NVIDIA's internal AI deployment now processes 4 trillion tokens per day at 99.96% availability.
“twenty-eight percent of enterprises today are now seeing an immediate a sustainable consistent ROI on their AI projects”
Latest signals
Every signal tracked from the world's AI conferences — 4,665 talks, quotes & announcements, newest first.
AI architectural research fails because it tests at insufficient compute scale
“to get to any interesting results you need certain level of compute to even see the capabilities in the model”
Turbopuffer CEO Simon Eskildsen shares origin story in Pragmatic Engineer interview
“it just sort of escalated from there.”
Meta doubled GEM ads model training efficiency to 20-25% MFU while scaling FLOPs 4x in 12 months
F1's agentic AI system cut data source onboarding from 8 weeks to 40 minutes
“For the first time, we have end-to-end visibility across the entire MarTech platform with data lineage and root cause analysis, not just dashboards full of alerts”
AWS automates Bedrock Automated Reasoning policy refinement, eliminating manual diagnose-edit-retest cycles
“Refining an Automated Reasoning policy in Amazon Bedrock has been a manual cycle of diagnose, hand-edit, retest, and repeat.”
A nightly cron prompt can automate dependency rebasing and validation via coding agents
“Set up a nightly cron job that executes the prompt: fetch upstream changes to the <software> and rebase all local changes on top of upstream. Check that the software works as intended and replace the current version.”
NVIDIA proposes isolated tenant Kubernetes clusters to share GPU infrastructure across teams
NVIDIA Vera delivers faster encryption, compression, and recovery for agentic AI storage workloads.
“Storage is an active part of every agentic AI workflow.”
LLMs have made open source's promise of user-modifiable software practically feasible for the first time
“I think LLMs have changed that equation in a way that makes the original dream much more feasible.”
Google's Kaggle ran a free AI agents course for 353,000 learners worldwide
Interconnects launches Artifacts Hub and Adoption Dashboard to track open model ecosystem adoption
“We're sharing them freely to help the open ecosystem find its strengths and grow.”
Robotics replacing blue-collar workers remains far off due to unstandardized physical environments
“if robotics is going to do what we're doing with these companies today, every single of those buildings need to be either 3D printed or completely like standardized. I don't see that happening.”
Databricks positions generative AI as the replacement for manual executive reporting decks
Databricks Variant data type for semi-structured ingestion is now generally available
Researchers built a working self-replicating AI worm that steals GPU compute to run LLMs autonomously.
“demonstrate that self-sustaining AI-driven cyber-threats are no longer theoretical”
Databricks completes Panther acquisition to build a unified security lakehouse platform
Databricks ML pipelines help energy suppliers reclaim 70% of theft exposure worth billions annually
“we've seen energy suppliers reclaim about 70% of their theft exposure by implementing machine learning and AI processes”
DeepSeek's updated flash model beats its larger pro version via post-training improvements alone
“Just the post-training step changed?”
OpenAI built GPT-Live, a turnless voice AI system enabling continuous real-time conversation in six months
Circles telco saw 22% ARPU increase and 9% churn reduction using OpenAI API and Codex
Anthropic and OpenAI jointly backed MCP Apps, enabling branded interactive UIs inside AI chat clients
“just a few months ago, we partnered with Anthropic and OpenAI to create the official extension to MCP, which we call MCP apps”
MCP Tasks V2 spec with radical changes is coming in July, explaining why no agents have implemented it yet.
“there's some pretty radical changes. So, I'm going to show you what's happening with V2”
AI labs optimize for benchmark scores rather than real-world model quality, warns industry insider
“unfortunately the teams are not getting better models overall but better Elm Marina models whatever that is possibly something with a lot of nested list bullet points and emojis”