Cognition built FrontierCode eval after SWE-bench saturated, focusing on code mergeability
“okay, this code is technically correct. But would you actually merge it? Would you feel happy? Would this improve the quality of your code base?”
Latest signals
Every signal tracked from the world's AI conferences — 4,665 talks, quotes & announcements, newest first.
Madrigal Pharmaceuticals cut AI agent time-to-production from 12 weeks to 2 using LangChain
“The first time I looked at a trace in LangSmith, I felt like I was peering into the brain of our AI agent.”
Mobileye cut support response times 90% using Amazon Bedrock AgentCore with zero infrastructure overhead.
“The results were so compelling that Mobileye transformed AgentCore into a self-service platform for teams across the company to deploy their own AI agents.”
Cold Card Bitcoin hardware wallet drained 1,600 BTC via flawed random number generator
“for the last 5 years, without them knowing, the cold card never did”
AWS engineers built a WebSocket/native-messaging bridge letting cloud agents call local MCP servers.
“This is the same pattern that powers products like Claude Cowork, a cloud agent calling local tools through MCP, but fully self-hosted on AWS with your own model and custom tool servers.”
Amazon Bedrock AgentCore harness is now generally available with an n8n community node
“AgentCore harness gives you that layer as a managed capability. You define an agent in configuration, including the model it uses, the tools it calls, the skills it has access to, and the instructions it follows, and AgentCore assembles and runs the loop for you.”
NVIDIA RTX Spark is a 1-petaflop superchip for running advanced AI agents locally on laptops
“The reinvention of the personal computer has begun and it starts with the RTX Spark”
Apple's Xcode agents now perform context-aware app localization across entire projects automatically
“This is more than a word-for-word translation. Xcode looks at each string in its context, the surrounding code, UI, the action.”
Microsoft QDK integrates GitHub Copilot as an AI tutor for in-editor quantum computing education
Agentic AI enables autonomous networks that reason about intent, not just automate scripts
“An agent is a model plus a harness. If the model is the brain, the harness is the body.”
Microsoft pitches unified data-to-intelligence stack via Fabric, Foundry, and Work IQ
“Microsoft is really working on this unified stack where you can take your data wherever it is and take it all the way to intelligence through this stack of Work IQ, Foundry IQ, and Fabric IQ.”
LangSmith now provides full observability for voice agents built on Gemini Live
“To take that agent to production safely, you need visibility into what your agent is doing.”
Apple introduces new App Store creative asset formats for app marketing beyond in-use visuals
Apple's August 2026 developer newsletter highlights visual intelligence integration guidance for app developers
A coding agent uses voice calls to escalate irreversible decisions to humans before acting
“Instead of guessing, it recognizes this is irreversibly my decision, loads the voice escalation skill, and calls me.”
Coding agent phones developer autonomously only for non-reversible decision points while working AFK
“I'm Echo, your coding assistant working on Tempo in Cloud Code. We're at a fork in the road and I need your decision.”
A coding agent autonomously phones the developer when it hits a judgment call it cannot make itself.
“I can't decide that for you.”
StarCloud raised $170M to build space data centers solving AI's terrestrial energy bottleneck
“we are very quickly running up on constraints on where we can build new energy projects terrestrially. And so by building them in space, we get access to this almost unlimited lowcost energy in the form of solar.”
Qwen 3.8 Max challenges OpenAI and Anthropic at 5-10x lower API pricing
“it sat there thinking for 16 days, starting from an empty folder, writing, testing, and repairing its own code”
a16z-backed Ulysses builds underwater robots for defense, infrastructure, and ocean restoration driven by AI supercycle demand
“AI super cycle is increasing demand for subcables. It's increasing demand for critical minerals. It's increasing demand for energy. These are all things which the ocean is a primary resource for.”
Over 90% of US critical minerals are refined in China, threatening AI hardware supply chains.
“Over 90% of the critical minerals we use are refined in China today, which is kind of wild to imagine given that that was a very fast transition just over the last several decades.”
Radiant and a16z argue mass-produced small nuclear reactors can revive US energy leadership
“we regulated nuclear out of business because we made it too expensive”
NVIDIA's Rubin GPU architecture kills megakernels, ending a major inference optimization research direction
“the GPU is designed in such a way that it kills mega kernels. So it seems like that entire research field won't be continued.”
LLM 0.32 adds visible reasoning traces, server-side tools, and OpenAI Responses API support
“the most significant new version of LLM since the initial launch of the project”