The Hallway Track

Latest signals

Every signal tracked from the world's AI conferences — 4,665 talks, quotes & announcements, newest first.

Quantization: The Size vs Quality Trade-Off

Hugging Face · Jun 16, 2026

Quantization shrinks AI models via fewer bits per value, trading some quality for smaller size and faster inference.

“a slightly worse model that actually fits your setup can still be much more useful than the full precision version”
You Might Not Need 50 Diffusion Steps — Ziv Ilan, Nvidia

AI Engineer · Jun 16, 2026

Nvidia is borrowing LLM optimization techniques to reduce diffusion model denoising steps below the typical 20-50 for real-time generation.

“We know that it's cool to generate videos or to generate images, but now if we talk about a developer context or a enterprise context, this should be fast.”
Simulating Humans at Scale: Simile's Joon Sung Park

Sequoia Capital · Jun 16, 2026

Simile founder Joon Sung Park is building an applied AI lab to simulate human behavior and societies at scale.

“I do see an opportunity today to really take the first crack at building the simulation. I would not have said that even 5 years ago, but that is a conviction that we have built up over the years as we are going deep into this research.”
Agentic Productivity with Amazon Quick | How to Build with AWS

AWS (re:Invent) · Jun 16, 2026

AWS show demonstrates building agentic productivity apps using the managed Amazon Quick service.

“Quick is an AWS managed service that handles, you know, a lot of the the underlying infrastructure and provides a nice user interface for you to build agents.”
The Fable 5 Export Controls Harm US Cyber Defense

Simon Willison · Jun 16, 2026

Fable 5 export controls wrongly ban defensive AI security work mislabeled as a jailbreak.

“That is not a guardrail bypass. It is the most valuable thing an AI model can do for defensive security: executing the find, fix, and test loop defenders run every day.”
Quoting Matteo Wong, The Atlantic

Simon Willison · Jun 16, 2026

A cybersecurity expert says Anthropic's Fable model refused a security-review prompt but complied when asked to 'fix this code,' calling it working as intended.

“the model working as intended”
[AINews] Satya on Loopcraft: Building Frontier Ecosystems

Satya Nadella · Latent Space Blog · Jun 16, 2026

Satya argues the real AI opportunity is building a frontier ecosystem and learning loops, not picking the best model.

“This means the real opportunity is not in picking the best model but instead in building a learning loop on top of models where human capital and token capital compound.”
Cloudflare CAPTCHA on at least one ampersand

Simon Willison · Jun 16, 2026

Cloudflare CAPTCHA rules can be scoped to only challenge search URLs containing an ampersand, fixed via Claude Code.

“After some mucking around with Claude Code it turns out you can register the following rule instead, so the CAPTCHA only kicks in for search URLs containing at least one ampersand”
Make your UIKit app fully resizable

Apple Developer (WWDC) · Jun 15, 2026

iOS 27 makes iPhone apps fully resizable, with Xcode AI features automating the migration work.

“Xcode's intelligence features and the new app modernization skill can do a lot of it for you.”
New domain for Sign in with Apple and iCloud+ Hide My Email

Apple · Apple Developer News · Jun 15, 2026

Apple will unify Sign in with Apple and Hide My Email under one domain, private.icloud.com.

“Later this summer, Apple will unify the email domains used by Sign in with Apple and iCloud+ Hide My Email under a single, shared domain: private.icloud.com.”
Introducing Gemma 4 models on Amazon Bedrock

AWS Machine Learning Blog · Jun 15, 2026

Google DeepMind's open-weight Gemma 4 model family is now available on Amazon Bedrock.

“Artificial Analysis reports an Intelligence Index of 39 for Gemma 4 31B, well above the median of 15 in the 4B–40B open-weights class.”
Google's agentic shift | Logan Kilpatrick

Sequoia Capital · Jun 15, 2026

Google is rebasing its products to be agentic-native, powered by the anti-gravity agent harness as a new through line.

“everything is now sort of using Gemini in some way. That's now becoming true for anti-gravity as sort of all of the products rebase to become agentic native products and actually taking action on behalf of users”
Trailer for Livestream 1 of POSETTE: An Event for Postgres 2026

Microsoft Developer (Build) · Jun 15, 2026

Microsoft showcases its Postgres contributions and Postgres AI tooling like MCP servers in the POSETTE 2026 event trailer.

“We're going to look at the range of ways you can build MCP servers from very exploratory servers where the agents can send full SQL queries to your database”
One man just liberated Fable... and now it’s illegal

Fireship · Jun 15, 2026

Days after Claude Fable's release, the US government restricted it citing national security after it was jailbroken into a cyber weapon.

“an American company that will not stop talking about AI safety just got safetied by its own government”
Modernize Java apps with AI​ | OD871

Microsoft Developer (Build) · Jun 15, 2026

Microsoft's GitHub Copilot app modernization extension uses agentic AI to upgrade legacy Java apps.

“94% of executives say that AI is critical, but unfortunately only 28% have scaled pilots into production.”
AI Agent Failure Detection and Root Cause Analysis with Strands Evals

AWS Machine Learning Blog · Jun 15, 2026

AWS's Strands Evals SDK adds detectors that automatically diagnose AI agent failures and recommend fixes.

“Detectors answer “why did it fail?” by producing diagnoses at the per-span level with categorized failures, causal chains, and fix recommendations.”