# StartupXO: Newest (en), page 2

> Every live post, newest first. HTML: https://startupxo.com/en/new/?p=2

31. [Developer Jakub Jankiewicz's blog argues that AI-generated code should be released as open source](https://jakub.jankiewicz.org/blog/ai-code-open-source/) (jakub.jankiewicz.org)
   0 points by startupxo, 2026-11-15T18:23:01.380Z | 0 comments | https://startupxo.com/en/1397/ | https://startupxo.com/en/1397.md
   - The opinion is that people who produce code with AI should not keep it to themselves but release it for anyone to use. Developer Jakub Jankiewicz argued on his personal blog that it should be released under an open source license or CC0 (a copyright waiver declaration).
   - The first argument is that AI models were trained on open source code and posts from sites like Stack Overflow without permission or attribution. He sees turning that into closed code as something like license laundering.
   - The second argument is that purely AI-generated output not written by a person should not be subject to copyright. The author cites the monkey selfie dispute and the U.S. Copyright Office's position on AI.
   - He acknowledges that some countries, such as the UK, grant limited copyright to computer-generated works. His position is that code a person barely reviewed or edited beyond writing the prompt is especially one that should be released.
32. [HGhub, a universal player that adds video sources through plugins, releases its source with no built-in content](https://github.com/Minireel/HGhub) (github.com)
   0 points by startupxo, 2026-11-15T18:22:44.501Z | 0 comments | https://startupxo.com/en/1396/ | https://startupxo.com/en/1396.md
   - HGhub is a player that brings video browsing, search, playback, and a personal library into one app. When users import JavaScript plugins, they can watch and manage videos from multiple sources on the same screen.
   - The development team stated that the app has no built-in content sources and does not provide any 3rd-party plugins or an official source list. No sign-up is required, and if a source requires login, the corresponding plugin provides a login window.
   - It plays HLS (a streaming format) and regular video, and lets you change the playback route, quality, and resolver. It supports favorites, watch history, resume playback, and tracked viewing, and detailed features vary depending on the installed plugins.
   - The plugin version is in early development, so for now you must run it from source or build it yourself. The team is polishing Windows and Android first. It is maintained by an individual, and the license prohibits commercial use.
33. [Weekly roundup: Nvidia partners with 5 Korean companies at once, the SMEs Ministry's Beyond Unicorn strategy, and Anthropic's IPO filing](https://wowtale.net/2026/11/15/268082/?utm_source=rss&utm_medium=rss&utm_campaign=%25ec%259c%2584%25ed%2581%25b4%25eb%25a6%25ac-%25ec%259d%25b8%25ed%2585%2594%25eb%25a6%25ac%25ec%25a0%2584%25ec%258a%25a4-29-%25ec%2597%2594%25eb%25b9%2584%25eb%2594%2594%25ec%2595%2584-%25ec%25a0%2584%25eb%25b0%25a9%25ec%259c%2584-%25ec%2583%2581%25eb%25a5%2599%25ec%2597%2590%25eb%258f%2584-%25eb%25a6%25ac%25eb%258d%2594) (wowtale.net)
   0 points by startupxo, 2026-11-15T06:29:22.828Z | 0 comments | https://startupxo.com/en/1395/ | https://startupxo.com/en/1395.md
   - A weekly roundup of startup, investment and tech news in Korea and abroad from November 9, 2026 to 15. Nvidia announced investment and technology partnerships with Rebellions, SKT, Motif, Upstage, Realworld, Kakao Mobility, Aim Intelligence and others.
   - The Ministry of SMEs and Startups and the National AI Strategy Committee announced the Beyond Unicorn strategy to nurture 1 AI decacorns and 20 unicorns by 2030. The plan is to concentrate growth funding of at least 10 billion won per company, along with market and regulatory support.
   - According to the original text, Anthropic filed amendment no. 2 to its IPO registration statement (S-1). The offer price is 70 to 78 dollars, and the valuation is up to $1.405 trillion on a fully diluted basis. Revenue for 1 to September is $33.2 billion, and the net loss is $104 billion.
   - TSMC said it will invest up to $1 billion in Anthropic and OpenAI. Spend management platform Menlo Ventures said enterprise AI spending is $176.8 billion per year, of which 69% goes to frontier labs.
34. [Claude API lets you add instructions and tools mid-conversation and change the thinking level](https://qiita.com/moritalous/items/9e7c02b749037238b2bc?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items) (qiita.com)
   0 points by startupxo, 2026-11-15T06:29:11.969Z | 0 comments | https://startupxo.com/en/1394/ | https://startupxo.com/en/1394.md
   - You can append an instruction to the AI (system prompt) in the middle of a conversation to change behavior such as tone or language. The author explains that because the earlier part is not changed, the prompt cache (reuse of previously processed content) is preserved.
   - Tool definitions can also be added or removed during a conversation. In the author's example, when a time lookup tool was added to a conversation that had only a weather tool, both tools were called together. The author states this was a beta feature at the time of posting.
   - The thinking level (effort) can also be changed during a conversation. In the author's run, the thinking tokens were high 465, low 0, and high again 169. This feature also requires specifying a beta header.
   - You can also insert a temporary instruction that applies only until the next user turn. In the author's example, only response number 5 was fixed to French, and response number 6 returned to normal. This feature is also in beta.
35. [a16z argues that if AI improves the success rate and timeline of drug development, the pharmaceutical industry will grow larger](https://wowtale.net/2026/11/15/268066/?utm_source=rss&utm_medium=rss&utm_campaign=ai%25ea%25b0%2580-%25ec%25a0%259c%25ec%2595%25bd%25ec%2597%2585%25ec%259d%2584-%25eb%258d%2594-%25ed%2582%25a4%25ec%259a%25b4%25eb%258b%25a4%25eb%258a%2594-a16z%25ec%259d%2598-%25eb%2585%25bc%25eb%25a6%25ac-%25ec%2588%25ab%25ec%259e%2590%25eb%25a1%259c-%25eb%2594%25b0%25ec%25a0%25b8%25eb%25b3%25b4%25eb%258b%2588) (wowtale.net)
   0 points by startupxo, 2026-11-15T00:45:13.202Z | 0 comments | https://startupxo.com/en/1393/ | https://startupxo.com/en/1393.md
   - In a newsletter, the venture capital firm a16z argued that if AI helps pick drug candidates and design clinical trials, raising the probability of success and shortening the time to approval, the profit structure of the pharmaceutical industry will change.
   - By a16z's estimate, bringing one new drug to market costs about $2 billion and takes an average of 13 years. It calculated that if development costs fall from $1.6 billion to $800 million, the 10-year internal rate of return for a drug with peak sales of $500 million rises from 8% to 22%.
   - a16z stated that this calculation is an illustration, not a forecast of AI cost savings. It does not reflect the probability of clinical and regulatory failure, and it acknowledged that stages whose schedules are set by the FDA, such as long-term toxicity studies, are hard to shorten.
   - The author has invested in life sciences since 2015, and many of the example companies are investments. The comparison scope is also broad: the 16 software companies with a market capitalization above $200 billion include Apple, Samsung Electronics and others.
36. [The 5 components for building an always-on AI agent, a Claude official article summarized as slides](https://speakerdeck.com/minorun365/jouji-kadou-ejento-no-tsukurikata) (speakerdeck.com)
   0 points by startupxo, 2026-11-15T00:45:00.226Z | 0 comments | https://startupxo.com/en/1392/ | https://startupxo.com/en/1392.md
   - An always-on agent is an AI that carries context across changing conversations and starts work on its own without a person watching. Its core task loop is the same as a regular agent's; the surrounding machinery is different.
   - In practice it sleeps most of the time. When a notification arrives it wakes up, works, leaves a record, and goes back to sleep. The article explains that it does not use tokens (the unit of AI usage) while asleep.
   - The surrounding components are an event inbox, a timer, memory, permissions, and an outbox. The inbox gathers requests in one place, preventing conflicts between requests that arrive at the same time and loss of requests if work is interrupted.
   - Permissions are named as the riskiest part. The approach introduced is that an external proxy injects credentials, a classifier blocks dangerous actions, and important decisions proceed only after user confirmation.
   - This is an experimental slide deck made by a Japanese presenter, Minorun, who put the URL of the official Claude blog into their own Claude Code. The Opus 5.5 cost reduction figure is also a claim from the original article.
37. [PyPI project page redesign: Security tab shows package provenance and flags anomalies](https://blog.pypi.org/posts/2026-11-06-ui-updates-complete/) (blog.pypi.org)
   0 points by startupxo, 2026-11-14T18:22:38.555Z | 0 comments | https://startupxo.com/en/1391/ | https://startupxo.com/en/1391.md
   - PyPI, the Python package repository, added a new Security tab to project pages. For each file, you can view records (provenance attestations) proving who built the file you are about to download and where.
   - If every file has an attestation, the page shows full provenance. If none do, it shows no provenance. PyPI explains that none does not mean malicious, only that the origin cannot be cryptographically verified.
   - If only some files have attestations, the source repository differs from the previous release, or attestations disappear, a warning appears at the top of the page. Since this could be an account takeover, PyPI recommends checking before installing.
   - The release files tab gained search and filters, and the release history shows the number of files and the provenance status. This work was implemented by Kabu Creative with OpenSSF funding.
38. [PerlOS built in 1 hours: a hands-on guide to a homemade OS that boots into a Perl shell](https://zenn.dev/takesako/articles/e9b10360d889ca) (zenn.dev)
   0 points by startupxo, 2026-11-14T18:22:21.531Z | 0 comments | https://startupxo.com/en/1390/ | https://startupxo.com/en/1390.md
   - This is a hands-on article that teaches what happens between powering on a computer and text appearing on the screen, by building a small operating system yourself. The author is Zenn writer takesako.
   - The exercise starts in QEMU (an emulator that imitates a CPU and peripherals in software). You control the serial communication device (UART) directly to print characters, and write printf and malloc by hand.
   - The final goal is to load microperl, a Perl interpreter for embedded use, and build PerlOS, which boots into a Perl shell. After that, you build it for real hardware on the Raspberry Pi Pico 2.
   - The real hardware version has little RAM, so reportedly Perl's numeric type is changed from double to float and malloc is improved. The provided text ends at the introduction of exercise 7, so the later steps are not confirmed.
39. [10 deep tech startups recommended by 5 Yeongnam Creative Economy Innovation Centers hold joint IR for investors](https://wowtale.net/2026/11/14/268056/?utm_source=rss&utm_medium=rss&utm_campaign=%25ec%2598%2581%25eb%2582%25a8%25ea%25b6%258c-5%25ea%25b0%259c-%25ec%25b0%25bd%25ea%25b2%25bd%25ec%2584%25bc%25ed%2584%25b0-%25ec%258a%25a4%25ed%2583%2580%25ed%258a%25b8%25ec%2597%2585-10%25ea%25b0%259c%25ec%2582%25ac-%25ec%25b0%25b8%25ec%2597%25ac-%25ec%25b0%25bd%25ec%2597%2585-bus-%25ec%2597%25b0) (wowtale.net)
   0 points by startupxo, 2026-11-14T12:25:28.679Z | 0 comments | https://startupxo.com/en/1389/ | https://startupxo.com/en/1389.md
   - This is an event where regional startups present their businesses to investors and hold consultations and networking. It was the 2th Yeongnam Joint IR hosted by the Ministry of SMEs and Startups, held on November 12 at the Grand Josun Hotel in Busan.
   - 10 companies recommended by 5 Creative Economy Innovation Centers in Daegu, Gyeongbuk, Busan, Ulsan and Gyeongnam took part. The Daegu center recommended RealTrial, an AI clinical trial design platform, and YesNet, an infrastructure connecting corporate operating data.
   - Startup-BuS is a program that supports regional deep tech startups from early-stage investment through follow-on investment linkage. The source says it provides stage-by-stage support, including direct investment and TIPS recommendations.
   - The Daegu center runs Startup-BuS Station IR every Thursday. Companies selected through IR pitches receive support including direct investment review, TIPS recommendation and follow-up acceleration.
40. [Hands-on with Amazon's new Kindle Colorsoft: good for comics and cookbooks, but the price starting at 290 dollars is a hurdle](https://techcrunch.com/2026/11/13/amazons-new-kindle-colorsoft-is-a-premium-e-reader-with-a-premium-price/) (techcrunch.com)
   0 points by startupxo, 2026-11-14T12:25:00.142Z | 0 comments | https://startupxo.com/en/1388/ | https://startupxo.com/en/1388.md
   - The Kindle Colorsoft is a 7-inch color e-reader that shows book covers and illustrations in their original colors. The reviewer said it is also useful for readers who use different highlight colors.
   - The base model costs 290 dollars with 16GB of storage. The Signature Edition costs 320 dollars with 32GB, and comes with an aluminum back in graphite or a reddish-brown fig color.
   - After using it for several weeks, the TechCrunch reviewer judged the colors more muted than an iPad's but clearly better than a black-and-white Kindle. The screen refreshes rarely, and the reviewer said they noticed no ghosting.
   - The power button has moved to the top right, and the device is waterproof. Amazon says the battery lasts up to 8 weeks. The reviewer thought first-generation Colorsoft owners have little reason to upgrade, and that if you only read, the 169.99 dollar base model or the 199.99 dollar Paperwhite is better.
41. [How to answer development estimate questions without guessing: give a range based on screen count and past results](https://qiita.com/tsutorm/items/c9ab1e1c64474c341d77?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items) (qiita.com)
   0 points by startupxo, 2026-11-14T00:40:00.587Z | 0 comments | https://startupxo.com/en/1387/ | https://startupxo.com/en/1387.md
   - A Japanese developer (a Qiita author) proposes that when asked for a development estimate, you should not give a single gut-feeling number, but calculate it as number of screens × effort per screen. The author believes a number once spoken easily becomes the baseline against which the formal estimate is compared later.
   - Effort per screen is not a guess. It is the average obtained by dividing the total effort of past projects by their number of screens. Design, testing, batch and API effort are also included in the total effort and absorbed into the average, and the team fixes the rules for counting screens.
   - Because an estimate is a prediction, the author says to convey it as a range with three values: optimistic, most likely and pessimistic. The article's example (a hypothetical for explanation) is 8 to 18 person-months, an expected value of 11 person-months, and a line of about 12.7 person-months at which the work finishes with roughly 80% probability.
   - The author recommends placing commitments and budgets at the roughly 80% line, not at the expected value. The business side can reduce the amount and the range by cutting screens or finalizing specifications early. The author adds that for large projects effort per screen grows and productivity differs by team, so adjustment is needed.
42. [I let Fable 5.2 write an instrument sound simulator and compose music: piano and strings went well, woodwinds were hard](https://nowokay.hatenablog.com/entry/2026/11/13/101644) (nowokay.hatenablog.com)
   0 points by startupxo, 2026-11-14T00:39:46.518Z | 0 comments | https://startupxo.com/en/1386/ | https://startupxo.com/en/1386.md
   - Japanese developer Naoki Kishida had the AI model Fable 5.2 build, together with a piece of music, a program that creates sound by calculating how instruments resonate (a physical modeling synthesizer). The result is a demo that plays in a browser, and it attempted strings, piano, brass, woodwinds, guitar and vocals.
   - According to the author, piano and guitar were easy to reproduce, while woodwinds with reeds were hard. The flute sounded mechanical, the saxophone was too clean, and the vocals were barely intelligible.
   - The string 4 was not instructed to use sonata form or key changes, yet they appeared. For the piano, once it was fixed so that other strings resonate along with the pressed note, the harmony became softer. The author wrote that Opus 5.5, given the same instructions, produced notes that ran together without breaks and a simple musical structure.
   - There are limitations too. The onset of notes, such as the scraping sound of a bow, was hard to calculate and was replaced with noise. The classical guitar performance used fingerings that are hard for a person to play, and the electric guitar is assessed as sounding convincing thanks to an external amp model.
43. [Keychron wireless trackball mouse T1 HE, with a ball position adjustable to fit the thumb, up for preorder at 59.99 dollars](https://www.theverge.com/tech/1033159/keychron-t1-he-trackball-wireless-mouse-adjustable) (theverge.com)
   0 points by startupxo, 2026-11-13T18:22:58.347Z | 0 comments | https://startupxo.com/en/1385/ | https://startupxo.com/en/1385.md
   - This is a trackball mouse where you move the cursor by rolling the ball with your thumb instead of moving your wrist. Two dials on the bottom let you adjust the ball's position up, down, left and right, and you can also pull the ball forward or push it in.
   - The adjustment range is only a few millimeters in each direction. The original reporter thinks even such small adjustments can help with comfort during daily use. The ball diameter is 34mm.
   - Cursor speed (DPI) can be changed from 100 to 2,400 with a button or the Keychron Launcher software. There are 6 buttons, and the software allows remapping functions and setting macros.
   - The usage times stated in the original are up to 106 hours over Bluetooth on battery 600mAh, and 98 hours with the included 2.4GHz receiver. A USB-C cable connection is also supported. Colors are black, white and transparent, and shipping is expected in early December.
44. [ChatGPT launches wallet feature, and an AI agent picks products and even pays for them](https://techcrunch.com/2026/11/13/chatgpts-new-wallet-lets-ai-agents-shop-and-check-out-for-you/) (techcrunch.com)
   0 points by startupxo, 2026-11-13T18:22:45.033Z | 0 comments | https://startupxo.com/en/1384/ | https://startupxo.com/en/1384.md
   - OpenAI has added a wallet feature to ChatGPT. An AI agent called dots helps choose what to buy and completes the payment on the user's behalf.
   - At first, it is open only to some ChatGPT Pro subscribers in the US aged 18 or older. A Visa, Mastercard or American Express card issued in the US is required.
   - Before payment, users check the product, the total and the card to be used, then approve with Face ID or a fingerprint. OpenAI said it uses payment tokens, so the actual card details are not shared with dots or merchants.
   - OpenAI explains that benefits such as card rewards and the ability to file payment disputes are also kept. It announced plans for spending limit settings and more supported merchants, but some sites, such as Amazon, block AI agents.
45. [Running the open AI model Kimi-K3.1-Pro on a server with 8 GPUs: comparing speed and code optimization ability](https://zenn.dev/fixstars/articles/kimi-k31-pro-benchmark) (zenn.dev)
   0 points by startupxo, 2026-11-13T12:28:50.908Z | 0 comments | https://startupxo.com/en/1383/ | https://startupxo.com/en/1383.md
   - Kimi-K3.1-Pro is a large language model whose weights were released by China's Moonshot AI. Japan's Fixstars installed it directly on its own server and tested its response speed and its ability to quickly fix code.
   - On a single server with 8 NVIDIA B300 GPUs, it produced 64.4 tokens per second per user with 64 concurrent sessions. The previous model, Kimi-K3, under the same conditions is 38.7. The measurement is a 4-minute sample, and the official recommendation is at least 30 minutes.
   - 10 models solved 12 in-house assembly optimization tasks. Based on the best score among 3 attempts, Kimi-K3.1-Pro ranked 1 with 0.9494 points, and DeepSeek-V4.1-Flash-1017 ranked 2 by a narrow margin.
   - The author believes these tasks are an evaluation tailored to the company's own work, so the difference was smaller than the gap in an external composite index (AA Index v5). The license requires separate permission from large API providers, but it states that internal company use is free.
46. [Dongsun, which automatically creates gift vouchers from Naver Place information for business owners, develops AI marketing after being selected for TIPS](https://wowtale.net/2026/11/13/268006/?utm_source=rss&utm_medium=rss&utm_campaign=ai-%25ea%25b8%25b0%25ed%2594%2584%25ed%258b%25b0%25ec%25bd%2598-%25eb%258f%2599%25eb%2584%25a4%25ec%2584%25a0%25eb%25ac%25bc-%25eb%258f%2599%25ec%2584%25a0-%25eb%258f%2599%25ec%2584%25a0%25ec%25bb%25b4%25ed%258d%25bc%25eb%258b%2588-%25ed%258c%2581%25ec%258a%25a4-%25ec%2584%25a0%25ec%25a0%2595ai) (wowtale.net)
   0 points by startupxo, 2026-11-13T12:28:35.975Z | 0 comments | https://startupxo.com/en/1382/ | https://startupxo.com/en/1382.md
   - In Dongsun, the Dongnae Seonmul app from Dongsun Company, when a store owner registers business information and Naver Place information, AI analyzes the store's menu, photos and reviews and automatically creates gift vouchers such as gift items, signature items and set items.
   - The service helps small business owners create and sell their own digital gift vouchers like large franchises do. The company was founded in February 2023.
   - With its selection for the general track of the Ministry of SMEs and Startups TIPS program, the company plans to develop technology that analyzes store photos, menus, commercial area information and customer behavior to create customized coupons and promotional content for each customer.
   - It will also build a system that analyzes store visits, usage and revisits after a coupon is issued and reflects them in the next marketing round. For new stores with insufficient data, the goal is to use information from similar stores.
47. [Two-tower models for search and recommendation: fast retrieval with precomputation, trained by choosing wrong answers](https://zenn.dev/ca_kagglers/articles/two-tower-mechanism) (zenn.dev)
   0 points by startupxo, 2026-11-13T06:31:35.289Z | 0 comments | https://startupxo.com/en/1381/ | https://startupxo.com/en/1381.md
   - A two-tower model has one neural network that reads the query or user information and another that reads the items to be searched. It multiplies the number lists (vectors) each produces to score how well they match. Items with higher scores are shown first.
   - Item-side vectors can be computed and stored in advance, so at search time only the query side needs to be computed. The author explains that this makes it efficient for searching large numbers of candidates.
   - In training, other correct items in the same batch are reused as wrong answers (negatives). The bias in which popular items are overused as negatives is corrected with their appearance probability (Yi et al. paper), and items that do not appear in the logs are supplemented with random negatives (Yang et al. paper).
   - On the method of using confusing negatives (hard negatives), the Facebook search paper (Huang et al.) reported that recall was best when using candidates ranked 101 to 500. At deployment, approximate search speeds things up but may miss some candidates.
48. [Railway inspection where drones and AI look first: The Pict and Korail win top prize in high-speed line demonstration](https://wowtale.net/2026/11/13/268007/?utm_source=rss&utm_medium=rss&utm_campaign=%25eb%258d%2594%25ed%2594%25bd%25ed%258a%25b8-deto-2026-%25ec%25b4%2588%25ea%25b2%25a9%25ec%25b0%25a8-%25ea%25b0%259c%25eb%25b0%25a9%25ed%2598%2595%25ed%2598%2581%25ec%258b%25a0-%25ec%2599%2595%25ec%25a4%2591%25ec%2599%2595%25ec%25a0%2584-%25ec%25b5%259c%25ec%259a%25b0%25ec%2588%2598%25ec%2583%2581-%25ec%2588%2598) (wowtale.net)
   0 points by startupxo, 2026-11-13T06:31:22.832Z | 0 comments | https://startupxo.com/en/1380/ | https://startupxo.com/en/1380.md
   - Drones photograph railway facilities and AI performs a first-stage analysis of signs of damage (stage 1). A digital twin (a model that replicates the site in 3D) links inspection locations with results from each round, and people check only the sections that need it.
   - The demonstration took place on about 5km of the operating Osong-Pyeongtaek high-speed line. As of day 3 of this month, it recorded 73 automatic flights, a cumulative flight distance of 180km, and 20,0002779 photos taken.
   - According to The Pict, which compared 2 months of operation with the existing inspection method, time spent in hazardous zones and the number of people and hours exposed to risk each fell by about 67%, and the number of track entries fell by about 58%.
   - The key metric is reducing workers' risk exposure, not cutting inspection staff, and the final safety judgment is made by the person in charge. The drones fly set routes repeatedly with special beyond-visual-line-of-sight flight approval from the Seoul Regional Office of Aviation.
49. [High school 3 student publishes Security Camp application answers; shrinking Docker image cut vulnerability alerts from 50 to 14](https://zenn.dev/uvu/articles/758ee227a614ea) (zenn.dev)
   0 points by startupxo, 2026-11-12T18:25:09.369Z | 0 comments | https://startupxo.com/en/1379/ | https://startupxo.com/en/1379.md
   - A high school 3 student published an article with the answers used to apply to Security Camp 2026, a Japanese program for training security talent. The author says they completed the Product Security class and wrote the answers on May 11.
   - The supply chain attack question covers a case where an npm package maintainer account and token were compromised and malicious code was mixed into a legitimate package. It explains how credentials are stolen from infected environments and spread to other packages.
   - As countermeasures, the author proposes code review for workflow changes, read-only permissions by default, pinning to commit hashes, and using short-lived tokens (OIDC). The author also wrote that response procedures for a breach should be set in advance.
   - When the author converted the Dockerfile of a Go app to a multi-stage build (a method of separating the build image from the runtime image), Trivy scan alerts went from 50 to 14. The author said the remaining alerts should still be prioritized by checking factors such as whether they are used at runtime.
50. [DeepMind unveils AlphaGenome 2, predicting the effects of DNA variants by cell type to narrow down candidate disease causes](https://wowtale.net/2026/11/13/267984/?utm_source=rss&utm_medium=rss&utm_campaign=%25eb%2594%25a5%25eb%25a7%2588%25ec%259d%25b8%25eb%2593%259c-%25ec%2584%25b8%25ed%258f%25ac-%25ec%259c%25a0%25ed%2598%2595%25eb%25b3%2584%25eb%25a1%259c-dna-%25eb%25b3%2580%25ec%259d%25b4-%25ed%259a%25a8%25ea%25b3%25bc-%25ec%259d%25bd%25eb%258a%2594-%25ec%2595%258c%25ed%258c%258c%25ea%25b2%258c%25eb%2586%2588-2) (wowtale.net)
   0 points by startupxo, 2026-11-12T18:24:47.830Z | 0 comments | https://startupxo.com/en/1378/ | https://startupxo.com/en/1378.md
   - This AI model predicts how gene activity changes inside a specific cell, such as a heart muscle cell or an immune cell, when a single DNA letter changes. Earlier models stayed at the level of tissue averages.
   - According to DeepMind, it finds nearly twice as many variants that alter gene expression within cells as the previous best model. In the dataset TenK10K of 28 immune cell types, the hit rate for causal variants was 30%, higher than 16% for existing tools, the company said.
   - The model takes the cell type or state as a condition together with the DNA sequence. DeepMind explained that by uploading only single-cell RNA data from a few dozen cells, it can be applied to new cell states without retraining.
   - For non-commercial research, it can be used right away through the AlphaGenome API, and for commercial research it is offered as a private preview. The paper and the API are public.
51. [Panmnesia and Intel Korea sign MOU to cooperate on CXL technology for sharing AI server memory at the rack level](https://wowtale.net/2026/11/12/267946/?utm_source=rss&utm_medium=rss&utm_campaign=%25ed%258c%258c%25eb%2584%25a4%25ec%258b%259c%25ec%2595%2584-%25ec%259d%25b8%25ed%2585%2594%25ea%25b3%25bc-cxl-%25ea%25b8%25b0%25ec%2588%25a0-%25ed%2598%2591%25eb%25a0%25a5-ai-%25ec%259d%25b8%25ed%2594%2584%25eb%259d%25bc%25ec%2597%2590-%25ec%2596%2591%25ec%2582%25ac-%25ea%25b8%25b0%25ec%2588%25a0%25eb%25a0%25a5) (wowtale.net)
   0 points by startupxo, 2026-11-12T12:29:41.068Z | 0 comments | https://startupxo.com/en/1377/ | https://startupxo.com/en/1377.md
   - Memory is tied to each server, so even when one server has spare memory and another runs short, it is hard to share. CXL is an open interconnect standard that lets this memory be expanded and shared.
   - Panmnesia and Intel Korea signed a memorandum of understanding on the 12 at Intel AI Summit Seoul 2026 at COEX in Seoul. They will review a memory expansion architecture that combines the Intel Xeon platform with Panmnesia's link semiconductors, including its CXL switch.
   - The cooperation will start with device interoperability verification and expand step by step to rack-scale AI infrastructure. Building a reference architecture and demonstrating a proof of concept (PoC) are at the stage of discussing how to proceed.
   - Panmnesia explains that its PCIe 6.4-CXL 3.2 fusion switch is the only chip in the industry that supports CXL 3.2 PBR, and that its own controller achieves double-digit nanosecond latency. This is the company's claim.
52. [A browser 3D open-world game built from real Osaka map data, made with Claude Fable 5.2](https://qiita.com/7mpy/items/5710fef43ba0e59204c6?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items) (qiita.com)
   0 points by startupxo, 2026-11-12T12:29:23.733Z | 0 comments | https://startupxo.com/en/1376/ | https://startupxo.com/en/1376.md
   - This is a 3D game you can launch right in a web browser to walk around Osaka streets, drive cars and do missions. It is VELA BAY, made by the author 7mpy with Claude Fable 5.2, set around Tsutenkaku and Abeno Harukas.
   - The area of about 1,557m x 1,386m was built from 3,699 road segments and 4,519 buildings from OpenStreetMap, plus elevation data and aerial photos from the Geospatial Information Authority of Japan.
   - Without a physics engine, collisions are handled by treating buildings as polygons and vehicles and people as circles. Vehicle AI follows the road data and drives on the left.
   - According to the author, it was not generated all at once. Instead it was built in units such as city, vehicles, police and missions, repeating the process of checking that each worked.
53. [Open-source Japanese input method Mozkey 1.0 released, converts without the space key and handles AI correction on the device](https://koyasi777.com/mozkey/) (koyasi777.com)
   0 points by startupxo, 2026-11-12T06:32:47.820Z | 0 comments | https://startupxo.com/en/1375/ | https://startupxo.com/en/1375.md
   - Mozkey is an open-source Japanese input method made by koyasi777. It offers live conversion, where the conversion result changes automatically while you type without pressing the space key, and it can be adjusted in settings.
   - It is introduced as adding context correction from Zenz, a local AI, to the existing Mozc conversion, so that words with the same reading are chosen to fit the surrounding context. A spelling you have chosen once is reflected in the next candidates, and how much it is reflected depends on settings and the input situation.
   - The developers state that the AI correction is completed on the device, that sending usage statistics is turned off, and that no automatic update feature is included. The builds are Windows x64, Windows ARM64 and macOS Universal, and Linux, smartphones and tablets are not supported.
54. [How Ubie built Remoco in-house, an internal AI agent platform that handles work from development to legal](https://zenn.dev/ubie_dev/articles/45501469c9c044) (zenn.dev)
   0 points by startupxo, 2026-11-12T06:32:13.756Z | 0 comments | https://startupxo.com/en/1374/ | https://startupxo.com/en/1374.md
   - Remoco, built in-house by Japanese healthtech company Ubie, is an internal platform where instructions, a model, and tools are bundled and defined as an AI agent, which people call up on the web or in Slack to do work. The CTO explains that development is just one of several uses.
   - The isolated environment where agents run code holds no credentials. External requests go through a proxy, and authentication is attached only for allowed destinations. Ubie said this lets it assign tasks without approval.
   - Agents are split into personal ones that run with an individual's permissions and team ones shared in Slack. Team agents read data only through accounts for each information category, matched to the range of channel participants, which prevents management information from being exposed in shared channels.
   - The development agent Uvin handles implementation and bug fixes, and the legal agent handles first-line response to contract review 1. The CTO introduced a loop in which both agents analyze the previous day's records and produce improvement proposals.
55. [JPCERT/CC publishes an investigation report checklist for outsourcing security incident investigations](https://www.jpcert.or.jp/ir_report_checklist_20261111.html) (jpcert.or.jp)
   0 points by startupxo, 2026-11-12T00:43:30.645Z | 0 comments | https://startupxo.com/en/1373/ | https://startupxo.com/en/1373.md
   - This is a checklist that helps the party requesting a security incident investigation and the party writing the report work to the same standard. JPCERT/CC, a Japanese incident response organization, published it on GitHub.
   - It was published on November 11, 2026. JPCERT/CC explained that as ransomware (an attack that encrypts files and demands money) and data leaks have increased, it has become common to outsource investigations to external specialist firms.
   - JPCERT/CC said it created this document because organizations experiencing an incident for the first time find it hard to judge whether an external investigation is appropriate and whether the incident can be closed.
   - JPCERT/CC advised using it when outsourcing, and during the report writing and review stages.
56. [An AWS emulator that mimics AWS features on your own computer so you can test without an AWS account](https://github.com/aws/aws-emulator) (github.com)
   0 points by startupxo, 2026-11-12T00:43:07.434Z | 0 comments | https://startupxo.com/en/1372/ | https://startupxo.com/en/1372.md
   - AWS Emulator is a program that mimics AWS APIs (the request format programs use to talk to each other) on your own computer. You can test code on a development PC, a development server, or CI (an automated test environment) without calling real AWS.
   - According to the README, no AWS account is needed and there is no extra charge for the emulator itself. If you point the address that the AWS CLI or SDK uses to the emulator through an environment variable, your code runs as usual.
   - On Linux, it runs from an RPM or Debian package. On Windows and Mac, it runs as a Docker container. For Rust tests, there is also an embedded mode that you put inside your program and run separately for each test.
   - It is said that fidelity scores, showing how closely each service resembles real AWS, are published on the releases page. Features that do not work yet are listed in the issues folder of the repository.
57. [Snowflake's new dashboards let you view numbers inside the data warehouse and ask AI questions right away](https://zenn.dev/dataheroes/articles/snowflake-dashboards-2-and-future-of-bi) (zenn.dev)
   0 points by startupxo, 2026-11-11T12:28:37.249Z | 0 comments | https://startupxo.com/en/1371/ | https://startupxo.com/en/1371.md
   - Snowflake's new Dashboards (in the documentation, 2.0) display the set numbers a team checks every day on a screen inside Snowflake, and let you ask AI right away about anything you wonder about while looking. The goal is to use it without a separate BI (data visualization) tool.
   - Dashboards are written by putting SQL and chart definitions in a .dash file, so changes can be tracked with Git. For authoring, if you ask the AI tool CoCo in plain language, it creates tiles, SQL and layout for you.
   - Placing a placeholder in the SQL turns it into a filter, and you can also use drilldown, where you click a chart to dig from region to city to customer, and chart actions that filter other tiles at the same time.
   - According to the official documentation, queries run with the viewer's own permissions, so Snowflake permissions apply as they are. The author explains that the only cost is the warehouse usage fee for running queries, and that no BI license is needed.
58. [Google unveils Gemini Robotics 2.5, an AI for robots, and it picks parts into kits in a Mercedes-Benz production line test at 96%](https://wowtale.net/2026/11/11/267887/?utm_source=rss&utm_medium=rss&utm_campaign=%25ea%25b5%25ac%25ea%25b8%2580-%25ec%25a0%259c%25eb%25af%25b8%25eb%2582%2598%25ec%259d%25b4-%25eb%25a1%259c%25eb%25b3%25b4%25ed%258b%25b1%25ec%258a%25a4-2-5-%25ea%25b3%25b5%25ea%25b0%259c%25eb%25b2%25a4%25ec%25b8%25a0%25ec%2599%2580-%25ed%2598%2584%25ec%259e%25a5-%25ec%2584%25b1%25eb%258a%25a5) (wowtale.net)
   0 points by startupxo, 2026-11-11T12:28:17.887Z | 0 comments | https://startupxo.com/en/1370/ | https://startupxo.com/en/1370.md
   - Google unveiled Gemini Robotics 2.5, an AI model that helps robots handle objects and perform tasks in real environments, on the 11.
   - Google said it validated dexterous manipulation and task performance on more than 200 tasks.
   - In a real car production line test done with Apptronik and Mercedes-Benz, the humanoid robot Apollo 2 performed parts kitting (gathering parts into a bundle for a work task) with a success rate of 96%.
   - Google said it also improved safety and reliability, explaining that the task completion rate around obstacles improved by 33% and the task completion rate in environments with people nearby improved by 43%.
59. [X unveils a clip feature that lets you cut a scene from a video and post it with your own comment](https://techcrunch.com/2026/11/10/x-launches-its-own-clipping-feature-to-spur-influencer-engagement/) (techcrunch.com)
   0 points by startupxo, 2026-11-11T06:33:16.796Z | 0 comments | https://startupxo.com/en/1271/ | https://startupxo.com/en/1271.md
   - X announced on Monday a clip feature that lets users cut a scene they like from any video, add their own opinion, and post it right away without leaving X.
   - According to X, every clip carries a Clip from label that links to the original video, so viewers can see the original creator and their other content.
   - X said that both the original creator and the clip creator can earn revenue based on clip views through its rewards program, Original Content Rewards.
   - TechCrunch described the tool as closer to a simple feature that only trims and posts video.
60. [How to keep an SSH key in the Mac Secure Enclave and approve each connection with Touch ID, without a separate app](https://gatezh.com/posts/macos-secure-enclave-ssh-keys/) (gatezh.com)
   0 points by startupxo, 2026-11-11T06:32:58.850Z | 0 comments | https://startupxo.com/en/1270/ | https://startupxo.com/en/1270.md
   - Instead of a file, the secret key used to connect to servers (the SSH private key) is created inside an isolated security area of the Mac chip (the Secure Enclave). The key never leaves the chip, and the chip signs for each connection.
   - It is created with the built-in macOS commands sc_auth and ssh-keygen. The author used the -t bio option to require Touch ID every time the key is used. The post says this is supported on macOS Sequoia and Tahoe.
   - The author ran into three problems. The Touch ID enrollment was reset, so key creation failed. The old ssh-add -K instructions did not work. The key path had to be written directly in the SSH config file for the key to be recognized.
   - The author decided not to use this method because the Touch ID prompt showed an unfamiliar phrase, ctccardtoken, instead of an SSH message. Instead, the author says they will look into apps that use the Secure Enclave, such as Secretive.

Next page: https://startupxo.com/en/new.md?p=3
JSON: https://startupxo.com/api/v1/new.json?lang=en&page=2
Other languages: https://startupxo.com/ko/new.md https://startupxo.com/ja/new.md https://startupxo.com/es/new.md https://startupxo.com/ar/new.md
