Después de muchos años como BurritoMobile…
Hoy nace oficialmente Loki IT 🟢
Misma experiencia en servidores, cloud y soporte,
pero ahora con enfoque en:
* Managed IT
* Infraestructura cloud
* Voz y mensajería empresarial
* Automatización
Si alguna vez trabajaste conmigo antes… seguimos aquí 🙂
Solo evolucionamos.
🌐 https://t.co/6Tir2aw6Me
Get an intro to the Transformer architecture and the BERT model.
Explore the components of the Transformer architecture and the tasks BERT can be used for, such as text classification, question answering, and natural language inference → https://t.co/zMXXc5hCsq
Get @GoogleCloud certified. ☁️
Google Cloud customers can join GEAR to apply and access an up to 9-week cohort featuring technical mentorship and instructor-led training into various topics → https://t.co/zt92SyA7hy
What is responsible AI, why is it important, and how does Google implement it in our tools?
Take this 30-minute course on Google Skills to get an intro to all things responsible AI. Plus, learn about Google's 3 AI principles → https://t.co/jN6T8kz5K5
Prompt injection tops the OWASP LLM Top 10 and there's no single fix.
Instead, you stack defenses, each one catching what the others miss.
Defenses come in two families: model-level and system-level.
Model-level defenses teach the model to resist injection.
- Spotlighting wraps untrusted text in control tags like <UNTRUSTED>…</UNTRUSTED> and tells the model to treat anything inside as data, not instructions.
- Instruction Hierarchy fine-tunes the model to rank the developer's system prompt above the user's message, and both above third-party content.
System-level defenses build a system around the LLM that bounds the damage.
- Least-Privilege Tools: Give the agent the minimum tools it needs.
- Human-in-the-Loop: Require explicit user approval before any sensitive action runs.
- Planner / Executor Split: Two separate LLMs. The planner has tool access but never sees untrusted content. The executor reads untrusted content but has no tools.
No single defense is enough. Production systems like Gmail stack them, and together they make indirect injection manageable.
Over to you: what's the one defense you've seen work in production that isn't on this list?
All you need to learn about LLMs is 10 minutes!
This microlearning Google Skills course explores what large language models are, the use cases where they can be applied, and how you can use prompt tuning to enhance LLM performance → https://t.co/Cv0MzO1LnC
Available now: Google’s official Agent Skills repository on @github!
Learn how to equip agents with additional, condensed expertise with Agent Skills—and stay tuned for additional skills in this repo in the coming weeks and months → https://t.co/K3CIaC5LTX
Introducing Agents CLI in Agent Platform: create to production in one CLI!
Agents CLI gives your AI assistant a direct, machine-readable line to the Google Cloud agent stack—turning a fragmented ecosystem into a seamless assembly line → https://t.co/jlA4TtgttD
New GEAR learning paths are officially LIVE.
Master enterprise agent scaling and production deployment to earn exclusive new skill badges.
✅ 35 monthly credits? Ready.
✅ New paths? Launched.
✅ Your AI career? Accelerated.
Get started: https://t.co/e1SiSihvtE