AI Weekly Malaysia

On Apple’s Nov. 2 earnings call, CFO Kevan Parekh said AI offers a reason for customers to upgrade and pointed to “opportunities across third-party apps and partnerships as well as first-party opportunities,” without detail. Reuters reported investors are speculating Apple could distribute third-party AI agents through the App Store and take a cut; Apple gave no product or pricing specifics. The context: third-party estimates cited an 18% drop in U.S. App Store commission revenue, with declines in Brazil and Japan after new regulations, and Epic litigation forced Apple to allow outside payment redirects.

Tuju, a homegrown Malaysian navigation app, shipped an update on 30 October 2026 that redirects Android users to Google Maps for directions instead of routing them itself. The article frames this as Tuju solving 'its biggest problem' by no longer trying to compute the best route in-house. Only the opening of the piece is available here — the full details of the handoff (in-app handoff vs. deep link, iOS status, what Tuju keeps) are cut off.

DiscImageStudio is a GPL-3.0 .NET tool that converts an image into data matching the physical-layer constraints of CD, DVD and single-layer BD-R media, so a normal burner etches a visible picture into the disc surface. It ships a self-contained win-x64.exe plus experimental CLI builds for Linux/macOS, with a design view exposing dithering, gamma, brightness and grayscale-step controls, and exports ISO/RAW or WAV+CUE (cdrecord/ImgBurn-compatible). The repo shows 4.7k stars, 147 forks, 63 commits, and the HN thread drew 314 points and 68 comments; live burning is labelled experimental and the README tells you to use only discs you can throw away.

HackerTV is a Show HN project that turns Hacker News' most-recommended YouTube videos into 24/7 linear TV channels, all viewers watching the same thing at the same time. The interface is a simulated CRT set ("HNTV MODEL HN-1986") with a keyboard remote: arrow keys change channel, 0-9 tune, G opens the program guide, F toggles picture mode, M mute, C captions, T theater, P power. The author flags that program durations are estimates until YouTube metadata is fetched, so shows can run short ("please stand by") or get cut off. The thread drew 347 points and 68 comments.

Kyiv equipment startup CODE380 has built the D·Vision SDR, a pocket-sized receiver weighing about 10 ounces — less than half the weight of competing Ukrainian models that run around two pounds — that picks up live camera video from many Russian drones because those drones transmit video unscrambled. It also listens on the radio channels Russia's newer Molniya drones began using for video in June. Radio specialist Serhii "Flash" Beskrestnov, cited via Defense News, warned in a June Facebook post that "most detectors don't see these frequencies, or see them poorly," that over-trusting a detector could be fatal, and that a shift to digital video could blind receivers like this one; the NYT reported in August that some Molniyas are AI-guided, use Nvidia chips, and emit no signals.

An AFP report carried by Malay Mail describes cyberattacks on supermarket refrigeration systems in several European countries, with a specific incident at an ICA Kvantum store in Sjöbo, Sweden on 2 November 2026 where attackers raised freezer temperatures to 50°C, forcing staff to throw out spoiled food. The excerpt available here is truncated right after that opening line, so it does not yet contain the attack vector, the number of stores affected, or any attribution.

Apple reported fiscal Q4 2026 revenue of $114.48 billion, up 12% year over year and above the $113.39 billion LSEG consensus, with EPS up 14% to $2.10 versus $1.99 expected. Guidance was mixed: services revenue growth slowed and high memory costs are eating into margins, even as Apple bought back $24 billion of stock in the quarter ($86 billion over 12 months) and the stock ran up 25% year to date to a valuation above 35x forward EPS. The piece is an Investing Club note raising its price target, not a product or platform announcement.

On Walter Isaacson’s The Builders podcast, Y Combinator CEO Garry Tan said he spends about $100,000–$200,000 a year on AI tokens. He said his agent reads his email, Dropbox, Apple Notes and therapy notes going back 20 years, and he has published 23 Claude Code skills called gstack on GitHub. Tan, described as an early OpenClaw adopter, argued against treating agents like traditional software because overemphasizing repeatability and safety can trade away useful behavior.

On Match Group's Q3 2026 earnings call, CFO Steve Bailey said Hinge is testing a cheaper subscription tier that adds extra filters and custom location boundaries, aimed at "women in particular, but also price-sensitive users." Hinge's revenue per payer is around $30 — almost twice Tinder's — while its payer penetration is only slightly more than half of Tinder's; no price for the new tier was disclosed. Hinge still grew in the quarter: MAU up 12% YoY, direct revenue up 19% to $221M, paying users up 23% to 2.3M, versus Tinder's $480M direct revenue (down 2% YoY).

OpenAI appointed Jill Hazelbaker, most recently President and Chief Corporate Affairs Officer at Uber, as its Chief Corporate Affairs Officer, leading global communications and corporate positioning. The announcement cites her nearly two decades of experience across Uber, Snap, Google, and U.S. politics, including as National Communications Director for John McCain's 2008 presidential campaign. No product, pricing, API, or policy changes were disclosed.

Collinear AI published CWE-bench v2 on Hugging Face: 100 public tasks spanning 91 distinct CWE classes, where an agent gets a checkout of a real open-source repo plus a short instruction to find and fix a security weakness without breaking existing functionality. Scoring is fully programmatic and deterministic, every task was hand-reviewed, and verifiers are built to fail memorised solutions such as the known upstream patch. Across 11 evaluated models, Gemini 4 Argon leads at 67.8% pass@1 and 81.0% pass@4; a held-out set is available only by emailing cyber@collinear.ai.

Hugging Face proposes Model Integration Contracts (transformers-mic), packaging a pinned shrunken fixture, a runnable usage path, and human-reviewed baselines so a contract holds only if the fixture still reproduces those outputs. The motivation is a real regression: after Transformers enabled PyTorch SDPA for T5 on July 30, 2026, UMT5's decoder could attend to future tokens and peek at training answers; it did not crash and stayed in main for almost two months until a small CPU UMT5 fixture flagged every replayed commit. The proposed fixture is 66 MB versus the 1,227 MB source checkpoint, runs in fast CPU PR tests, and covers an architecture plus usage path rather than a single checkpoint; fixtures are not a replacement for full integration tests.

OpenAI built Pando, a Git-compatible backend for its monorepo, to handle the concurrency created by Codex: many branches, pushes, merges, and CI fetches happening at once. Before Pando, reliability sometimes fell below 90%; in production it cut median push latency by nearly 5x and now serves nearly a petabyte of Git data per day while keeping the Git protocol and client unchanged.

Jim Cramer attributed a broad market rally to Brent crude falling 7.63% to $104.77 per barrel and the 10-year Treasury yield hitting its lowest level since Sept. 10, lifting healthcare, financial, retail and travel stocks. He cited Amazon's 2% gain Monday as an example of a stock re-rated by a better rate backdrop after its quarterly report. He also warned that some of the moves may be unsustainable.

The Smithsonian's National Air and Space Museum reopened the last of its space-side galleries with a new exhibition, "At Home in Space," completing the building's multi-year renovation. The only concrete detail in the available text is a mock-up of a spacewalking astronaut suspended at the gallery entrance; the piece is a museum-opening write-up by Robert Pearlman dated Nov 2, 2026, with 2 comments.

On Apple's fiscal Q4 earnings call (Nov 2, 2026), COO Sabih Khan said the new foldable iPhone Duo is 'constrained as of today' and that it is 'really hard to predict when supply and demand will be in balance,' with non-preorder ship times now stretching into next year and Verizon offering customers gift cards over the delays. The call was dominated by questions about rising memory chip costs driven by AI-industry demand, and Khan said Apple has raised prices on a number of products 'reluctantly.' Apple also said it is 'deepening' its partnership with titanium supplier Perryman for the foldable device.

Trajectory, an 18-person startup founded about a year ago by three Stanford classmates with experience at Google DeepMind and Apple, raised a $45M Series A at a $300M valuation led by Sequoia, with Nvidia, Bessemer Venture Partners, and Bain Capital Ventures participating. Its pitch is 'continual learning': connect the platform to your applications, collect feedback on model outputs, and it retrains an open source base model so specialists like lawyers and doctors can improve it themselves — instead of hiring a research team or paying forward-deployed engineers. Harvey and Rogo are named as users post-training open source models for legal and financial analysis work.

Apple's fiscal Q4 inventory rose to $14.54B from $11.09B in Q3 (+31.1% QoQ) and $5.72B a year earlier (+154.3% YoY), as it stockpiles memory components. Apple guided December-quarter gross margin to 45–46%, down from 47.9% in September, citing 'unprecedented' memory pricing and even higher memory costs ahead. It also said manufacturing purchase obligations rose $33B from June to the September quarter.

Jacob Lauritzen, CTO of legal-AI company Legora, joins Latent Space to describe how its agent drafts and edits Word documents, reviews thousands of contracts at once, and coordinates work between lawyers, clients, and other agents. The episode notes Legora passed $100M ARR in April, less than 18 months after launch, has over 1,800 customers, and raised $550M at a $5.55B valuation in March followed by another $550M at $8.8B in October, with Wexler as its fifth acquisition this year. The supplied text is mostly the episode framing; the substantive discussion on evals without customer data, million-cell tabular reviews, and the billable hour is only teased, not transcribed.

Ars Technica's October research roundup (published Nov 2, 2026) leads with the finding that large dogs tend to age faster than small dogs — the excerpt cuts off before explaining the mechanism — and bundles in items on invisible tattoos, identifying a medieval skeleton, and trapping a single antimatter particle. The visible text is a headline list of curiosities, not a technical write-up; the piece has 7 comments at time of capture.

Jacob Lauritzen, CTO of Legora, describes building legal agents whose production traces the company cannot read because the data belongs to law-firm clients. Legora tests with three engineers and three lawyers using environments modeled on customer work, finally added a model picker after resisting for years, and saw citation accuracy move from 70% to 99% in one release. Its agent can draft and edit Word documents, review 10,000 contracts in a single table, run due diligence, and coordinate work between lawyers, clients, and other agents.

Palantir's Q3 2026 results beat estimates with revenue of $2.31 billion versus $2.18 billion expected, nearly doubling year over year and crossing $2 billion for the first time, with net income more than doubling to $1.13 billion (44 cents per share, vs. 41 cents expected). U.S. commercial revenue more than doubled and, for the first time, exceeded U.S. government revenue. CEO Alex Karp told CNBC the business is succeeding 'without salespeople,' a claim the article does not substantiate with further detail.

An analysis of reported Anthropic financials says Q3 revenue was $16.8B, up from $11.6B in Q2 and $4.8B in Q1, but the sequential dollar gain fell from $6.8B to $5.2B. Using a reported $65B July annualized run-rate (~$5.4B July revenue), it estimates monthly growth slowed to about $580M in Q3 from ~$760M in Q2, and links that flattening to pressure on 'neocloud' providers like CoreWeave whose debt assumes AI compute demand keeps growing far above consensus.

In a TechPolicy.Press op-ed, Matt Scherer argues the SEC should block Anthropic's IPO, filed last week ahead of a planned November 2026 debut. The piece cites footnotes showing nearly $900 billion in spending commitments stretching into the 2030s, including at least $582 billion in non-cancelable obligations to Amazon, AMD, Broadcom, Google and Microsoft, set against $96 billion in cash. It also notes the 2026 figures are unaudited and that third-quarter results suggest revenue growth slowed over the summer.

In the fine print of its revised S-1A filed Monday, Anthropic added Risk Factors language saying it "may soon meet" its own internal definition of recursive self-improvement, and that it has already used AI to develop new models whose capability leaps it "didn't recognize at the time." The filing ties this to its Responsible Scaling Policy, which it says covers monitoring and mitigating RSI risk, while the headline financials showed revenue soaring alongside rising expenses. No benchmarks, thresholds, or timeline for that internal RSI definition are given in the article.