Tech News
← Home  ·  All topics

Code

407 GoKawiil briefs on this topic · RSS for code

Anthropic set to price IPO above $70-$78 range ahead of Nov. 19 Nasdaq debut

Sources tell CNBC that Anthropic's IPO is likely to price above its current $70-to-$78 range, with order books set to close Tuesday at 4 p.m. ET. The AI company will list on the Nasdaq under ticker 'ANT' and begin trading Nov. 19. Anthropic reported $33.24 billion in revenue for the first nine months of 2026, up about 1,200% year-over-year, alongside a $104 billion net loss largely tied to changes in the value of convertible notes and derivatives.

Anthropic details how Claude Code team verifies AI-written code

Anthropic published an account of how its Claude Code team approaches verification, noting that roughly 90% of its codebase is written and reviewed by Claude. The team says structured verification loops, such as a separate grader checking work against a rubric, boosted task success by up to 10 points in internal tests, with the biggest gains on harder problems.

New solver 'Soksolv' outperforms all published Sokoban solvers, cracks 24 world-first levels

A new Sokoban solver called Soksolv has solved more levels than any previously published solver on two standard benchmark suites, including 24 levels never automatically solved before. It also solved shared levels roughly 3.2x faster in total time than the next-best solver, Festival, built with help from AI coding assistants over a 14-day session.

New Rust tool 'unref' finds dead Python code using Ruff's parser

A developer released unref, an open-source dead code detector for Python written in Rust and built on Ruff's parser. The tool's benchmarks claim it scans the entire CPython codebase (about 2,300 files) in under a second, roughly 20 times faster than the existing vulture tool, and ships with built-in rules for pytest, Pydantic and SQLAlchemy.

Essay 'Do Repeat Yourself' Critiques Corporate Platform Team Standardization Push

The piece uses a fictional scenario at 'Megabank' to illustrate how separate engineering teams independently built similar features using different tech stacks, then describes a newly hired VP who creates a centralized platform team to enforce one shared stack across all groups. The narrative frames this standardization effort as driven by a desire to eliminate perceived duplication and inefficiency, despite each team having shipped working products on time.

TPC16 recreates Turbo Pascal 7 command-line compiler, source released

A project called TPC16 offers a from-scratch reimplementation of Borland's Turbo Pascal 7.0 command-line compiler (TPC.EXE), written entirely in Pascal. It is described as source-code compatible, using the same syntax and producing identical compiled units and 16-bit x86 code as the original Borland tool.

Developer releases free self-hostable PancakeSlicer for non-planar 3D printing

A developer known as Alpaca Engineering released PancakeSlicer, a free, self-hostable slicing tool that lets 3D printers deposit material along curved, non-planar paths across all three axes rather than flat horizontal layers. The tool uses 3D models of specific print heads to calculate safe clearance angles and avoid collisions, and already supports a wide range of printers from brands including Bambu Lab, Prusa, Creality, and Voron.

Developer argues AI-generated code should default to Open Source licensing

In an opinion piece, developer jcubic argues that code produced entirely by AI should be released under Open Source licenses or dedicated to the public domain via CC0. He bases this on AI models having been trained on scraped Open Source and Creative Commons-licensed code without attribution, and on the legal precedent from the Monkey Selfie copyright case, which established that non-human-created works cannot be copyrighted.

IETF draft proposes HTTP status code 483 'I'm Not a Freezer'

An Internet-Draft submitted independently to the IETF by R. Bánffy proposes a new HTTP status code, 483, labeled 'I'm Not a Freezer.' The code would be returned when a server understands a request requiring freezing or refrigeration capabilities but cannot provide them. The draft also attempts to distinguish between freezers, refrigerators, low-temperature servers, frozen software processes, and services marketed as 'cold storage.'

Lean-based audit finds 138 drafting errors in US Tax Code's Title 26

Researchers behind the global-law project ran a Lean formalization of the entire Internal Revenue Code, checking every textual cross-reference against the official structured data from the Office of the Law Revision Counsel as of November 3, 2026. The automated suite flagged 257 findings, with an AI agent verifying each against official text and amendment history, identifying 138 confirmed drafting mistakes such as broken or outdated references.

Essay argues formal verification's rise via AI doesn't guarantee bug-free systems

A technology essay observes that formal verification has gone mainstream over the past year, driven by AI tools like Claude generating large volumes of verification code (citing a billion lines of Rocq for one project) and growing interest from figures associated with Claude Code, Vercel, and Codex/Cursor. The author notes startups such as Theorem and Harmonic are demonstrating economic value from verification work with Anthropic and in verifying C programs and hardware designs, while cautioning that 'formally verified' systems can still contain bugs.

Lawyer upgrades homemade V100 AI server, moves DeepSeek model to 4-bit precision

Nick Mermiges, a South Carolina trial lawyer and founder of PLI Labs, upgraded his private eight-GPU V100 AI server by adding 1TB of RAM and an RTX 3090, shifting roughly 80% of its workload from 2-bit to 4-bit NVFP4 precision. The server runs DeepSeek V4.1 Flash with a custom legal workflow he built using Claude Code, and the upgrade traded some speed—code output dropped from 160–215 to 150–180 tokens per second—for stability, after crashes traced to a faulty power supply were resolved.