Machine Learning on MacWorks

2026-04-13

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Engineering @ Scale — 2026-04-13#

Signal of the Day#

When using large language models for recommendation systems, passing raw numerical counts ruins the signal because the model processes digits as text tokens rather than magnitudes. By converting raw engagement counts into percentile buckets wrapped in special tokens (e.g., <view_percentile>71</view_percentile>), LinkedIn increased the correlation between popularity and embedding similarity 30x, offering a highly reusable pattern for safely encoding structured numerical data into transformer contexts.

2026-04-12

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-12#

Watch First#

Building Towards Self-Driving Codebases with Long-Running, Asynchronous Agents offers a highly credible look into the mechanics of long-running coding agents from Cursor’s founder, cutting through the hype to explain the concrete architectural hurdles of scaling AI from autocomplete to massive, unsupervised pull requests.

Youtube Tech Channels

Mon, 01 Jan 0001 00:00:00 +0000

Tech Videos — Week of 2026-04-04 to 2026-04-10#

Watch First#

[Why, and how you need to sandbox AI-Generated Code? — Harshil Agrawal, Cloudflare] from the AI Engineer channel is the single best watch this week because it strips away agent hype to deliver a stark reality check: executing generated code means running untrusted internet code in production. It provides a strict, capability-based security framework for deciding when to use V8 Isolates versus full Linux containers to prevent compute exhaustion and credential leaks.

2026-04-08

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Tech Videos — 2026-04-08#

Watch First#

Why, and how you need to sandbox AI-Generated Code? — Harshil Agrawal, Cloudflare from the AI Engineer channel is the most critical watch of the day. It strips away the AI hype to state a fundamental truth: if your agent executes generated code, you are running untrusted code from the internet in production. It delivers a strict, pragmatic capability-based security framework for deciding when to use V8 Isolates versus full Linux containers to prevent credential leaks and compute exhaustion.

2026-04-07

Mon, 01 Jan 0001 00:00:00 +0000

Hacker News — 2026-04-07#

Top Story#

The standout technical feat today is “Solod”, a new strict subset of Go that translates directly to C. It strips away Go’s heavy runtime and garbage collector, offering a “Go in, C out” workflow for systems programming with manual memory management and native C interop.

Front Page Highlights#

[Netflix Void Model: Video Object and Interaction Deletion] · Github Netflix open-sourced a fascinating video inpainting model built on CogVideoX that doesn’t just erase objects—it calculates physical interactions. If you remove a person holding a guitar from a video, the model understands that the person’s effect on the guitar is gone, causing it to naturally fall to the ground. It relies on a clever two-pass pipeline using Gemini and SAM2 for masking, solving long-standing temporal consistency issues with warped-noise refinement.

2026-04-04

Mon, 01 Jan 0001 00:00:00 +0000

Sources

Company@X — 2026-04-04#

Signal of the Day#

Anthropic is restricting Claude subscription access for third-party tools like OpenClaw, prompting Hugging Face to aggressively push users toward open-source local models like Gemma 4. This policy shift highlights a growing fracture between closed API ecosystems moving to lock down interfaces and the open-source community’s push for self-hosted AI.

Company@X

Mon, 01 Jan 0001 00:00:00 +0000

Company@X — Week of 2026-04-04 to 2026-04-10#

Signal of the Week#

Meta’s launch of Muse Spark marks a massive strategic shift, as the newly formed Meta Superintelligence Labs abruptly abandons the company’s recent open-weights strategy. By releasing a proprietary, natively multimodal reasoning model equipped with “Contemplating mode,” Meta is signaling its intent to directly rival extreme test-time reasoning systems like Gemini Deep Think and GPT Pro.

Key Announcements#

Meta · Muse Spark Meta introduced Muse Spark, its first major model since Llama 4, built on a completely overhauled data pipeline, architecture, and infrastructure. Keeping the model proprietary is a massive pivot to compete in the high-end reasoning space, with the company deploying it exclusively via the Meta AI app and an upcoming private API.

Hacker News

Mon, 01 Jan 0001 00:00:00 +0000

Hacker News — Week of 2026-04-04 to 2026-04-10#

Story of the Week#

Anthropic’s frontier AI models crossed a terrifying new threshold in autonomous cybersecurity, completely shifting the industry’s threat model. First, Claude Code uncovered a complex, 23-year-old vulnerability in the Linux kernel’s NFS driver that predated Git itself. Days later, the infosec community went into full meltdown when Anthropic’s unreleased “Mythos” model autonomously wrote a 200-byte ROP chain exploit for FreeBSD and demonstrated the ability to reliably escape Firefox’s JavaScript virtualization sandbox in 72.4% of trials.