The Hallway Track
Every signal tracked from the world's AI conferences — 4,665 talks, quotes & announcements, newest first.
Claude Opus 5.5 composed surprisingly competent retro game music via a custom text format
“I wonder if the ability to compose competent music is similar to the 3D graphics thing - a new capability for text models that emerged in the past few months?”
Apple launches iPhone Duo, a foldable smartphone with multi-posture form factor for developers
NVIDIA introduces Green Contexts for fine-grained GPU resource sharing between concurrent workloads
“Controlling how GPU resources are shared between them remains difficult.”
Salesforce Agentforce splits agent reasoning into probabilistic LLM and deterministic business logic layers
“Agentforce splits the agent brain into two layers. The first is a generative layer, the LLM, that does all of the language and reasoning. The second is a deterministic layer.”
Banning open-weight models will hurt US AI competitiveness while increasing long-term cyber risk
“I feel we're going to end up making policy decisions that both limit American AI competitiveness and increases long-term cyber risk.”
NVIDIA GTC Berlin keynote announcement with no substantive content available
Meta's ads MCP server integrates with Databricks to connect enterprise customer data to ad campaigns
OpenAI frontier model solves open math problems, releases formal Lean proofs on GitHub
YDUQS used Databricks to achieve 90% productivity gains across 1.3M students
“we can bring context to our IAIS, to our agents, who are IAIS who understand our business, who understand our students, who understand our employees and our teachers”
OpenAI and Ironclad are training AI agents on complex legal contracting workflows
Azure launches flexible spend-based savings plans for database cost optimization
“It's a flexible spend-based commitment that gives you meaningful discounts across eligible Azure database services without locking you into a specific service, region, or configuration.”
Falcon-Emirati LLM is trained to understand Emirati dialect, culture, and nuance.
Reflection AI launches Beam, a 501B/23B-active US-trained open MoE model for coding and agentic work
OpenAI Codex now runs on mobile and can be self-hosted via open-source application server
“You can host Codex anywhere thanks to its open source code and application server”
PayPal is integrating with OpenAI and Google agent commerce protocols to enable AI-initiated payments
“PayPal Enterprise Payments is building bridges for merchants so they can accept payments from many different places.”
Braintrust workshop teaches isolating signal from AI agent trace noise using observability tooling
AI agents should operate with zero credentials, secured via sandboxes and MCP gateways
“Agents do a lot more than I thought...than I expected them to do.”
Apple iPhone Duo launches October 23, 2026, requiring developer app updates.
Inference engine performance bottlenecks live in the scheduler and GPU execution layer
Claude Cowork moves model inference and VM execution to the cloud for mobile and battery improvements
“The "new" version of Cowork runs model inference and the VM in the cloud. Each session gets its own sandbox, not sharing state with other sessions.”
LLM judges used as agent benchmark verifiers are unreliable and overconfident evaluators
“The fact is that the web is a very open system. There is no single path to correctness. There is no absolute truth.”
Z.ai's 753B-parameter GLM 5.3 MoE model is now available on Amazon Bedrock for enterprise use
Collibra-acquired DZee Labs targets the POC-to-production gap for enterprise unstructured data RAG pipelines
“we didn't have unstructured data, and that's where DZee shines”
Arize AI presents a 101 framework for evaluating AI agents from tracing to LLM-as-judge meta-evaluation