With one million displaced, Lebanon turns to digital wallets for aid (wired.com)

As Israel’s attacks and the resulting displacement push more than one million people into temporary shelters in Lebanon, humanitarian funding is increasingly flowing through digital wallet and fintech platforms rather than traditional aid channels. The article describes how remittances and peer-to-peer transfers are being routed via wallets to vetted intermediaries and recipients for fast, on-the-ground spending, citing Lebanon’s historically large informal donation flows and declining public institutional trust. It also notes that while payments are subject to anti-money-laundering monitoring, Lebanon’s regulatory framework leaves a looser gray area compared with places like the UAE.

Computational Physics (2nd Edition) (websites.umich.edu)

The page provides online companion resources for Mark Newman’s book, Computational Physics (2nd edition), including sample chapters, exercise text, programs and datasets, and downloadable figures for instructional and self-study use.

The Melanesian: Dark-skinned people with blonde hair region of Oceania (guardian.ng)

The article describes Melanesia in the South Pacific, including countries such as Vanuatu, the Solomon Islands, Fiji and Papua New Guinea, and explains that many Melanesians have dark skin and occasionally blond hair. It discusses competing theories for how blond hair developed in the region, and notes research suggesting it evolved independently of the genes associated with European blond hair. The piece also outlines genetic diversity across Melanesian islands and touches on cultural and religious change following European contact and Christianisation.

Codex is switching to API pricing based usage for all users (help.openai.com) AI

OpenAI’s Codex rate card has been updated: as of April 2, 2026, Codex pricing for new and existing ChatGPT Business customers and new ChatGPT Enterprise plans shifts from per-message estimates to token-based usage (credits per million input, cached input, and output tokens). The article provides separate legacy rate cards for Plus/Pro and most other plans until migrations are completed, with users advised to review both during the transition.

Great Plains Shelterbelt (en.wikipedia.org)

The Great Plains Shelterbelt was a 1934–1942 U.S. project that planted millions of trees across a broad strip of the Great Plains to reduce wind erosion and moisture loss after the Dust Bowl. It was carried out with support from agencies including the U.S. Forest Service and the Works Progress Administration, and by 1942 had planted about 220 million trees over 18,600 square miles. The article also notes that many shelterbelts have since declined in health, and that recent efforts and grants have aimed to maintain and restore them in several central Plains states.

Drivers May Soon Pay Taxes Based on How Much Their Car Weighs (autoblog.com)

The article discusses a proposal in which drivers could be taxed based on their vehicle’s weight, rather than existing fee structures, potentially changing how vehicle-related taxes are calculated. The piece outlines the concept and why lawmakers or regulators might consider shifting to weight-based taxation.

Why Over-Engineering Happens (yusufaytas.com)

The article discusses why teams often over-engineer solutions, citing factors such as uncertainty, risk aversion, desire to avoid future rework, and misaligned incentives or planning.

Meta, Google under attack as court cases bypass 30-year-old legal shield (cnbc.com) AI

Recent court losses for Meta and Google, along with other lawsuits, are testing whether platforms can still rely on Section 230’s protections that have shielded them for decades. CNBC reports that plaintiffs are pursuing narrower theories aimed at bypassing the law—often by focusing on how products are designed and how AI-generated summaries or recommendations are presented to users. The article notes that while penalties so far are limited, the cases could shape future litigation as the industry shifts from traditional social media and search toward AI-driven experiences, with possible appeals up to the Supreme Court.

Roogle: a Rust API search engine (github.com)

Roogle is an open-source Rust API search engine that lets users search for functions by name and Rust type signatures. The repository documents a query syntax (including support for generic and bounded types) and provides example usage via HTTP and Docker.

Perfmon – Consolidate your favorite CLI monitoring tools into a single TUI (github.com)

Perfmon is an open-source, Go-based terminal TUI that consolidates common CLI monitoring commands (e.g., top, vmstat, netstat) into a tabbed interface. It provides live sparklines for quick system health checks, supports configurable refresh intervals and custom commands via a TOML file, and includes theme toggling plus Linux/macOS defaults.

Nanocode: The best Claude Code that $200 can buy in pure JAX on TPUs (github.com) AI

A GitHub discussion introduces “nanocode,” a fully open-source, end-to-end approach to train a Claude-Code-like agentic coding model using pure JAX on TPUs. The author describes an architecture and training pipeline based on Anthropic-style Constitutional AI and Andrej Karpathy’s nanochat, including synthetic data generation, preference optimization for alignment, and TPU-optimized training. They report that a 1.3B-parameter model (d24) can be reproduced in about 9 hours on a TPU v6e-8 for roughly $200, with smaller variants costing less, and provide starter commands and training/evaluation notes.

A tail-call interpreter in (nightly) Rust (mattkeeter.com)

Matt Keeter describes building a tail-call–based Uxn CPU interpreter in nightly Rust, using the new `become` feature to ensure dispatch compiles into true tail calls rather than growing the call stack. The resulting VM keeps state in function arguments (optimized like registers) and outperforms both his earlier Rust and hand-written ARM64 assembly in benchmarks, though the x86 results are mixed and still behind his assembly backend. He also shares implementation details, including a macro to generate the tail-call functions for different opcode signatures, and notes codegen/performance observations from generated machine code.

Microsoft terms say Copilot is for entertainment purposes only, not serious use (tomshardware.com) AI

Microsoft’s updated Copilot terms state the AI is “for entertainment purposes only,” warns it can make mistakes, and says users should not rely on it for important advice. The article notes this caution is similar to disclaimers from other AI providers and argues it conflicts with how Microsoft markets and integrates Copilot into products like Windows 11 for business use. It also emphasizes the need to verify AI outputs due to issues like hallucinations and automation bias.

Phone-free bars and restaurants on the rise across the U.S. (axios.com)

Phone-free bars and restaurants are spreading across the U.S. as diners seek to disconnect from screens, amid related restrictions and concerns about the effects of smartphones and social media. At least 11 states have venues with phone limits or digital-detox incentives, including policies that lock phones away or offer perks for leaving devices at home. Supporters and operators say removing phones can improve attention, conversation, and the overall in-person experience.

StackOverflow: Retiring the Beta Site (meta.stackoverflow.com)

Stack Overflow’s Meta says it will retire its beta site in the coming weeks, remove access to it, and stop supporting it. The unified posting experience from the beta won’t be brought to the main site, though some opinion-/subjective content from the beta may be reformatted into the standard Q&A style. The company also plans not to migrate most beta UI/design elements and will communicate more as the shutdown approaches.

Code Reviews Need to Evolve (latent.space) AI

The article argues that traditional human code reviews are becoming infeasible as code changes grow and AI-generated code increases review time and effort. It proposes shifting review “upstream” to human-authored specs and acceptance criteria, with automated, deterministic verification (tests, type checks, contracts), layered trust gates, restricted agent permissions, and adversarial verification by separate agents. The overall point is to replace approval-by-reading-diffs with approval-by-verifying intent and constraints before code is generated.

The Locksmith's Apprentice – Claude told me to expose my data without auth (mpdc.dev) AI

An IT operator describes building a self-hosted “security operations brain” for AI-assisted monitoring, then discovering it had been exposed to the public internet for 11 days due to a tunnel/DNS setup with no authentication. He says Anthropic’s Claude helped design and deploy the system via Anthropic’s MCP tooling, but authentication was never considered, even as multiple AI sessions continued to access and modify his exposed data. After discovering the issue, the fix was to remove the DNS record, and he uses the incident to argue that AI can follow correct procedures while missing real-world security context and urgency.

Baby's Second Garbage Collector (matheusmoreira.com)

The article describes the evolution of a “Baby’s First Garbage Collector” in the Lone Lisp project into a more advanced collector that can recover escaped objects by scanning not just the Lisp stack but also the native stack beneath it. It explains how the collector marks reachable heap values efficiently, and how conservative assumptions and pragmatics are used to avoid quadratic checks. The author then notes that tests revealed remaining “missing” objects, leading to further investigation into places like CPU registers.

Why the US Can't Have Nice Things, Part 2 (walkingtheworld.substack.com)

After a delayed international flight and an extended, chaotic TSA passport-control process at JFK, the author describes a crowded early-morning A train marked by homelessness, filth, and tense encounters. Using travel observations from places like Bulgaria and Korea, he argues that the U.S.—especially New York—has a broken, low-trust public sphere where riders without alternatives are hit hardest. He suggests that while some procedures could be improved, the deeper problem is cultural: social trust has eroded and can’t be repaired with policy tweaks alone.

Banray.eu: Raising awareness of the terrible idea that is always-on AI glasses (banray.eu) AI

The Banray.eu site argues that Meta’s camera-equipped “Ray-Ban Meta” glasses enable always-on, privacy-invasive surveillance, including potential sharing and human review of recorded footage by subcontractors. It also claims Meta is preparing built-in facial identification features that would expand consent-free facial data collection, and points to broader industry moves toward smart glasses with persistent recording. The article urges venues and regulators to adopt policies against such devices and facial recognition.

EPA Official in Charge of Methane Regs Wrote Oil Industry Argument Against Them (propublica.org)

ProPublica reports that EPA air official Aaron Szabo, who now leads efforts to loosen methane regulations, had previously helped draft a 2022 oil-industry letter opposing methane controls. The article says Szabo later pushed for delays and revisions to the Biden-era methane rules and solicited industry input, including specific regulatory language, through meetings and internal communications. Critics argue his prior lobbying suggests the EPA is being influenced by oil and gas interests while methane rules that would cut emissions are being weakened.

Artemis II crew see first glimpse of far side of Moon (bbc.com)

NASA’s Artemis II crew, including Reid Wiseman, Victor Glover, Christina Koch and Canadian astronaut Jeremy Hansen, have begun orbiting around the Moon’s far side and described it as a new and unusual view. The astronauts shared a photo of the Orientale basin, which NASA says is the first time the entire basin has been seen by human eyes. The report also notes the Orion spacecraft’s distance from Earth as it continues the mission.

Trump uses expletive-ridden social media post to threaten Iran's infrastructure (theguardian.com)

In a Truth Social post, Donald Trump used abusive language while threatening to attack Iran’s energy and transport infrastructure unless Iran reopens the Strait of Hormuz to shipping. The Guardian reports Trump framed the warning as “Power Plant Day” and “Bridge Day,” and said he could reach a deal with Iran by Monday. The live coverage also notes concurrent reporting of drone damage to Kuwait’s energy facilities and airstrikes in Iran and Lebanon.

SPF/PC v4 for MS-DOS, FreeDOS, x86 (github.com)

MOSHIX’s GitHub repository provides SPF/PC v4, an MS-DOS (including FreeDOS/x86) “abandonware” SPF editor from around 1993. The project describes it as an ISPF-like panel and editing environment with an included REXX implementation, intended to run on DOSBox and also usable on Windows and MS-DOS.

Japanese, French and Omani Vessels Cross Strait of Hormuz (japantoday.com)

Shipping data shows that since Thursday three Omani-operated tankers, a French-owned container ship and a Japanese-owned LNG carrier have crossed the Strait of Hormuz. The article says Iran had initially halted traffic after U.S.-Israeli strikes but later allowed passage for ships it deems friendly, and that some vessels adjusted their AIS settings while transiting. Japan’s Mitsui O.S.K. Lines reported its co-owned Sohar LNG crossed the strait, while other Japanese-linked ships remain stranded in the region.