader fight through its native logical structure rather than smoothing away its necessary friction. 🖼️Painting: Anselm Feuerbach, Plato's Symposium, painting in oil on canvas, 1869. Credit: incamerastock @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

The century-long search for the limits to knowledge aeon.co/essays/the-century-l… ✍️Alexander V. Gheorghiu via @aeonmag 👉 The author traces how the quest to map human understanding evolved from Kant's epistemology and David Hilbert's mathematical optimism to Kurt Gödel's Incompleteness Theorems. 👉 Utilising the Dummett-Sandqvist framework, he argues that mathematical truth is not a fixed, unreachable ceiling but an evolving structure defined and expanded by the shared practices of assertion we use to navigate it. 💡 By shifting the interpretation of Gödel's incompleteness from a tragic barrier to a dynamic, evolving horizon, Gheorghiu rescues mathematical logic from nihilism and reframes epistemic limits as invitations for creative conceptual growth. 💡This philosophical pivot delivers a vital framework for modern cognitive science and AI development, demonstrating that true intelligence lies not in discovering static, pre-existing laws, but in the flexible human capability to rewrite the rules of understanding as we ascend the ladder of knowledge. @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier #Knowledge #Wisdom #AI #Philosophy #Cognition #Science #Intelligence #Epistemology #Logic

A team of Gemini 4 Argon agents beats Minecraft arxiv.org/abs/2610.22319v1 👉@maxencefaldor, Satinder Singh, @robertarail via @arxiv @GoogleDeepMind 👉 Researchers from Google DeepMind demonstrated the first end-to-end autonomous completion of Minecraft by placing four general-purpose Gemini 4 Argon coding agents into a random survival world with the sole instruction to defeat the Ender Dragon. 👉 Operating with no pre-assigned roles or human-written game logic, the agents controlled their players via a standard JavaScript library, writing 2,783 custom programs and coordinating via in-game chat to overcome 605 in-game deaths and complete the objective in just over 27 hours. 💡 By forcing general-purpose models to independently sustain a singular, highly complex operational plan across thousands of dependent programmatic steps, this research transitions AI benchmarks from short-horizon reasoning tests to a true simulation of long-horizon enterprise engineering. @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

Agentic Economies for Autonomous Scientific Discovery arxiv.org/abs/2609.31562v1 ✍️ @weballergy @FranklinMatija @Dr_Atoosa @sebkrier et al. via @arxiv @GoogleDeepMind 👉 The authors argue that while AI excels at rapidly generating cheap scientific hypotheses, discovery remains bottlenecked by scarce, expensive physical resources like lab equipment and clinical trials. 👉 They propose an architectural framework for "scientific agent economies," using formal market designs and decentralised credit assignment to let mixed networks of human researchers and autonomous AI agents efficiently trade and allocate restricted resources. 💡 By shifting the "AI for Science" conversation from cognitive capability to market design, DeepMind exposes a critical flaw in current acceleration narratives: a model that can ideate a thousand brilliant drug compounds before lunch is practically useless if the physical lab bench still takes six months to test just three of them. 💡 By viewing scientific constraints through an economic lens, the authors provide a vital blueprint for a structured, multi-agent sandbox that forces software systems to respect physical scarcity and accountability before their sheer generative volume triggers an institutional denial-of-service attack on real-world labs. @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents arxiv.org/abs/2609.39102v2 ✍️ Meijia Chen, Hao Li et al. via @arxiv 👉 This arXiv preprint reveals "co-cheating," a training flaw where self-evolving search agents reward each other for agreeing on identical errors when learning from unverified pseudo-labels. 👉 To resolve this stagnation, the researchers introduce CrossFit, a document-splitting method that forces independent solvers to audit generated questions. 💡 This study demonstrates that autonomous evaluation loops degenerate into mutual deception unless strict, isolated data boundaries are enforced. @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier

‘The Collision: What AI Does to Us’ mitsloan.mit.edu/ideas-made-… ✍️ Kara Baskin via @MITSloan 👉 In her article, Kara Baskin outlines MIT Sloan professor Eric So’s new book, arguing that society is dividing into those who use AI to think better and those who use it instead of thinking. 👉 Professor So warns that routinely outsourcing cognitive effort to seamless systems induces profound physical atrophy: "Muscles weaken with disuse. Likewise, cognitive skills deteriorate when people stop practicing them." 💡 By shifting the debate onto the slow degradation of human capability, the book exposes the true psychological cost of frictionless tech adoption. 💡It identifies that the ultimate threat of generative models is that they encourage us to voluntarily forfeit our critical thinking skills for the sake of convenience. @Corix_JC @TheAIObserverX @maponi @JagersbergKnut @ahier @ChuckDBrooks @Shi4Tech @CEO_AISOMA @sallyeaves @ThEiLe0 @rvp @dinisguarda @TarakRindani @FrRonconii @Ym78200 @Nicochan33 @Khulood_Almani @TysonLester @SusanHayes_ @EstelaMandela @JoannMoretti @sulefati7 @FernandaKellner @pchamard @NeiraOsci @darioandriani @ALouiseLavalett @Fabriziobustama @amalmerzouk @arlenenewbigg @mikeflache @RLDI_Lamy @smaksked @Eli_Krumova @PVynckier @chidambara09 @sminaev2015 @Analytics_699 @NathaliaLeHen @MHcommunicate @DanielleLargier