GoKawiil - Tech News

1.

Liquid AI's smallest model yet LFM2.5-230M beats models 4X its size at data extraction, can run 'anywhere' (venturebeat.com)

2026-06-25 | tags: liquid ai, lfm2.5-230m, qwen3.5

2.

Alibaba's model never trained as an agent — and improved agent performance across seven benchmarks (venturebeat.com)

2026-06-24 | tags: alibaba, qwen-agentworld, qwen3.7-max

3.

Xiaomi's HarnessX rewrites its own AI scaffolding mid-task — and smaller models gain the most (venturebeat.com)

2026-06-24 | tags: xiaomi, harnessx, qwen3.5

4.

Two Qwen3 models on one DGX Spark: the residency math (news.ycombinator.com)

2026-06-18 | get NVIDIA DGX Station A100 → | tags: qwen3, dgx spark, vllm

5.

SimpleHelp bug lets hackers create rogue remote support accounts (bleepingcomputer.com)

2026-06-15 | get Remote Support Software License → | tags: simplehelp, openid connect, horizon3.ai

6.

Rio de Janeiro's city government model Rio3.5 beats Qwen3.7 in recent benchmarks (news.ycombinator.com)

2026-06-14 | tags: rio de janeiro, rio 3.5, qwen3.7

7.

Alibaba's Qwen3.7-Plus supports text, video and imagery inputs at low cost of $0.4/$1.6 per 1M token — but it's proprietary (venturebeat.com)

2026-06-02 | get AI Text and Video Generator → | tags: alibaba, qwen3.7-plus, ai large language model

8.

Pinterest cut AI costs 90% by gutting a frontier model's vision layer (venturebeat.com)

2026-05-29 | get AI Vision Model Toolkit → | tags: pinterest, qwen3-vl, openai

9.

Theseus: Translating Win32 to WASM (news.ycombinator.com)

2026-05-25 | get WebAssembly for Windows Apps → | tags: win32, wasm, theseus

10.

A Comma and a Question Mark (news.ycombinator.com)

2026-05-22 | get Comma and Question Mark Keyboard → | tags: zsh, llama.cpp, qwen3.6

11.

Alibaba's proprietary Qwen3.7-Max can run for 35 hours autonomously and supports external harnesses like Anthropic's Claude Code (venturebeat.com)

2026-05-21 | get AI Development Kit → | tags: qwen3.7-max, alibaba, anthropic

12.

Qwen3.7-Max: The Agent Frontier (news.ycombinator.com)

2026-05-20 | get AI Assistant Developer Kit → | tags: qwen3.7-max, agent, frontier

13.

Deal: Sony WH-1000XM6 headphones drop below $400, matching the record-low price (androidauthority.com)

2026-05-19 | get Sony WH-1000XM6 Headphones → | tags: sony, wh-1000xm6, qn3

14.

Orthrus-Qwen3: up to 7.8×tokens/forward on Qwen3, identical output distribution (news.ycombinator.com)

2026-05-15 | get AI Model Tokenizer Upgrade → | tags: orthrus, qwen3, diffusion

15.

Microsoft CTO confesses that 30-year-old code from the mid-90s still forms the bedrock of Windows 11 — ancient Win32 API still the backbone, but CTO says it's 'more relevant than ever in 2026' (tomshardware.com)

2026-05-08 | by Mark Tyson | get Windows 11 Win32 API Book → | tags: windows 11, win32 api, microsoft azure

16.

Diskless Linux boot using ZFS, iSCSI and PXE (news.ycombinator.com)

2026-05-07 | get ZFS Boot Server Kit → | tags: zfs, pxe, iscsi

17.

Unofficial Notepad++ macOS app forced to rebrand after creator objects (techspot.com)

2026-05-05 | get Notepad for Mac → | tags: notepad++, macos, andrey letov

18.

Usage-based pricing killing your vibe, here's how to roll your own local AI (news.ycombinator.com)

2026-05-04 | by Tobias Mann | tags: anthropic, github copilot, qwen3.6-27b

19.

Inside Google's TPU V8 strategy, delivering two chips for two crucial tasks at incredible scale — network scales up to 1 million TPUs per cluster, an advantage over Nvidia AI accelerators (tomshardware.com)

2026-04-27 | by Luke James | get Google TPU V8 Developer Kit → | tags: google tensor processing units, tpu 8t, tpu 8i

20.

DJI’s new Lito Series beginner drones start at less than $400 (theverge.com)

2026-04-23 | by Andrew Liszewski | get DJI Lito Series Drone → | tags: dji, lito series, lidar

21.

This ultra-bright Google TV projector promises not to be a pain to set up (androidauthority.com)

2026-04-23 | get Google TV Projector Stand → | tags: google tv, jmgo, n3 ultimate

22.

Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model (news.ycombinator.com)

2026-04-22 | get AI Coding Assistant Tool → | tags: qwen3.6-27b, flagship, coding

23.

Actively exploited Apache ActiveMQ flaw impacts 6,400 servers (bleepingcomputer.com)

2026-04-21 | get Apache ActiveMQ Security Patch → | tags: apache activemq, cve-2026-34197, shadowserver

24.

Even 'uncensored' models can't say what they want (news.ycombinator.com)

2026-04-20 | get AI Text Generation Software → | tags: uncensored models, google gemini, qwen3.5

25.

We got 207 tok/s with Qwen3.5-27B on an RTX 3090 (news.ycombinator.com)

2026-04-20 | get RTX 3090 GPU Cooling Fan → | tags: qwen3.5, rtx 3090, hugingface

26.

Kimi K2.6: Advancing open-source coding (news.ycombinator.com)

2026-04-20 | get Kimi K2.6 Coding Robot → | tags: kimi, open-source, qwen3.5

27.

Qwen3.6-Max-Preview: Smarter, Sharper, Still Evolving (news.ycombinator.com)

2026-04-20 | get AI Language Model Keyboard → | tags: qwen3.6-max, ai model, tech preview

28.

Theseus, a Static Windows Emulator (news.ycombinator.com)

2026-04-20 | get Windows Emulator Software → | tags: theseus, retrowin32, emulation

29.

How Motorola’s 2N2222 and 2N3904 transistors became the default NPNs (news.ycombinator.com)

2026-04-17 | get Electronics Starter Kit → | tags: motorola, 2n2222, 2n3904

30.

CISA flags Apache ActiveMQ flaw as actively exploited in attacks (bleepingcomputer.com)

2026-04-17 | get ActiveMQ Messaging Broker → | tags: apache activemq, cisa, cve-2026-34197