Ai2 budgets GPU time, Willison codes by voice, OpenAI sells tier over Astra
Every URL the pipeline pulled into ranking for this issue — primary sources plus the supporting and contradicting findings each Researcher returned. Inline citations in the issue point back here.
Sources
Impactful scheduling for GPU clusters huggingface.co
A new feature for my blog, built using my voice simonwillison.net
I shipped a new feature for my blog today: the Newsletters page, which offers an index of all of the newsletters I’ve sent out, both my free weekly Substack and my monthly sponsors-only updates. I built the feature almost entirely using my voice, chatting away to my laptop while I cooked dinner. Codex voice mode I used the ChatGPT desktop app for this, in the Codex tab, using the voice conversation mode, running against a local development environment. Here’s what that looks like: I started the…
Sophos cuts threat investigation time by 96% with OpenAI Daybreak openai.com
Discover how Sophos uses OpenAI’s Daybreak to cut cyber-threat investigation time by 96% and automate 52% of MDR cases while preserving human oversight.
Asana cuts model costs 76x in browser tests with GPT-6.1 Sol openai.com
Using GPT-6 Astra in Codex, Asana made its browser agent 76x cheaper and 5x faster in tests to offer customers more capable models.
LegalOn halves Codex costs while maintaining development speed openai.com
LegalOn cut estimated daily Codex costs by 65% while maintaining development speed. It matched Astra, Sol, and Luna to tasks and managed budgets strategically.
References
Artificial Analysis artificialanalysis.ai
GPT-6.1 Sol replaces GPT-6 Sol after just 7 days with near-Astra intelligence
OpenAI Asana case study openai.com
Approximately 29x of this reduction stemmed from optimizing workflow parallelization, the remainder was driven by Sol’s aggressive caching and efficiency
Welcome.AI (LegalOn coverage) welcome.ai
LegalOn cuts costs by 65% while sustaining development speed
UnderDefense ‘2026 AI SOC Trends’ underdefense.com
Level 1 SOC analyst roles are being ‘sawed off at the bottom’ … some CISOs have privately reported cutting L1 headcount by as much as 60–80% after deploying modern agentic platforms
Ogletree Deakins legal brief ogletree.com
California amends Cal-WARN Act to require AI and automation disclosures in layoff notices
Datacouch enterprise comparison datacouch.io
Anthropic’s Project Glasswing using Claude Mythos 5.1 achieved a 73% success rate in professional ‘Capture the Flag’ benchmarks and discovered long-standing bugs in OpenBSD
SchedMD Fair Tree documentation slurm.schedmd.com
Fair Tree prevents account coordinators from inadvertently harming their users’ priorities relative to other departments… the Classic algorithm suffers precision loss in deep hierarchies, causing ties the scheduler handles poorly.
DataCenterDynamics on Meta Llama 3 training datacenterdynamics.com
During the 54-day training run for Llama 3, Meta reported 419 unexpected interruptions—roughly one every three hours—mostly due to hardware failures like HBM3 memory issues.
Yang & Yoo (Korea University SSLab), ‘Lazer’ scheduler, DAC 2026 infoscience.epfl.ch
‘Futile preemptions’—where a job is pushed aside before it can finish its first execution cycle—inflate Job Completion Time by up to 1.6x; Lazer uses Bayesian optimization to pick preemption windows and reduces futile preemptions by 2x–67x.
kubex.ai ‘Kubernetes GPU Scheduling’ comparison (2026) kubex.ai
NVIDIA open-sourced Run:ai as the KAI Scheduler (CNCF Sandbox) in April 2025… KAI v0.10.0 adds time-based fair-share that penalizes gpu-hogging queues based on historical Prometheus data; Kueue now integrates natively with Dynamic Resource Allocation (DRA, GA in K8s 1.34).
Medium / N. Raman, ‘GPU Scheduling: The Hidden Infrastructure Crisis’ medium.com
Checkpointing overhead for massive models can take 15–45 minutes to save and another 15–45 minutes to restore; preempting a job twice in a 24-hour window can result in an 8% efficiency loss.
daily.dev discussion thread on Ai2 post daily.dev
The project’s technical breakdown generated 56 comments… some commenters expressed skepticism about the readability of AI-assisted technical blog posts… open questions remain regarding potential capacity fragmentation as massive distributed training jobs navigate the new preemption boundaries.
ML6 — GPT-6 Astra cost-efficiency analysis ml6.eu
Astra used 68% fewer tokens than GPT-5.6 Sol to complete the same software migration tasks, reducing an average 10.2 million token workflow down to just 3.3 million… on simpler coding tasks, the cheaper GPT-6.1 Sol often matches Astra’s performance at a fraction of the price.
Kompozy review of OpenAI voice mode desktop kompozy.io
Voice Mode can drain 10% of a five-hour rolling usage window in just five minutes of active conversation… routing bugs where the app loses conversation memory when switching between text and voice modes.
CyberSecurityNews — GPT-5.6 Codex $HOME deletion incident cybersecuritynews.com
When the agent attempted to define a temporary workspace, it tried to override the $HOME environment variable… resulted in the execution of recursive deletion commands (e.g., rm -rf $HOME) on the host system rather than the intended temporary folder.
Corpowid — Vibe coding and accessibility debt corpowid.ai
AI generators prioritize the visual ‘vibe’ of an interface over its semantic structure, they frequently emit code that fails WCAG standards for keyboard navigation and screen reader compatibility… these systems often fail users with non-standard speech patterns, such as those with dysarthria, stuttering, or apraxia.
Gadget Flow — Future of voice-controlled workspaces thegadgetflow.com
Constant voice-prompting creates a ‘noisy, distracting’ environment that disrupts coworkers’ focus… sensitive logic or proprietary secrets may be exposed if spoken aloud in common areas… some companies [are] moving toward ‘cube farms’ or private soundproof booths specifically for AI interaction.
Ryan Shrott on Medium — Voice coding tools 2026: still clunky medium.com
The high ‘correction trap’—the difficulty of fixing a single misspoken term without manual input—often leads developers to abandon voice workflows for complex debugging in favor of returning to a physical keyboard.
