thank you, yeah, there's still a lot i need to do around the LLMs. getting the underlying stuff working was most of the work up to now. thx for the pointer on model, will update when i fix tool calling
LLMS
-
SambaCloud platform for deploying frontier AI models and agent pipelines
By
–
SambaCloud is SambaNova’s cloud platform for deploying frontier AI models.
— SambaNova (@SambaNovaAI) 20 mai 2026
It runs models like @MiniMax_AI M2.7 at extremely low latency thanks to purpose-built AI hardware—making it a strong option for production agent pipelines where speed is critical.
Start building:… pic.twitter.com/mUYKhx8rMZSambaCloud is SambaNova’s cloud platform for deploying frontier AI models. It runs models like @MiniMax_AI M2.7 at extremely low latency thanks to purpose-built AI hardware—making it a strong option for production agent pipelines where speed is critical. Start building:
-
Google I/O 2026 AI Announcements Summary
By
–
Google ran I/O 2026 last night.
— AlphaSignal AI (@AlphaSignalAI) 20 mai 2026
Google I/O Day 2 starts in an hour.
Day 1 was packed with AI announcements across models, agents, browser tooling, and infrastructure:
– Gemini 3.5 Flash became Google’s new default inference model
– Managed Agents added persistent sandboxed… https://t.co/SThQG9aq8m pic.twitter.com/KWscqhwCZPGoogle ran I/O 2026 last night. Google I/O Day 2 starts in an hour. Day 1 was packed with AI announcements across models, agents, browser tooling, and infrastructure: – Gemini 3.5 Flash became Google’s new default inference model
– Managed Agents added persistent sandboxed -
Improving Grok’s reasoning capabilities for agentic tasks
By
–
As Jason mentions, it’s a top priority. Grok was initially trained to give quick answers in chat mode, rather than think hard for as long as needed to get the task right, which is what matters in agentic mode. We are fixing that.
-
Nature Editorial on the Role and Impact of AI Scientists
By
–
Important editorial @Nature on the new "AI-scientist" papers
"AI scientists can and should empower human
researchers. They cannot and should not replace them." https://
nature.com/articles/d4158
6-026-01551-3
… -
Anti-Self-Distillation Technique for Reasoning Reinforcement Learning
By
–
Anti-Self-Distillation for Reasoning RL via Pointwise Mutual Information
-
Gemini Shifts from Chatbot to Autonomous Task-Oriented AI System
By
–
6/ Gemini is no longer a chatbot. It is now a system that runs tasks on its own. Spark in the background. Antigravity for development. Omni Flash for media. The chat box is the smallest piece of the product.
-
Breakdown of Gemini 3.5 Flash architecture and its operational layers
By
–
5/ Gemini 3.5 Flash is the core model. The other three sit on top of it.
Omni Flash handles video. Spark runs your tasks in the background. Antigravity is the environment developers use to build with it. One model, three different surfaces. -
Google’s Antigravity platform uses AI agents to automate software development
By
–
4/ Antigravity 2.0.
— AI Highlight (@AIHighlight) 20 mai 2026
Google's coding platform. Like Cursor or Claude Code, but built around running many AI agents at once.
In the I/O demo, 93 subagents built a working OS in 12 hours for under $1,000.
Use for: building software. pic.twitter.com/fwGJi9mhpP4/ Antigravity 2.0. Google's coding platform. Like Cursor or Claude Code, but built around running many AI agents at once. In the I/O demo, 93 subagents built a working OS in 12 hours for under $1,000. Use for: building software.
-
Gemini Omni Model Released for Multimodal Video Creation and Editing
By
–
2/ Gemini Omni.
— AI Highlight (@AIHighlight) 20 mai 2026
A new model that creates and edits video from any input. Text, audio, image, or another video.
Live in the Gemini app, Flow, and YouTube Shorts.
Use for: video creation and editing. pic.twitter.com/AO0mldizDr2/ Gemini Omni. A new model that creates and edits video from any input. Text, audio, image, or another video. Live in the Gemini app, Flow, and YouTube Shorts. Use for: video creation and editing.