SciTech Brief Titelbild

SciTech Brief

SciTech Brief

Von: @LocutorCo
Jetzt kostenlos hören, ohne Abo

Daily science, technology and digital culture — in minutes.@LocutorCo Science Fiction Täglich Wissenschaft
  • IBM’s Atomic Leap: The World’s First Sub-1nm Chip and the Future of AI Power
    Jun 26 2026
    Is Moore’s Law finally back from the dead? On June 25, 2026, IBM unveiled a landmark semiconductor breakthrough: the world’s first sub-1 nanometer (0.7 nm) chip technology. In today’s episode, we explore the revolutionary "nanostack" architecture, a 3D vertical design that packs nearly 100 billion transistors—double the density of previous generations—onto a chip the size of a fingernail.

    We break down the staggering performance projections: up to 50% more power or a massive 70% increase in energy efficiency. We analyze why this is a "holy grail" for the generative AI era, where data centers are currently straining power grids and cooling systems. We also dive into the technical details of the 7 angstrom node, the 40% improvement in SRAM scaling, and the high-stakes roadmap that aims for mass production within the next five years. Join us as we discuss how IBM is reinventing the transistor to keep the digital revolution alive for another decade.

    This episode includes AI-generated content.
    Mehr anzeigen Weniger anzeigen
    2 Min.
  • Alibaba stole Claude triggering US export bans
    Jun 25 2026
    The Great AI Heist: Anthropic vs. Alibaba and the War for Claude’s "Secret Sauce"

    Did China just "clone" the world's most advanced AI at a fraction of the cost? In today’s episode, we investigate the explosive letter sent by Anthropic to U.S. officials on June 24, 2026, accusing the Chinese giant Alibaba of a massive, industrial-scale "distillation attack". We break down the data: 28.8 million exchanges using nearly 25,000 fraudulent accounts designed to extract Claude’s prized capabilities in software engineering and agentic reasoning.

    We dive into the geopolitical fallout, including Alibaba’s plummeting stock and its place on the Pentagon’s blacklist. We also analyze the "adversarial distillation" technique that allows rival labs to bypass years of R&D, and the Trump administration's controversial decision to lock down Anthropic’s own models, Mythos and Fable, citing national security risks. Join us as we explore whether the era of open AI borders is officially over and what this means for the $1 trillion race to the IPO.

    This episode includes AI-generated content.
    Mehr anzeigen Weniger anzeigen
    2 Min.
  • Google s 75 million A24 AI deal
    Jun 24 2026
    Google’s $75M Hollywood Bet: The A24 Partnership and the Fight for "Authentic" AI

    Is Google buying artistic credibility for its AI? In today’s episode, we break down the shocking $75 million investment by Google DeepMind into A24, the independent powerhouse behind Everything Everywhere All at Once and the 2026 box-office hit Backrooms. We investigate why this "anti-studio studio" is the perfect partner for Google to prove that AI can be an assistant rather than a replacement for human creativity.

    We dive into the unprecedented terms of the deal: Google gets zero access to A24’s film library and the agreement is non-exclusive, meaning A24 can still work with other AI models. We analyze the first concrete tool in development—an AI storyboard assistant—and discuss the "administrative warfare" of filmmaking, from scheduling to pre-visualization. Finally, we address the deep skepticism from within A24’s own ranks, including Backrooms director Kane Parsons, who admitted he would make generative AI disappear if he could. Join us as we explore if this partnership can bridge the gap between Big Tech automation and handmade cinema.


    This episode includes AI-generated content.
    Mehr anzeigen Weniger anzeigen
    2 Min.
adbl_web_anon_alc_button_suppression_t1
Noch keine Rezensionen vorhanden