WAITING FOR YOU

Unread

No deadlines. Just things worth spending time with.

58 articles

Post 50: Good Research Takes are Not Sufficient for Good Strategic Takes — Neel Nanda

The author argues that being good at AI research is only weak evidence of being good at strategic thinking about AGI. Research offers clear feedback; strategy does not, so researchers can hold bad strategic views and never correct them. He lists skills good strategic takes need and asks readers to check what evidence supports a person's strategic claims.

Post 51: Socratic Persuasion: Giving Opinionated Yet Truth-Seeking Advice — Neel Nanda

The author recommends giving advice by breaking your argument into key steps and asking the other person questions about each step. If you get an unexpected answer, you likely missed context, so you can pivot early. The method requires real humility and willingness to change your mind, and costs more time and effort than direct advice.

MATS Applications Open (Due Aug 29) — Neel Nanda

This is a recruitment call for the MATS mechanistic interpretability program run by Neel Nanda. Applicants complete a 12-20 hour research task and write-up by August 29. About 32 candidates enter a 5 week paid online phase, then about 8 enter a 12 week paid in-person phase writing a paper. The author says past scholars often publish at top ML venues.

Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights

This article reports that a frontier LLM can read and write neural network weights without gradient descent. It identifies trained behaviors from raw weights, writes game-playing networks zero-shot, and edits a small share of a language model's weights to fix looping and raise math scores. Results are early; a model without looping defects could not be improved, showing limits of the method.

The era of personal software

The author describes building a custom health tracking web app to collect data from his devices. He uses automated exports, a hosted Postgres database, and AI agents to log meals and give insights. He argues that LLMs make personal software cheap to build, but not a full replacement for existing tools yet.

Battling Impostor Syndrome

This essay explains impostor syndrome among software engineers: doubting your skills even with proof of success. The author suggests reframing thoughts, comparing yourself only to your past self, and writing down daily accomplishments. Techniques include mentoring both ways, asking co-workers questions, and noting what improved your mood during highs and lows.

How to make (good) pixel art effects with Claude

The author argues that AI coding agents cannot one-shot good pixel art visual effects; you must iterate. The method: have the agent write many code-driven effect drafts rendered as short videos, then pick and refine them by eye. Quality comes from repeated human judgment, not prompts. The author shares open source effects as starting references.

PHP2Shell: AI uncovers critical RCE vulnerabilities affecting 70% of the Internet

Wiz Research found memory-safety bugs in core PHP that let attackers run code or read worker memory without login. The bugs affect PHP versions back to 5.4, and for a decade exploit building was too complex for attackers. Frontier AI models made building exploit chains practical, and PHP released a patch on October 21, 2026.

Book Review: Seeing Like A State

This review covers James Scott's book about state planning. Scott argues that governments simplify complex systems to make them easy to measure, tax, and control. These plans often fail because they ignore local practical knowledge. The reviewer likes the historical examples but questions how far the thesis applies.

The Work Required to Have an Opinion

You must work hard before you earn an opinion. You must know the other side's arguments, find facts that contradict your view, and test your ideas. The essay says most people skip this work, so most opinions stay weak.

dlab Open Source Week: Frontier AI on Your Own Hardware — Tim Dettmers

Tim Dettmers argues that AI agents make research cheap, so small university labs can compete with big labs. His lab will release an agent harness, a local inference framework, and a compaction method that halves agent costs. He tells students to change how they work, keep their identity, and attack problems that need few resources.

Everyone is doing Git on S3 wrong

The article argues that Git-on-object-storage designs using a write-ahead log, like Cursor's Continuity, gain durability and read scaling but remain effectively single-writer and still need full repositories on disk. The author presents Enroute, which instead uses per-repository Raft consensus, stateless compute nodes, optimistic locking, and a segment/layer format confining deltas to commit ancestry, with S3 for durability. Reported peak exceeds 7,600 pushes per second into one repository; reads benchmark near stock Git, but clients pay more to decode served deltas, and garbage collection remains unsolved.

How to keep enjoying programming in a world of LLMs

A Haskell developer, worried about LLM-driven burnout and skill atrophy, argues against full vibe coding: you should keep writing code yourself while delegating planning, research, review cycles, and low-risk cleanup to agents. The tradeoff is accepting modest productivity gains (roughly 2x by his estimate) instead of boasting numbers, in exchange for retaining enjoyment, codebase ownership, accountability, and independence from token limits and frontier-model vendors.

repr(Rust) - The Rustonomicon

This Rustonomicon chapter explains the default repr(Rust) type layout rules: alignment, size, and compiler-inserted padding. It argues that field ordering is deliberately unspecified because generic monomorphizations may need different orderings to avoid wasted space, and enums can use tricks like the null pointer optimization to eliminate tags. The tradeoff is compiler layout flexibility versus the inability to assume a stable memory layout.

Rebuilding AUTOMATIC1111 with Gradio Workflow

This article walks through Workflow1111, a rebuild of AUTOMATIC1111's stable-diffusion-webui as a 73-node Gradio Workflow canvas, arguing that node graphs can mix local Python functions with remote model calls. It shows that every output node becomes a REST endpoint and MCP tool, and that visitors can run pipelines under their own quota. Tradeoffs are acknowledged: there is no loop operator, and roughly two-thirds of nodes run in-process without a network call.

Async GRPO with LoRA across HF Jobs: a bucket, a proxy, and no NCCL

A rank-1 LoRA adapter is only a few MB, so async GRPO can span separate HF Jobs: trainer and vLLM replicas swap adapters via a mounted Storage Bucket instead of NCCL, with a proxy adding auth, broadcasting adapter loads, and routing rollouts by KV prefix. Five instrumented runs chase the bottleneck between training and generation; packing, dropping gradient checkpointing, and raising in-flight caps cut 500 steps from 3h27m to 53min at equal reward. Tradeoffs: FUSE sync lag, retry logic, concurrency caps.

Your Agent Aced the Task. Will It Do It Again?

The article argues that average benchmark accuracy hides agent unreliability: a ReAct agent on GPT-4.1 scored 77.4% Mean@5 on AppWorld but only 53.0% Pass^5 (succeeding on every run). Their Consistency Analyzer resamples decision points from a single recorded trajectory to find flip-prone steps, and the resulting injected guidelines halved the gap (24.4pp to 12.0pp) without costing average accuracy. Tradeoffs: diagnosis needs one extra LLM call per decision step, and consistency is orthogonal to capability, so a bigger model won't fix it.

A 0-click exploit chain for the Pixel 10: When a Door Closes, a Window Opens

The article documents a zero-click-to-root exploit chain on Pixel 10: an updated Dolby decoder exploit (CVE-2025-54957) paired with a new VPU kernel driver flaw whose mmap handler maps unbounded physical memory, exposing the kernel image. The author credits Android's improved triage (High severity rating, 71-day fix) but argues the same vendor again shipped a shallow, easily found driver bug, so proactive driver security remains weak.

On the Effectiveness of Mutational Grammar Fuzzing

The author critiques mutational coverage-guided grammar fuzzing: coverage feedback misses semantic dependencies (e.g., bugs requiring chained function calls), and greedy corpus accumulation produces highly similar samples. The proposed fix is periodically restarting fuzz workers with empty corpora while syncing with a server holding accumulated coverage and samples, alternating independent generation with combined-corpus mutation. In week-long libxslt experiments this found more unique crashes faster than uninterrupted runs, though the optimal interval T was target-dependent.

Testing race conditions with memory access tracing and stack-based delay injection

The author presents MAccConc, tooling that traces kernel memory accesses via ASAN outline instrumentation fed through KCOV, and uses count-augmented stack traces to stably identify accesses across runs, enabling delay injection that forces specific thread interleavings. This supports confirming race-condition candidates, writing reliable regression tests, and eventually fuzzing. Tradeoffs include ASAN missing some stack/global accesses, coverage loss on kernel panics, and constraint-style orderings being less deterministic than fully-specified ones; kernel patches are not yet upstream.

Optimizing CPU-side Rendering Code – Godot Engine

A Godot engine developer demystifies CPU-side renderer optimization, outlining a methodology—profile hotspots, understand the cause, fix, re-measure—and illustrating CPU/GPU tradeoffs (2D batching helps CPU-bound games; 3D occlusion culling offloads the GPU). Two case studies show the process: reusing vertex buffers for animated Polygon2D (28 to 83 FPS) and assigning per-thread heaps to remove allocation stalls in SPIRV-to-DXIL transpilation (11 seconds saved). The core lesson: optimizations are often counter-intuitive, so measurement is mandatory.

Dev snapshot: Godot 4.8 dev 5 – Godot Engine

This Godot 4.8 dev 5 snapshot, one of the last before feature freeze, highlights mip-level texture streaming: only needed mipmaps load based on camera distance, cutting VRAM for large 3D worlds, though it requires opt-in via project settings, an editor restart, and re-importing textures as 'Texture2D Streamed'. Other changes include preserved alpha test coverage for distant alpha-scissor materials, a simplified 2D editor toolbar, Feral GameMode support on Linux, and roughly 12 MB of runtime RAM savings in core types.