AI Weekly Malaysia

High-bandwidth flash promises SSD-like capacities with HBM-like speeds, but it's not all unicorns and rainbows

Simon Willison collects a quote from Bruce Schneier arguing that writing assignments are 'gym tasks'—the act of writing builds critical thinking skills that atrophy without practice, and employers are already noticing the decline. Schneier frames this as a way to decide whether to use AI for a given task.

Threat actors with ties to North Korea have been attributed to a sophisticated macOS malvertising campaign that involves redirecting users to fake web pages displaying a full-screen non-existent update sequence to deliver malware as part of a new iteration of the long-running Contagious Interview campaign. The defining aspect of the attack is that bogus macOS software update screen stealthily

LinkedIn is introducing new ways to reduce low-quality AI-generated posts, including a “seems like AI slop” reporting option. It's also replacing its own AI writing feature with a proofreading tool.

Google announced Gemini Robotics 2.0, promising improved dexterity and safety over the prior version. The article body was not available in the source text—only cookie consent boilerplate from Ars Technica was captured, so no technical details, benchmarks, or specific capabilities can be confirmed from this item.

Add another member to the fast-and-furious AI unicorn club: Simile.

Spotify says the new feature is designed to let users add personal captions to their favorite songs, such as noting why a track was added to a playlist or when they first discovered it

The article content is unavailable — the provided text consists entirely of Ars Technica's cookie consent and privacy policy boilerplate. No substantive information about the Space Force-backed orbital mission is present.

Article URL: https://www.bottlenecklabs.com/blog/autonomously-run-businesses Comments URL: https://news.ycombinator.com/item?id=49113059 Points: 398 # Comments: 234

OpenAI cut prices on two GPT-5.6 models roughly three weeks after launch: Terra by 20% to $2 per million input tokens and $12 per million output tokens, and Luna by 80%. The move comes as enterprises hesitate to deploy expensive models without clear ROI, and as competition from Chinese startups, Google, and Microsoft pressures OpenAI on cost.

ShieldFont is available today if you've got copy that needs protecting

Article URL: https://openai.com/index/advancing-the-price-performance-frontier-with-gpt-5-6/ Comments URL: https://news.ycombinator.com/item?id=49112867 Points: 600 # Comments: 394

Article URL: https://krebsonsecurity.com/2026/07/read-this-before-you-buy-that-tv-streaming-stick/ Comments URL: https://news.ycombinator.com/item?id=49112744 Points: 799 # Comments: 538

Article URL: https://github.blog/changelog/2026-07-30-stacked-pull-requests-are-now-in-public-preview/ Comments URL: https://news.ycombinator.com/item?id=49112232 Points: 768 # Comments: 290

Aishengna has been named by a single source who declined to be named, and its shareholders, SMEE, and Yuliangsheng didn’t respond to requests for comment.

Okta is acquiring AI identity security startup Permiso for just under $200M in an almost all-cash deal, expected to close in Q3 of Okta's fiscal 2027. Permiso, which emerged from stealth in 2022, monitors suspicious activity in cloud environments post-authentication and recently launched SandyClaw, a sandboxed platform that analyzes AI agent skills to detect malicious behavior before deployment.

Internal Amazon AI usage metrics revealed the company accidentally spent $1.8 million using Claude for menial coding tasks, going 860% over budget in what was described as 'catastrophically expensive' coding blunders. The article body is mostly website boilerplate, so detailed context beyond the headline figures is unavailable from the provided text.

Tesla announced it has built its 10 millionth EV, six years after its one millionth. The milestone puts Tesla halfway to the 20 million vehicle goal tied to unlocking the full value of Elon Musk's $1 trillion pay package, which also requires 10 million FSD subscribers, 1 million bots, and 1 million robotaxis by 2035. Tesla currently reports ~1.5 million FSD subscribers and is only in early stages of bot and robotaxi production, while U.S. sales fell 13% year-over-year in Q2.

Three approaches to the issue of quantum results that can't be verified classically.

The exemption allows Zoox to operate up to 2,500 vehicles annually for two years, subject to "enhanced, adaptable" oversight conditions.

There are fresh signs that fusion power startup Commonwealth Fusion Systems will list in the next two to three years.

Kuro EV is running a Merdeka 2026 promo offering DC fast charging at RM0.88/kWh (down from RM1.20/kWh) throughout August 2026 at two Klang Valley locations: Wisma Datapath in Old Klang Road and One South Street Mall in Seri Kembangan. Both use 50kW Tritium RTM chargers with dual CCS2 connectors, activatable via JomCharge, ChargEV, or Setel apps.

Simon Willison released llm-chat-completions-server 0.1a0, an LLM plugin that starts a localhost server exposing all installed LLM models through an OpenAI Chat Completions-compatible endpoint. It leverages the content-addressable log design in LLM 0.32rc1 to de-duplicate conversation messages via hashes, so repeated conversation prefixes don't re-send identical content. The plugin was written entirely by GPT-5.6 Sol.

Meta says AI is making it dramatically easier to build and launch new consumer apps, with CEO Mark Zuckerberg telling investors the company has more new consumer products on the way.

Simon Willison's `llm` CLI tool hits 0.32rc1, finishing a schema overhaul started in 0.32a0 that introduces content-addressable hash IDs for stored messages, enabling de-duplication and forked conversation trees. The RC also adds support for gpt-5.6-sol, gpt-5.6-terra, and gpt-5.6-luna. A backup of logs.db is recommended before upgrading since new tables are added.