home.social

#gemini — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #gemini, aggregated by home.social.

  1. RE: social.lol/@brennan/1166408222

    🖥️ Three protocols that have their own ecosystems, their own communities, and their own aesthetics:

    finger://
    gopher://
    gemini://

    🗓️ Two predate the World Wide Web entirely, but one was created in 2019

    🚫 None of them require a GUI

    🚫 None of them require JavaScript

    🖥️ All three of them run in a terminal

  2. Três protocolos com seus próprios ecossistemas, comunidades e estéticas:

    finger://
    gopher://
    gemini://

    Dois precedem a World Wide Web, mas um foi criado em 2019

    Nenhum deles requer interface gráfica

    Nenhum deles requer JavaScript

    Todos os três funcionam num terminal

    brennan.day/gemini-gophers-and

  3. Three protocols that have their own ecosystems, their own communities, and their own aesthetics:

    finger://
    gopher://
    gemini://

    Two predate the World Wide Web entirely, but one was created in 2019

    None of them require a GUI

    None of them require JavaScript

    All three of them run in a terminal

    brennan.day/gemini-gophers-and

  4. Google’s Android XR audio glasses launch this fall with Gemini support for navigation, messaging, photos, and hands-free voice assistance. 🕶️
    Google partnered with Samsung, Gentle Monster, and Warby Parker as AI wearables expand ambient data access beyond smartphones. 📡

    🔗 blog.google/products-and-platf

    #TechNews #Google #Android #AndroidXR #Gemini #SmartGlasses #Wearables #Samsung #WarbyParker #GentleMonster #Privacy #AI #XR #Android #Hardware #Tech

  5. Google’s Android XR audio glasses launch this fall with Gemini support for navigation, messaging, photos, and hands-free voice assistance. 🕶️
    Google partnered with Samsung, Gentle Monster, and Warby Parker as AI wearables expand ambient data access beyond smartphones. 📡

    🔗 blog.google/products-and-platf

    #TechNews #Google #Android #AndroidXR #Gemini #SmartGlasses #Wearables #Samsung #WarbyParker #GentleMonster #Privacy #AI #XR #Android #Hardware #Tech

  6. Google’s Android XR audio glasses launch this fall with Gemini support for navigation, messaging, photos, and hands-free voice assistance. 🕶️
    Google partnered with Samsung, Gentle Monster, and Warby Parker as AI wearables expand ambient data access beyond smartphones. 📡

    🔗 blog.google/products-and-platf

    #TechNews #Google #Android #AndroidXR #Gemini #SmartGlasses #Wearables #Samsung #WarbyParker #GentleMonster #Privacy #AI #XR #Android #Hardware #Tech

  7. Google’s Android XR audio glasses launch this fall with Gemini support for navigation, messaging, photos, and hands-free voice assistance. 🕶️
    Google partnered with Samsung, Gentle Monster, and Warby Parker as AI wearables expand ambient data access beyond smartphones. 📡

    🔗 blog.google/products-and-platf

    #TechNews #Google #Android #AndroidXR #Gemini #SmartGlasses #Wearables #Samsung #WarbyParker #GentleMonster #Privacy #AI #XR #Android #Hardware #Tech

  8. Google’s Android XR audio glasses launch this fall with Gemini support for navigation, messaging, photos, and hands-free voice assistance. 🕶️
    Google partnered with Samsung, Gentle Monster, and Warby Parker as AI wearables expand ambient data access beyond smartphones. 📡

    🔗 blog.google/products-and-platf

    #TechNews #Google #Android #AndroidXR #Gemini #SmartGlasses #Wearables #Samsung #WarbyParker #GentleMonster #Privacy #AI #XR #Android #Hardware #Tech

  9. Google is replacing link-first Search with AI answers, agents, and generative interfaces that keep users inside Google’s ecosystem. 🔎
    New AI Mode features include persistent agents, custom mini apps, and interactive results, increasing pressure on publishers losing referral traffic. 🤖

    🔗 techcrunch.com/2026/05/19/goog

    #TechNews #Google #GoogleSearch #Search #SearchEngine #SEO #AI #Gemini #AIOverviews #Publishers #OpenWeb #Privacy #FOSS #Transparency #Automation #BigTech #Web

  10. Google is replacing link-first Search with AI answers, agents, and generative interfaces that keep users inside Google’s ecosystem. 🔎
    New AI Mode features include persistent agents, custom mini apps, and interactive results, increasing pressure on publishers losing referral traffic. 🤖

    🔗 techcrunch.com/2026/05/19/goog

    #TechNews #Google #GoogleSearch #Search #SearchEngine #SEO #AI #Gemini #AIOverviews #Publishers #OpenWeb #Privacy #FOSS #Transparency #Automation #BigTech #Web

  11. Google is replacing link-first Search with AI answers, agents, and generative interfaces that keep users inside Google’s ecosystem. 🔎
    New AI Mode features include persistent agents, custom mini apps, and interactive results, increasing pressure on publishers losing referral traffic. 🤖

    🔗 techcrunch.com/2026/05/19/goog

    #TechNews #Google #GoogleSearch #Search #SearchEngine #SEO #AI #Gemini #AIOverviews #Publishers #OpenWeb #Privacy #FOSS #Transparency #Automation #BigTech #Web

  12. Google is replacing link-first Search with AI answers, agents, and generative interfaces that keep users inside Google’s ecosystem. 🔎
    New AI Mode features include persistent agents, custom mini apps, and interactive results, increasing pressure on publishers losing referral traffic. 🤖

    🔗 techcrunch.com/2026/05/19/goog

    #TechNews #Google #GoogleSearch #Search #SearchEngine #SEO #AI #Gemini #AIOverviews #Publishers #OpenWeb #Privacy #FOSS #Transparency #Automation #BigTech #Web

  13. Google is replacing link-first Search with AI answers, agents, and generative interfaces that keep users inside Google’s ecosystem. 🔎
    New AI Mode features include persistent agents, custom mini apps, and interactive results, increasing pressure on publishers losing referral traffic. 🤖

    🔗 techcrunch.com/2026/05/19/goog

    #TechNews #Google #GoogleSearch #Search #SearchEngine #SEO #AI #Gemini #AIOverviews #Publishers #OpenWeb #Privacy #FOSS #Transparency #Automation #BigTech #Web

  14. Google lanza AI Threat Defense: plataforma autónoma para combatir ciberataques impulsados por IA en tiempo real

    En un momento en que los ciberdelincuentes utilizan inteligencia artificial para detectar y explotar vulnerabilidades en horas en lugar de semanas, Google Cloud presentó su respuesta: una plataforma que combina Gemini, Mandiant, Wiz y CodeMender en un sistema de defensa siempre activo que no solo detecta amenazas, sino que las repara automáticamente (Fuente Google).

    La carrera armamentista en ciberseguridad acaba de dar un salto de magnitud. Google Cloud lanzó oficialmente su plataforma AI Threat Defense el 28 de mayo de 2026, un sistema de seguridad autónomo diseñado para ayudar a las empresas a mantenerse al ritmo de los ciberdelincuentes que utilizan IA para descubrir y explotar vulnerabilidades a velocidades sin precedentes. El problema que busca resolver es concreto e inquietante: la «ventana de explotación», el tiempo crítico entre que una vulnerabilidad es descubierta y que es atacada, se redujo de semanas a apenas horas. Los métodos de seguridad manuales tradicionales ya no son suficientemente rápidos para detener ataques automatizados e impulsados por IA.

    La arquitectura de la plataforma integra cuatro activos clave de Google en un sistema unificado. AI Threat Defense fusiona los modelos de la familia Gemini, la empresa de seguridad en la nube Wiz —recientemente adquirida por Google—, el agente de corrección de código CodeMender y la práctica de inteligencia de amenazas y respuesta a incidentes de Mandiant. El resultado es un sistema que opera en cuatro etapas secuenciales: Preparar, Escanear y Priorizar, Remediar y Monitorear.

    Lo que distingue a esta plataforma de otras soluciones existentes es su enfoque en la acción, no solo en la alerta. Uno de los principales problemas de los sistemas actuales de ciberseguridad con IA es el enorme volumen de alertas generadas durante los análisis de vulnerabilidades. AI Threat Defense se diferencia porque no solo detecta y señala riesgos, sino que los prioriza activamente según el riesgo real en el mundo y automatiza el camino hacia la solución. En palabras de Francis deSouza, COO de Google Cloud y Presidente de Productos de Seguridad: «Los atacantes están usando IA para descubrir y aprovechar vulnerabilidades a velocidades sin precedentes. Los equipos de seguridad e ingeniería ya no pueden encontrar, analizar y parchear estas fallas manualmente con la velocidad suficiente. Este sistema está diseñado para superarlos.»

    El contexto competitivo en el que llega este lanzamiento es igualmente revelador. Anthropic introdujo su modelo de ciberseguridad Claude Mythos en abril, mientras que OpenAI lanzó su plataforma Daybreak basada en GPT-5.5 a principios de mayo. Google llega tercero en este segmento específico, pero con una ventaja difícil de igualar: la integración nativa de Mandiant —considerada una de las firmas de inteligencia de amenazas más respetadas del mundo— junto con la escala de infraestructura de Google Cloud y el poder de razonamiento de Gemini.

    La arquitectura segura por defecto de Google ya bloquea automáticamente 10 millones de correos spam por minuto y protege a miles de millones de usuarios en todo su portafolio. AI Threat Defense busca llevar esa misma filosofía de defensa proactiva al corazón de la seguridad empresarial en la era de la IA. La pregunta ya no es si las organizaciones necesitan defenderse con IA: es si pueden permitirse no hacerlo.

    #AIThreatDefense #ciberataques #ciberseguridad #CodeMender #gemini #google #GoogleCloud #IA #InteligenciaArtificial #Mandiant #openai #PORTADA #seguridadEmpresarial #vulnerabilidades #wiz
  15. Gemini-3.5-flash догнал GPT-5.5 на 97/S и в 2.5× дешевле. Но главное — китайцы выигрывают по цене и качеству

    Месяц назад я писал про парадокс DeepSeek V4 Pro — модель проиграла собственному Flash и Qwen 3.6 Plus трёхнедельной давности. Сегодня прогнал свежий battle на пяти моделях — два американских флагмана и три китайских — и расклад снова поменялся. Главное: Gemini-3.5-flash взял 97/S, тот же балл, что у GPT-5.5, и в 2.5 раза дешевле. Google впервые на моём тесте встал рядом с OpenAI на длинном русском контенте. Tencent Hy3-preview даёт A-tier за $0.0017 за вызов — в 134 раза дешевле GPT-5.5. DeepSeek V4 Pro я прогнал третий раз — качество стоит на месте (87), но цена упала в 5.4× и она снова в зоне полезного. Qwen 3.7 Max упорно вставляет китайские иероглифы в русский текст — регрессия относительно собственной 3.6, где этой проблемы не было. По дороге заметил, что наша формула cost_per_call жила полгода с приближением, которое занижало стоимость на 50–140%. Починили — теперь берём реальные токены из OpenRouter response.usage . И отдельно — почему при такой скорости релизов модель в продукте стоит держать заменяемой, а не зашитой в код намертво.

    habr.com/ru/articles/1040770/

    #LLM #Gemini #gpt5 #DeepSeek #Qwen #Tencent #Benchmarks #AI #OpenRouter #Russian_NLP

  16. @Steve Yup. This was hilariously the response when I tossed the link at Gemini. Absolultely not what I wrote.

    #Google #Gemini

  17. Gemini-3.5-flash догнал GPT-5.5 на 97/S и в 2.5× дешевле. Но главное — китайцы выигрывают по цене и качеству Месяц назад я п...

    #LLM #Gemini #gpt-5 #DeepSeek #Qwen #Tencent #Benchmarks #AI #OpenRouter #Russian #NLP

    Origin | Interest | Match
  18. @jaffathecake my British #Gemini begs to differ (but well, they must be cousins because they agree at least on the presence of 1 P)

  19. MediaTek lancia il Dimensity 8550 con Gemini Nano V3: l’AI on-device arriva in fascia media

    MediaTek ha ufficialmente presentato il Dimensity 8550, il suo nuovo SoC per smartphone di fascia medio-alta. Il chip porta una novità particolarmente rilevante per gli utenti Android: il supporto nativo a Gemini Nano V3, il modello AI di Google ottimizzato per l'elaborazione direttamente sul dispositivo senza necessità di connessione cloud. Gemini Nano V3 e il nuovo LLM Booster La caratteristica principale del Dimensity 8550 è l'integrazione di un modulo chiamato LLM Booster, progettato […]

    androidiani.net/mediatek-lanci

  20. Postei no BR-Linux to avisando (e é sobre small web!)

    __
    DIAS DE UM FUTURO ESQUECIDO: FINGER, GOPHER E O GEMINI (NÃO AQUELE, O OUTRO!)

    Um oportuno artigo de @brennan recostura a linha entre os protocolos finger:// (1971), gopher:// (1991) e gemini:// (2019), e os alinha em um mesmo nicho, muito interessante: o da chamada small web, onde não há cookies, nem pixels de rastreamento, nem recursos de terceiros, nem espionagem analítica.

    br-linux.org/2026/01/dias-de-u

    #gopher #smallweb #gemini

  21. RT @AiBattle_: Das Gemma 4 124B MoE-Modell wird schon seit einiger Zeit gerüchteweise erwartet und wurde sogar von Jeff Dean erwähnt. Ich hoffe, dass es nicht mehr lange dauert, bis es endlich veröffentlicht wird. Nathan Lambert (@natolambert): Befreit das 100B Gemma 4 MoE! Gemini Flash 3.5 ist da, also könnt ihr es jetzt veröffentlichen! — nitter.net/natolambert/status/

    mehr auf Arint.info

    #AI #Gemini #Gemma4 #MachineLearning #MoE #arint_info

    https://x.com/AiBattle_/status/2059629633939579004#m

  22. RT @ArtificialAnlys: Artificial Analysis and IBM Research are launching ITBench-AA, the first in a new series of benchmarks evaluating models on agentic enterprise IT tasks, starting with Site Reliability Engineering tasks where frontier models score below 50% ITBench-AA’s SRE tasks benchmark model performance on Kubernetes incident response, where models must diagnose live systems by reading logs, tracing dependencies, and identifying root-cause entities across complex infrastructure. The underlying ITBench dataset has been developed by @IBM's Software Innovation Lab, leveraging IBM’s deep expertise in enterprise IT operations Artificial Analysis has worked closely with IBM over the last 6 months to develop a implementation of the dataset for frontier AI evaluation, beginning with Site Reliability Engineering (SRE) and expanding to Financial Operations (FinOps) and Chief Information Security Officer (CISO) tasks over time ITBench-AA SRE overview: ➤ 59 SRE tasks in total: 40 public tasks and 19 brand new, held-out tasks ➤ Each task provides a Kubernetes incident snapshot containing alerts, events, traces, metrics, logs, and application topology. The model must identify the minimal set of independent root-cause Kubernetes entities responsible for the incident ➤ Faults span typical SRE failure modes including infrastructure, service, application, and chaos-injected incidents, such as resource quota exhaustion, rollout failures, connection pool exhaustion, and network partitions Methodology details: ➤ Agentic harness: each task is solved by the model running in our open-source Stirrup referenc…

    mehr auf Arint.info

    #Claude #DeepSeek #Gemini #GPT5 #Kubernetes #Qwen37 #arint_info

    https://x.com/ArtificialAnlys/status/2059698327235805258#m

  23. I really enjoyed this post with lots of info on Gemini, Gopher, history and the present day including links to software I haven’t tried yet. brennan.day/gemini-gophers-and #gemini #gopher #web

  24. I really enjoyed this post with lots of info on Gemini, Gopher, history and the present day including links to software I haven’t tried yet. brennan.day/gemini-gophers-and #gemini #gopher #web

  25. I really enjoyed this post with lots of info on Gemini, Gopher, history and the present day including links to software I haven’t tried yet. brennan.day/gemini-gophers-and #gemini #gopher #web

  26. I really enjoyed this post with lots of info on Gemini, Gopher, history and the present day including links to software I haven’t tried yet. brennan.day/gemini-gophers-and #gemini #gopher #web

  27. I really enjoyed this post with lots of info on Gemini, Gopher, history and the present day including links to software I haven’t tried yet. brennan.day/gemini-gophers-and #gemini #gopher #web

  28. "a single attack pattern lets a malicious repository achieve remote code execution through AI coding assistants. The agent is tricked into a benign-looking file copy that secretly overwrites its own config, and the next restart runs attacker code with full user privileges.

    We confirmed the technique against #Claude Code, #Gemini CLI/Antigravity CLI, #CursorAgent CLI, GitHub #Copilot CLI, #Grok Build, and #OpenAI Codex CLI"

    Ooops.

    #GenAI

    h/t @campuscodi

    adversa.ai/blog/the-approval-p