読書リサーチレーダー投資フレームワーク
ログイン / 新規登録
すべて技術製品投資株式市場その他

読書

2026-10-30115 投稿

更新

Salesforce Signs Deal to Acquire Rune Technologies, a Deployed Defense Logistics AI Startup

重大
2026-10-30 04:00 GMT+8

US enterprise software giant Salesforce announced on October 29, 2026, that it has signed a definitive agreement to acquire Rune Technologies, a defense AI startup. Kendall Collins, President and CEO of Salesforce Missionforce and Government Cloud, described this as the company's "most significant national security-focused acquisition." This deal is the third in a series of October acquisitions by Salesforce, following deals for analytics platform Loops and customer engagement platform Braze.

Rune Technologies was founded in 2024 by David Tuttle and Peter Goldsborough, both former leaders in command-and-control at defense tech firm Anduril. The company's core product, TyrOS, is a super intelligence (SI) platform for tactical edge logistics designed to operate in Denied, Degraded, Intermittent, or Limited (DDIL) connectivity environments. Already deployed with the US Army and US Marine Corps, the platform connects enterprise system data with real-time operational data, enabling frontline units to forecast supply needs, optimize resources, and suggest transport routes even when disconnected, syncing automatically once connectivity returns.

Salesforce will integrate Rune into its Missionforce portfolio, aiming to complete a three-layer architecture spanning headquarters, strategic planning, and the operational edge. Missionforce currently serves as the system of record for the first two layers across the US military and allied forces. By acquiring Rune, Salesforce seeks to bridge the gap between frontline decisions and enterprise records. This move aligns with the US Department of Defense's shift since March 2025 toward prioritizing commercial software pathways, and builds on Salesforce's existing 10-year, $5.6 billion ceiling contract with the US Army.

While Salesforce did not disclose the acquisition price, the deal reflects a broader defensive strategy by traditional SaaS companies facing competition from AI-native solutions. According to Reuters, Salesforce acquired AI agent platform Fin for $3.6 billion earlier this year and Informatica for $8 billion last year. Investors have expressed concern that rapid AI adoption could erode demand for legacy software, prompting Salesforce to aggressively acquire specific AI capabilities to maintain its market position.

ソース:Salesforce

リサーチ

Korean Document AI Firm KDL and IBK Industrial Bank Complete VLM Agent PoC: Voice Phishing Relief Processing Time Cut by 88%

トピック · KDLクイックテイク
2026-10-30 12:54 GMT+8

Korean document AI startup KDL (한국딥러닝) and IBK Industrial Bank have completed a proof-of-concept (PoC) for an AI agent designed to automate voice phishing victim relief processes. According to Platum, the test conducted between April and August 2026 resulted in an 88% reduction in document processing time.

In traditional workflows, bank employees manually reviewed faxed documents such as relief applications and ID copies, then entered data into systems. This delay hindered rapid responses like payment freezes before funds could be transferred out. KDL's 'DEEP Agent for Finance,' built on a Vision-Language Model (VLM), goes beyond simple OCR by understanding document context to classify files, extract information, and infer crime types.

Data provided by KDL indicates the system achieved 100% accuracy on printed text and over 97% on handwritten text, successfully handling low-quality IDs and mixed handwriting that previous AI-OCR systems struggled with. Compared to existing AI-OCR, accuracy improved by up to 20.4 percentage points. The system uses a confidence scoring mechanism: high-confidence results are processed automatically, while low-confidence items are routed to human reviewers for final judgment.

The project won an excellence award in the PoC category of the '2026 Public Institution AI Innovation Challenge' hosted by South Korea's Ministry of Economy and Finance and Ministry of Science and ICT. IBK Industrial Bank plans to deploy the system in actual operations by the first half of 2027. KDL intends to expand this solution to other financial and public institutions.

ソース:스타트업 생태계의 모든 것 \'플래텀(Platum)\'

リサーチ

Chi Mei Hospital President Lin Hung-jung: Before AI Agents Enter Hospitals, a Secure Sandbox Environment Must Be Built

トピック · 奇美医院AI代理沙盒クイックテイク
2026-10-30 15:51 GMT+8

Lin Hung-jung divides Chi Mei Hospital's AI development into four steps: starting with data warehousing and business intelligence tools in 2005; establishing an AI Center in 2019 to extend biostatistical capabilities to machine learning; entering the generative AI assistant stage in 2023; and targeting Agentic AI starting in 2025. The logic driving each step is not technological availability, but whether the organization has already met the conditions for safely using the next layer of capability.

During the disease prediction phase, Chi Mei deliberately avoided medical imaging AI. Lin Hung-jung explained that imaging models require significant computing resources and manual annotation labor, whereas Chi Mei has accumulated years of clinical data and biostatistical experience. Focusing on disease risk prediction based on these existing capabilities is more pragmatic. This choice later yielded 32 online predictive models covering 13 departments, allowing physicians to view patient-related risks directly on the Hospital Information System (HIS) page.

Upon entering the generative AI stage, Chi Mei started with documentation tasks such as nursing records, converting vital signs and catheter logs into drafts, and then expanded to other professions like physicians and pharmacists. This collaborative mechanism operates independently of the existing HIS, avoiding the need to modify original programs and thus preventing version-change risks and cascading system crashes.

The A+ Medication Safety Specialist is currently the most notable achievement. Lin Hung-jung cited an example where a pharmacist might be responsible for the medication safety of over a hundred patients in an entire ward. Previously, rule-based systems filtered patients according to preset conditions, but medical knowledge updates with literature and clinical guidelines, causing existing rules to become outdated. The new system allows AI to first analyze medical records and medication data, screening out 10 patients out of 100 who may have medication issues and compiling their histories for pharmacists to prioritize checking. Lin stated this system saves pharmacists approximately 150 minutes per day and won the National Healthcare Quality Award (NHQA). To reduce hallucination risks, Chi Mei combines AI with the Medi-Span pharmaceutical database, using RAG to introduce continuously updated pharmaceutical knowledge. Medi-Span first screens for drug interactions, contraindications, and dosage issues, which are then processed by AI before final professional judgment by the pharmacist.

However, Lin Hung-jung explicitly stated that Chi Mei has not yet officially launched AI agents because the hospital has not established a sufficiently robust secure execution environment. He uses the 'lobster tank' metaphor: you cannot let lobsters run wild, cause damage, or escape. For a hospital, if AI agents are to read medical records, operate systems, or execute tasks, it must first ensure they only operate within permitted scopes. Building such an environment presupposes effective data governance—Lin points out that this is a technical capability lacking in many hospital IT departments.

ソース:iThome

リサーチ

Goldman Sachs Derivatives Expert Lee Coppersmith Discusses Shift in AI Stock Options Sentiment

クイックテイク
2026-10-30 12:00 GMT+8

Lee Coppersmith, from the Equity Derivative Sales team at Goldman Sachs FICC and Equities, analyzes current options trading dynamics in artificial intelligence (AI) stocks on an episode of The Markets podcast. According to the publisher's episode description, Coppersmith notes that options markets are showing signs of increased nervousness about AI as investors closely monitor large tech company earnings and attempt to separate likely winners from losers.

He explains that while overall demand for AI stocks remains strong and the S&P has remained resilient, put option buying has picked up in recent weeks, whereas activity in call options—which would reward a jump in equities—has waned. This observation offers readers a specific entry point into understanding AI investment sentiment through derivative market positioning.

The episode was recorded on October 29, 2026. This recommendation is based on publisher metadata and episode descriptions, without including the full transcript or specific argumentative details. Views expressed represent the guest's personal stance and do not constitute investment advice.

ソース:youtube.com

リサーチ

US chip giant Intel's stock plunges 22% as high memory prices suppress PC demand, missing Q4 guidance

重大
2026-10-30 15:30 GMT+8

U.S. semiconductor manufacturer Intel released its financial forecast for the current fourth quarter, indicating revenue between $15 billion and $16 billion, falling short of analysts' expectations of $17 billion. The company also projected adjusted earnings per share (EPS) of 25 cents, significantly below the market consensus of 43 cents. This guidance miss caused Intel's stock to plummet approximately 22% in after-hours U.S. trading.

Intel CFO David Zinsner stated that the company has observed a slowdown in business for weeks due to declining computer sales driven by soaring memory prices. He warned that this situation would not improve at least until the first half of 2027. CEO Lip-Bu Tan noted that the entire market is affected, with memory prices remaining a severe problem throughout next year. Intel expects the PC market to shrink by another 10% in 2027, following a 12% decline this year.

Although the AI boom has increased demand for data center memory, thereby driving up memory prices, this has simultaneously suppressed consumer PC sales. Additionally, Intel faces supply chain issues for data center chips, including shortages of certain raw materials and the temporary closure of a factory in Israel. While Intel's processors also benefit from AI demand, the company has been unable to fully meet market needs.

In the previous quarter, Intel's revenue rose 21% year-over-year to $16.5 billion, meeting analyst expectations, but net profit fell 18% year-over-year to $3.3 billion. Previously, investors were optimistic about Intel profiting from the data center business through new manufacturing processes and its emerging role as a foundry for other chipmakers. This year, Intel invested approximately $20 billion in factories and equipment, with Zinsner suggesting spending would increase further next year.

ソース:t3n.de

リサーチ

Waymo Partners with Toyota to Test Autonomous bZ4X in Japan and Expand to UK

トピック · Waymo与丰田合作重大
2026-10-30 15:01 GMT+8

Alphabet's self-driving subsidiary Waymo announced on Friday a broader partnership with Japanese automaker Toyota. Under the agreement, Toyota will manufacture the 2027 electric SUV bZ4X and integrate Waymo's self-driving system at its Takaoka factory in Japan. These autonomous vehicles will initially undergo testing in Japan, with plans to deploy them in other right-hand drive countries, including the United Kingdom.

This move signifies a concrete step in Waymo's international expansion strategy. Previously, Waymo relied heavily on the Jaguar I-Pace for its US operations and London tests. Shifting to the Toyota bZ4X indicates Waymo is leveraging the manufacturing capacity of a major traditional automaker to address supply chain scalability. Hiroki Nakajima, Executive Vice President of Toyota, stated that the collaboration aims to bring the technology to customers in Japan and beyond via Waymo's ride-hailing service, while Toyota continues to explore integrating vehicles with traffic infrastructure for improved safety.

It is worth noting that this is not the first interaction between the two companies. Eighteen months ago, Toyota and Waymo signed a preliminary agreement characterized by vague language, widely interpreted as a 'wait and see' arrangement. The current announcement specifies the model, production location, and timeline, demonstrating that the partnership has moved from intent to substantive supply chain integration. While both parties indicated they will continue exploring ways to bring Waymo's technology to personally owned Toyota vehicles, the immediate focus remains on securing a stable hardware source for Waymo's robotaxi fleet.

ソース:TechCrunch

リサーチ

Investment columnist Joachim Klement: Demographics cannot predict the outcome of US elections

重大
2026-10-30 15:01 GMT+8

The United States is approaching its midterm elections, and investment columnist Joachim Klement uses this occasion to reiterate a judgment he has long held: demographics is not destiny. His core argument is that using demographic trends to predict election outcomes has never succeeded historically.

The trigger for this judgment was an opinion piece by Thomas Edsall in The New York Times this spring. Edsall argued that the 2030 census will shift electoral votes from states with declining populations, such as New York and California, to states with growing populations, such as Texas and Florida; since the former are reliably Democratic states and the latter are reliably Republican states, this will increase the Republican advantage by approximately 12 electoral votes between 2032 and 2040.

Klement points out that this analysis implicitly assumes a key premise: voters in the 2030s will vote along the same demographic fault lines as they do today. That is, it assumes that Black voters and highly educated voters reliably vote Democratic, while older voters and working-class whites reliably vote Republican. Klement considers this a bold assumption—older people die, so no one knows if the elderly in the 2030s will still reliably vote Republican; working-class whites were historically a key constituency for the Democrats, and no one knows if they will still shun the Democrats ten years later; Hispanics currently lean more toward voting Democratic than Republican, but they are highly religious and could well switch to the Republican Party if they feel alienated by Democratic policies.

To support this criticism, Klement recommends a new paper by Richard Calvo and his collaborators. The paper analyzes the historical performance of election predictions based on demographic variables. Klement describes the results: comparing the forecast line based on demographic considerations (dark black) with the actual election result line (light gray) in the charts, there is no resemblance between the two. His conclusion is that predicting election trends using demographics is a complete waste of time.

For investors, this means that directly mapping long-term demographic trends to election outcomes and policy expectations when assessing US political risk and policy directions lacks historical empirical support. Klement acknowledges that demographic changes are indeed important, but he believes their pace is too slow, meaning many other factors will dominate election outcomes.

It should be noted that the specific methodology and statistical details of the Calvo et al. paper have not been independently verified in this article; Klement's retelling constitutes an attributed argument rather than independently verified facts. Furthermore, the mechanical process of redistributing electoral votes by population in Edsall's argument does not itself depend on voting behavior assumptions; Klement's rebuttal primarily targets the behavioral premise of coalition stability, rather than the arithmetic results of seat redistribution.

ソース:klementoninvesting.substack.com

リサーチ

Platformer Founder Casey Newton: The Less Profitable AI Search Becomes, the More It Consolidates Google's Moat

トピック · 谷歌AI搜索变现困境重大
2026-10-30 08:51 GMT+8

On October 29, 2026, Philipp Schindler, Chief Business Officer of Alphabet (Google's parent company), delivered an answer during an earnings call that unsettled investors. An analyst questioned why search query volume was increasing while revenue growth was slowing. In listing reasons, Schindler noted that increased usage of AI Mode is hurting the company because it earns less money than traditional search. Later that day, Alphabet's stock fell 12.6%, wiping out approximately $500 billion in market capitalization. Hours later, Reddit reported its earnings; search traffic had dropped significantly, with daily active users potentially declining by a low double-digit percentage in the fourth quarter. Its stock plunged 15%. Reddit CEO Steve Huffman subsequently announced plans to gradually require users to log in to use the website—a move for an open platform that has operated for over twenty years.

Casey Newton, founder of the technology media outlet Platformer, connected these two events to propose a counterintuitive judgment: the failure of AI search to monetize effectively may ultimately consolidate Google's competitive position. His core argument rests on a loop—Google's AI answers retain value on its own pages; publishers, having lost traffic, block content and sell access rights; AI products therefore become more dependent on paid licensed data; and companies that cannot afford this cost are either blocked or sued.

The premise of this loop is the collapse of the old contract between Google and the internet. In the past, publishers allowed Google to index their pages, and Google rewarded them with massive traffic. Google profited from ads embedded in search results, while publishers earned smaller profits from visitors. Reddit was one of the greatest beneficiaries of this mechanism. As the web became flooded with low-quality content, genuine human conversations became scarce, prompting Google to highlight forum discussions in search results, causing Reddit's search traffic to surge. According to Newton, reports indicate Google pays Reddit $60 million annually for data licensing.

However, after ChatGPT became a phenomenon in 2022, Google was forced to follow suit, launching AI Overviews and AI Mode. Newton points out that while Google executives claimed a year ago that Overviews monetized similarly to traditional search, Schindler's recent comments imply that AI Mode—the interface representing the future of search—is not profitable. Search revenue growth slowed from 19% in the first quarter to 17% in the second and 12% in the third, remaining nearly flat quarter-over-quarter, despite AI Mode having approximately 1 billion monthly active users and AI Overviews having about 2.5 billion.

Newton believes Google's response strategy—expanding ads in AI Mode to 10 new countries and adding ad placements to follow-up questions in the United States—is re-enacting the path taken during the mobile internet era: build the product first, then flood it with ads. The problem is that publishers will not wait passively. Reddit's login wall is a signal: when traffic no longer flows from search engines, websites lock down their data and turn to selling licenses to AI companies.

ソース:Platformer

リサーチ

AI Demand Triggers Massive PCB Shortage: Orders Booked Until 2030, Material Costs Surge 12x

トピック · AI引发的PCB短缺重大
2026-10-29 23:25 GMT+8

According to a report by Tom's Hardware citing Nikkei Asia, the AI industry is consuming manufacturing resources previously dedicated to serving PC, smart device, and gadget makers. Following RAM, GPUs, and SSDs, the Printed Circuit Board (PCB) has become the next basic electronic component facing severe shortages.

The report indicates that several leading PC and PCB industry managers have warned of shortages and price rises for PCBs and their subcomponents in the coming months. An anonymous executive from an American PC maker stated, “Even though PC shipments could decline again next year, we are still struggling to secure enough PCB supplies to meet our needs.” Several commenters attributed this crisis to AI computing demand eating up production capacity.

Upstream in the supply chain, the situation is more dire. One PCB maker noted that lead times for many products have doubled, with some orders taking as long as 30 months to fulfill. Another supplier explained that new orders placed today would not be fulfilled until late 2028. Additionally, costs for Copper Clad Laminates (CCL), a key raw material for PCBs, have surged at least 12 times since July. One of the world’s leading CCL suppliers told Nikkei Asia that its order book is currently filling up to 2030, despite plans to triple capacity by 2028.

In related developments, Nvidia recently announced partnerships with major chip substrate makers, including Japan's Ibiden, South Korea's Samsung Electro-Mechanics, and Taiwan's Unimicron. Chip substrates are essentially tiny, intricate PCBs on which CPUs and GPUs are mounted. Unimicron is looking at expanding capacity by 2.5x over the next three years.

Beyond material and capacity constraints, engineering staff shortages are exacerbating the crisis. It is increasingly difficult to recruit the engineers required at new production facilities, with salaries across the PCB industry already rising 20% to 30% to retain skilled employees.

ソース:Tom's Hardware

リサーチ

GitLab Launches Orbit Context Graph: Internal Tests Show 45x Fewer Agent Retries and 4.5x Lower Token Usage

トピック · GitLab Orbit重大
2026-10-29 08:00 GMT+8

GitLab officially launched GitLab Orbit on October 29, 2026, a "context graph" for the software lifecycle. The product aims to address the inefficiency and hallucinations common in current AI coding agents (such as Claude Code and Codex) when operating within large codebases due to a lack of holistic context. According to GitLab's official blog, Orbit integrates code, merge requests, work items, pipelines, vulnerabilities, and ownership into a single live, queryable view that updates continuously as development progresses.

In GitLab's internal testing, AI agents connected to Orbit required up to 45x fewer retries and consumed 4.5x fewer tokens to complete the same tasks using the same models. These figures suggest that providing structured project context significantly reduces the time and cost associated with agents repeatedly querying backends for missing information.

Beyond internal data, UK price comparison platform Compare the Market provided external validation as a design partner. The company tested four context approaches for its internal AI code reviewer across 79 real merge requests. Results showed that the GitLab Orbit-grounded reviewer matched approximately 70% of expected inline comments, whereas traditional Retrieval-Augmented Generation (RAG) methods achieved only about 58%, performing worse than having no added context at all. Ryan Harvey, Head of AI Engineering at Compare the Market, stated that Orbit enabled the AI code reviewer to truly understand the codebase rather than just the diff in front of it, improving comment placement accuracy and change summaries.

GitLab emphasizes that Orbit is not merely a code indexing tool but a full lifecycle graph covering everything from code changes to deployment, vulnerabilities, and owners. The graph leverages existing GitLab data without requiring connector configuration or additional ETL pipeline maintenance, while preserving original permission structures. Orbit is currently available for Premium and Ultimate users, supporting integration with mainstream AI agents like Claude Code and Cursor via MCP protocol, as well as custom integrations via REST API. Pricing is set at $30 per user per month (in GitLab Credits) for each query pack containing 1,000 queries, with overage billed at approximately $0.03 per query.

ソース:about.gitlab.com

リサーチ

Suspected Binance Hack Prompts Tether and Circle to Freeze $3 Billion in Assets

トピック · 币安遭黑客攻击事件構造的
2026-10-30 14:22 GMT+8

Binance, the world's largest cryptocurrency exchange, suffered a major security incident early on October 30. According to reporting by INSIDE, Tether and Circle, two leading stablecoin issuers, jointly froze approximately $2.54 billion in USDT and USDC assets. Binance later updated that the total frozen amount could approach $3 billion. This action aimed to prevent further movement of stolen funds but does not imply recovery.

The abnormal transfers occurred between 20:24 and 21:04 UTC on October 29, within a roughly 40-minute window. Preliminary investigations suggest a DevOps engineer's laptop may have been compromised, with no evidence yet of private key leaks. A Binance spokesperson stated that user funds remain safe and committed to covering all losses using the 'Secure Asset Fund for Users' (SAFU) and corporate reserves. Founder Changpeng Zhao also confirmed this stance on X.

Estimates of the actual stolen amount vary significantly. CoinDesk cited internal Binance estimates of approximately $6.3 billion, while Chainalysis estimated $5.9 billion. Other analysts suggest net losses after freezes might be around $1 billion, while reports referencing The Wall Street Journal indicate potential losses up to $5.5 billion. None of these figures have been officially confirmed.

The incident caused Bitcoin to briefly drop below $78,000, reflecting concerns about centralized exchange security. The rapid freezing capability demonstrated by Tether and Circle highlights their role as systemic 'circuit breakers' during crises, but also underscores the concentration risk inherent in relying on a few stablecoin issuers. Investors should monitor Binance's forthcoming official investigation report and regulatory responses.

ソース:INSIDE

リサーチ

UK AI Chip Startup Fractile Completes £577M Series C Following Reported $250M Anthropic Order

トピック · Fractile芯片融资重大
2026-10-30 14:00 GMT+8

According to filings with the UK's Companies House, London-based AI chip startup Fractile has allocated £577 million in Series C preferred shares since September. This raise significantly exceeds its $220 million Series B completed five months prior.

Founded in 2022 by Walter Goodwin and Yuhang Song, Fractile designs chips specifically for AI inference. Unlike traditional architectures that move data between separate memory and processors, Fractile uses near-memory computing to keep data adjacent to the transistors performing calculations. The company aims to increase processing speeds for long AI workloads from approximately 40 tokens per second to 1,200 tokens per second.

The funding comes amid reports of early commercial interest. Bloomberg reported that Fractile agreed to an initial deal to sell approximately $250 million worth of chips to Anthropic, with both parties interested in expanding the agreement. However, the hardware is not expected to be deployed in data centers until 2027.

Despite securing substantial capital and reported customer commitments, Fractile has not yet delivered a production chip and has no disclosed revenue. In a market driven by compute scarcity, investors appear willing to fund unproven technical approaches at high valuations. Fractile's implied valuation sits between recent peers Olix ($3.3 billion) and Etched ($21 billion), but unlike Etched, which has paying customers, Fractile faces significant execution risk in translating its architecture into working silicon.

ソース:Tech Funding News

リサーチ
次へ →
ログイン / 新規登録
読書リサーチレーダー投資フレームワーク

Another unresolved challenge is how AI agents will integrate with existing Robotic Process Automation (RPA). Chi Mei has used RPA to handle many repetitive tasks, but RPA executes predefined rules and steps. Theoretically, AI agents could understand tasks and plan steps, which RPA would then execute as fixed operations, forming agentic automation. Lin emphasizes that the specific integration method between the two remains an area of exploration.

Facing rapidly changing AI technology, Lin Hung-jung's strategy is not to chase the cutting edge, but to stay 'two steps behind the technological frontier,' maintaining caution while continuing to follow developments. This positioning means Chi Mei is willing to leave a buffer between capability verification and organizational readiness, rather than using technological availability as the sole criterion for deployment.

This is the closing point of the loop. Newton writes that regardless of how expensive content licensing becomes, Google can almost certainly afford it better than anyone else—and publishers are generally willing to sell. When Reddit closed crawler access in 2024, Google was the only major search engine to maintain real-time result access. A closed, paywalled network might increase the cost per answer for Google, but it still has the budget to pay. Most challengers do not.

Newton concludes that a shrinking network actually benefits Google because it makes competition increasingly expensive. Everyone is acting in their own long-term interests, even if it causes short-term damage. He acknowledges that AI Mode is indeed a better product for many questions, and few would miss scrolling through long lists of links to find answers. However, he warns: the internet used to be good largely because of the content funded by those links; without them, the internet may become smaller, less democratic, and more expensive. Google will be fine; he worries about everyone else.

It must be noted that this is Newton's attribution analysis, not an independently verified conclusion. Schindler did not specify exactly how much less AI Mode earns, nor did he clarify the comparison baseline. Newton's circular argument relies on a key assumption: that Google can continuously bear licensing costs and that competitors cannot find ways around paid content—such as synthetic data, new paradigms of user-generated content, or regulatory mandates for open access. If these conditions change, the loop could break. But based on evidence from October 2026, this mechanism is unfolding.