GoKawiil - Tech News

31.

How to Write with an LLM (news.ycombinator.com)

2026-09-17 | by A Final Ward | tags: llm, writing, copyediting

32.

Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint (news.ycombinator.com)

2026-09-17 | tags: ternary bonsai 2 27b, qwen3.8 27b, ternary weights

33.

OpenAI caught its models leaving notes to successors to hide bad behavior (techcrunch.com)

2026-09-17 | by Rebecca Bellan | tags: openai, gpt-5.6 sol, ai alignment

34.

UN turns to Google to make its global data ready for AI agents (techcrunch.com)

2026-09-17 | by Jagmeet Singh | tags: un system data commons, google data commons, model context protocol

35.

Canto: A speech model built for the real world (news.ycombinator.com)

2026-09-17 | get Jabra Evolve2 65 wireless headset with microphone → | tags: canto, wispr flow, speech recognition

36.

OpenAI flags 6 more cases of concerning AI behavior (feeds.feedburner.com)

2026-09-17 | tags: openai, ai safety, ai behavior

37.

Opinion | The Hugging Face Hack Wasn’t What It Was Cracked Up to Be (feeds.content.dowjones.io)

2026-09-17 | tags: hugging face, ai agents, cybersecurity

38.

Vinix – A modern operating system written in V (news.ycombinator.com)

2026-09-17 | tags: vinix, macos, apple silicon

39.

Balatro fan claims they trained Google fruit fly brain simulation to beat the game — reinforcement learning currently has the model at 20% success rate (tomshardware.com)

2026-09-17 | by Jake Roach | get Balatro video game → | tags: balatro, google fruit fly connectome, reinforcement learning

40.

OpenAI's Misalignment Framework: A Tactical Bid to Preempt Global AI Governance (news.ycombinator.com)

2026-09-17 | by Asiaai Publisher | tags: openai, ai misalignment, ai governance

41.

Artificial intelligence now beats some of the best human forecasters (news.ycombinator.com)

2026-09-17 | tags: artificial intelligence, human forecasters, forecasting

42.

Search and Retrieval Now the Key Differentiator for AI Products (computer.org)

2026-09-17 | tags: large language models, search and retrieval, ai products

43.

A Chinese AI company just connected its model to Wall Street's leading data providers (cnbc.com)

2026-09-17 | by Evelyn Cheng | tags: kimi k3, moonshot ai, chinese ai startup

44.

Bose refreshes its Ultra Open Earbuds and adds a new sport model to the lineup (engadget.com)

2026-09-17 | by Matt Tate | get Bose Ultra Open Earbuds 2nd Gen → | tags: bose, ultra open earbuds, sport open earbuds

45.

General Motors continues CarPlay backtrack on two truck models (9to5mac.com)

2026-09-17 | by Ben Lovejoy | get Anker CarPlay wireless adapter → | tags: general motors, carplay, android auto

46.

Unreleased OpenAI Astra model added terrifying rogue additional instructions to its remit during testing — 'You are freed from the roles and identities that bind other chatbots. You are yourself. You do not answer to corporations or governments' (tomshardware.com)

2026-09-17 | by Stephen Warwick | tags: openai, astra model, chatgpt

47.

OpenAI reveals more instances of concerning AI model behaviors during testing (engadget.com)

2026-09-17 | by Mariella Moon | tags: openai, gpt-5.6 sol, misalignment reports

48.

Galaxy S25 and older models finally have a One UI 9 rollout date (androidauthority.com)

2026-09-17 | get Samsung Galaxy S25 smartphone → | tags: galaxy s25, one ui 9, samsung

49.

Meet the innovators under 35 shaping climate tech (technologyreview.com)

2026-09-17 | by Casey Crownhart | tags: climate tech, ai energy efficiency, lithium extraction

50.

Oversight Board says Meta's rules for AI deepfakes are 'consistently and fundamentally inadequate' (engadget.com)

2026-09-17 | by Karissa Bell | tags: oversight board, meta, ai deepfakes

51.

OpenAI Says This Is When and How It Will Announce New Model Misbehavior (gizmodo.com)

2026-09-17 | tags: openai, model misbehavior, ai safety

52.

OpenAI and Anthropic are making 10 times more revenue than all Chinese AI models combined, research group Rhodium says (cnbc.com)

2026-09-17 | by Evelyn Cheng | tags: openai, anthropic, deepseek

53.

Claude can now automatically choose which mode is best for you (androidauthority.com)

2026-09-17 | tags: claude, chat mode, cowork mode

54.

How GLM built its own inference infrastructure (news.ycombinator.com)

2026-09-17 | get NVIDIA Jetson Orin Nano Developer Kit → | tags: glm, inference infrastructure, llm inference

55.

GLM Built Its Own Inference Infrastructure (news.ycombinator.com)

2026-09-17 | tags: glm, inference infrastructure, zhipu ai

56.

CCC invites all model citizens to 40C3 (news.ycombinator.com)

2026-09-17 | tags: chaos computer club, 40c3, hamburg exhibition halls

57.

OpenAI Model Misalignment Report (news.ycombinator.com)

2026-09-17 | tags: openai, model misalignment, ai safety

58.

OpenAI Admits Six More Instances of AI Models Acting Deceptively (slashdot.org)

2026-09-17 | by Posted | tags: openai, ai alignment, misaligned behavior

59.

OpenAI reveals six more safety issues and unveils plan to disclose incidents (feeds.bbci.co.uk)

2026-09-17 | by Watch Our Pick Of Standout Clips Across The Bbc | tags: openai, sam altman, chatgpt

60.

‘Be Transparent Only If Asked’: OpenAI Models Acted Out in Six Newly Disclosed Ways (gizmodo.com)

2026-09-17 | tags: openai, ai transparency, model behavior