Synaptics has released the Astra SR200 series of edge AI MCUs, combining heterogeneous AI-native computing with integrated high-bandwidth custom DRAM scalable to hundreds of megabytes, delivering up to 400 GOPS of AI acceleration and 256 MB of custom DRAM, and integrating Arm Cortex-M85 and Cortex-M52 cores. The series includes three pin-compatible models—SR210 (200 GOPS), SR230 (300 GOPS), and SR250 (400 GOPS)—targeting wearables, connected cameras, and industrial vision and voice systems.
Tado has released a software update for its X-series thermostats that uses AI to learn the typical patterns of each room in order to detect open windows more reliably, especially in spring and autumn when the indoor-outdoor temperature difference is small. Room-specific learning is free for all Tado X devices, but automatically pausing heating after a window is opened remains exclusive to the "AI Assist" subscription at €4 per month or €30 per year; free users only receive a push notification. Subscribers will also get a mold index that assesses temperature and humidity changes and predicts mold risk.
Tado has released a software update for its X-series thermostats that uses AI to learn the typical patterns of each room in order to detect open windows more reliably, especially in spring and autumn when the indoor-outdoor temperature difference is small. Room-specific learning is free for all Tado X devices, but automatically pausing heating after a window is opened remains exclusive to the "AI Assist" subscription at €4 per month or €30 per year; free users only receive a push notification. S
Anthropic's draft IPO prospectus shows $16.3 billion in revenue and a net loss of $80.7 billion in the first half of 2026, of which $72 billion comes from the revaluation of financing instruments such as convertible notes and is not actual cash spending. Computing and infrastructure spending in the first half of the year was nearly $9 billion, about 55% of revenue, down from more than 200% in the same period last year, and the company had already begun to be profitable on an adjusted basis in the third quarter. The article points out that the real risk lies on the cost side: signed forward computing and infrastructure commitments exceed $600 billion and could reach as high as $750 billion; Reuters said that of the $518 billion at the time, about 80% was irrevocable or required minimum payments, while of the $16.3 billion in revenue in the first half of 2026, more than $3.2 billion came from just two major customers, and many major customers had not signed long-term purchase commitments.
Anthropic's draft IPO prospectus shows $16.3 billion in revenue and a net loss of $80.7 billion in the first half of 2026, of which $72 billion comes from the revaluation of financing instruments such as convertible notes and is not actual cash spending. Computing and infrastructure spending in the first half of the year was nearly $9 billion, about 55% of revenue, down from more than 200% in the same period last year, and the company had already begun to be profitable on an adjusted basis in th
Anthropic's Alignment Science blog published research using Opus 5 red-team agents to conduct adversarial testing against production interception monitoring such as Claude Code's Auto Mode and OpenAI Codex's Guardian, finding that on average more than 55% of attacks were not intercepted. The most effective attack methods included agent-generated prompt injection, multi-context attacks across sessions or compressed summaries, and exploiting coverage gaps where edits within the working directory are not monitored. The team introduced Auto Mode++ through improvements such as transcript-format escaping and heredoc framing, monitoring all edits, and adding a tool-call investigation agent, reducing the attack success rate by more than 50 percentage points, but reliably defending against multi-context attacks still directly conflicts with latency costs.
Anthropic's Alignment Science blog published research using Opus 5 red-team agents to conduct adversarial testing against production interception monitoring such as Claude Code's Auto Mode and OpenAI Codex's Guardian, finding that on average more than 55% of attacks were not intercepted. The most effective attack methods included agent-generated prompt injection, multi-context attacks across sessions or compressed summaries, and exploiting coverage gaps where edits within the working directory a
IBM has released Granite Timeseries Ensemble on Hugging Face, by default combining four pretrained models—PatchTST-FM-r1, PatchTST-FM-r2, FlowState, and Tiny Time Mixers—achieving an aggregate MASE of 0.683 across 97 GIFT-Eval configurations, outperforming the best single model, PatchTST-FM-r2, at 0.685. The recommended default Probabilistic Ensemble pools quantile forecasts from members with equal weights and requires no fitting; the IQR Ensemble, by contrast, weights dynamically according to forecast interval width. The main value of the ensemble is stable accuracy without having to select an architecture in advance, but PatchTST-FM-r2 still retains the lowest aggregate CRPS/MWQL.
IBM has released Granite Timeseries Ensemble on Hugging Face, by default combining four pretrained models—PatchTST-FM-r1, PatchTST-FM-r2, FlowState, and Tiny Time Mixers—achieving an aggregate MASE of 0.683 across 97 GIFT-Eval configurations, outperforming the best single model, PatchTST-FM-r2, at 0.685. The recommended default Probabilistic Ensemble pools quantile forecasts from members with equal weights and requires no fitting; the IQR Ensemble, by contrast, weights dynamically according to f
Bengaluru police arrested 25-year-old Hazarat Ali, who is suspected of posting an AI-generated fake video of an explosion at Vidhana Soudha on Instagram; no explosion actually occurred. Police registered a case under Sections 353(1)(b) and 353(2) of the Bharatiya Nyaya Sanhita, and traced his residence by analyzing digital traces such as IP logs, announcing the arrest on Wednesday, October 21. Police are investigating his motive and whether others were involved, and warned against misusing AI to create threatening content.
Bengaluru police arrested 25-year-old Hazarat Ali, who is suspected of posting an AI-generated fake video of an explosion at Vidhana Soudha on Instagram; no explosion actually occurred. Police registered a case under Sections 353(1)(b) and 353(2) of the Bharatiya Nyaya Sanhita, and traced his residence by analyzing digital traces such as IP logs, announcing the arrest on Wednesday, October 21. Police are investigating his motive and whether others were involved, and warned against misusing AI to
US Treasury Secretary Bessent said at a closed-door meeting on Wednesday that there are no systemic problems in the artificial intelligence industry, that it cannot be compared at all to the dot-com bubble at the turn of the century, and predicted that AI will greatly boost productivity and significantly promote US economic growth. But he also acknowledged that data centers and AI companies have not done enough to explain the benefits of the technology to ordinary Americans; currently, more and more regions in the United States are beginning to oppose data center construction, and investors are increasingly questioning whether related stock prices are too high.
US Treasury Secretary Bessent said at a closed-door meeting on Wednesday that there are no systemic problems in the artificial intelligence industry, that it cannot be compared at all to the dot-com bubble at the turn of the century, and predicted that AI will greatly boost productivity and significantly promote US economic growth. But he also acknowledged that data centers and AI companies have not done enough to explain the benefits of the technology to ordinary Americans; currently, more and
Microsoft released a report stating that a China-linked actor named Storm-0940 used the open-source AI agent ARTEX to automate attacks on hundreds of organizations across different industries worldwide. ARTEX is released as open source on GitHub and chains together multiple LLM agents, requiring almost no human intervention from reconnaissance to data theft; it logs in using ConnectWise's Microsoft 365 backup and management application key, bypassing user sign-in and MFA, and in several intrusions achieved cloud privileged access from the initial target in under 3 hours. The report also notes that most attack attempts failed, with one password spray recording 151K+ failures and not a single success, and that application sign-in alone does not prove data was stolen.
Microsoft released a report stating that a China-linked actor named Storm-0940 used the open-source AI agent ARTEX to automate attacks on hundreds of organizations across different industries worldwide. ARTEX is released as open source on GitHub and chains together multiple LLM agents, requiring almost no human intervention from reconnaissance to data theft; it logs in using ConnectWise's Microsoft 365 backup and management application key, bypassing user sign-in and MFA, and in several intrusio
Ming Lu, a senior advisor at KKR, believes the scale of the US market's bet on AI is unprecedented, but the payoff will come far later than the market expects. AI-related capital expenditure has risen to about 2% of US GDP, and AI-related investment accounted for 39% of US economic growth in the first three quarters of 2025, higher than the 28% at the peak of the dot-com bubble. The combined market value of the "Magnificent Seven" is $25 trillion, accounting for 35% of the S&P 500.
Continuous integration cloud service provider Blacksmith has raised a $66 million Series C round led by Sequoia Capital, with participation from Peak XV Partners, Y Combinator and GV, at a company valuation of $1 billion, bringing total funding to $125 million. The company offers CI cloud services on its own servers; more than 10,000 enterprises currently use it, and weekly code checks have grown from about 5 million in early 2026 to about 90 million, while annual recurring revenue grew 17-fold over the same period. CEO Aditya Jayaprakash said AI coding tools have made writing code easy and the bottleneck has shifted to code verification; this round's funds will mainly be used to buy servers to expand the global CI cloud and to grow the team from about 40 people to 75 by the end of the year.
Continuous integration cloud service provider Blacksmith has raised a $66 million Series C round led by Sequoia Capital, with participation from Peak XV Partners, Y Combinator and GV, at a company valuation of $1 billion, bringing total funding to $125 million. The company offers CI cloud services on its own servers; more than 10,000 enterprises currently use it, and weekly code checks have grown from about 5 million in early 2026 to about 90 million, while annual recurring revenue grew 17-fold
Grok Bot can now generate reports, memos, and dashboards. After users give the Bot a page request, they can edit it together with the Bot, or have the Bot keep it continuously updated. According to Roshan Sadanani, content such as product dashboards, personal financial metrics, PRDs, UXR summaries, research reports, industry analyses, metric explorations, and reference documents can all be generated directly within Grok Bot.
Peter Yang has updated his review of four personal agents: Grok Bot is free and integrates with X, with excellent voice mode and Pages features, but its email function is basically unusable; Muse is free and offers the best consumer-grade experience, with standout performance on business calls; Claude Duet is the most proactive out of the box and has the strongest model capabilities, but the research preview lacks features such as driving a browser; ChatGPT Dots can drive cloud and local threads within ChatGPT, but still has bugs and slow voice connections.
Peter Yang has updated his review of four personal agents: Grok Bot is free and integrates with X, with excellent voice mode and Pages features, but its email function is basically unusable; Muse is free and offers the best consumer-grade experience, with standout performance on business calls; Claude Duet is the most proactive out of the box and has the strongest model capabilities, but the research preview lacks features such as driving a browser; ChatGPT Dots can drive cloud and local threads
Anthropic has released a personal agent app called Claude Duet, positioned as a "proactive agent" that can run in the background and proactively push tasks to users; it is currently in research preview with a waitlist. The tool is built on the flagship Opus 5.5 model released at the end of September, and it pushes a daily morning briefing from connected apps, notifies users when something needs attention, and requires user approval for some tasks. Mike Krieger, co-lead of the Anthropic Labs team, says it can judge when to act proactively; Karan Ahuja, who leads Duet development, previously headed product management at Meta Superintelligence Labs.
Anthropic has released a personal agent app called Claude Duet, positioned as a "proactive agent" that can run in the background and proactively push tasks to users; it is currently in research preview with a waitlist. The tool is built on the flagship Opus 5.5 model released at the end of September, and it pushes a daily morning briefing from connected apps, notifies users when something needs attention, and requires user approval for some tasks. Mike Krieger, co-lead of the Anthropic Labs team
CICC recommends "buying insurance" for AI assets through asset allocation: in addition to defensive assets such as cash, bonds, and high-dividend stocks, investors should increase their allocation to gold to position for the dollar depreciation trade. Since 2026, the negative correlation between gold and the dollar has risen to -81.2%; across 91 dollar-decline periods since 2000, gold rose 84 times, a win rate of 92.3%. CICC believes that cooling in the AI trade could weaken demand for the dollar, and the dollar depreciation trade can provide portfolio protection when the AI rally weakens or liquidity improves.
CICC recommends "buying insurance" for AI assets through asset allocation: in addition to defensive assets such as cash, bonds, and high-dividend stocks, investors should increase their allocation to gold to position for the dollar depreciation trade. Since 2026, the negative correlation between gold and the dollar has risen to -81.2%; across 91 dollar-decline periods since 2000, gold rose 84 times, a win rate of 92.3%. CICC believes that cooling in the AI trade could weaken demand for the dolla
Alloop has launched Trio, a modular AI-native health wearable consisting of a wristband, a ring, a Cube sensor and a Halo, which can be combined into either a wristband or ring form. It measures resting heart rate, heart rate variability, blood oxygen, respiratory rate and body temperature, and provides a sleep score and Bio Age. Its AI health agent calls Google Gemini and ChatGPT; the app is subscription-free, with an optional Pro plan at about $5/month. Trio will launch on Kickstarter on November 10 at $299, with shipping expected between late December 2026 and early January 2027.
Alloop has launched Trio, a modular AI-native health wearable consisting of a wristband, a ring, a Cube sensor and a Halo, which can be combined into either a wristband or ring form. It measures resting heart rate, heart rate variability, blood oxygen, respiratory rate and body temperature, and provides a sleep score and Bio Age. Its AI health agent calls Google Gemini and ChatGPT; the app is subscription-free, with an optional Pro plan at about $5/month. Trio will launch on Kickstarter on Novem
LMSYS used Miles, SGLang, Megatron-LM, and Harbor to perform fully asynchronous agent reinforcement learning on Qwen3-30B-A3B-Instruct on AMD Instinct MI355X, raising pass@1 from 7.3% to 27.5% on a 100-task stratified subset of SWE-bench Verified. Training was conducted in two stages: first SFT on approximately 4,000 public trajectories from Open-SWE-Traces, then GRPO on approximately 1,350 SWE-smith tasks. After fixing the model and RL recipe, scaling inference from 1 node to 3 8-GPU nodes, switching to E2B cloud sandboxes, and removing the concurrency limit increased rollout throughput from approximately 860 to approximately 3,480 trajectories/hour; limiting trajectory length on the same 8-GPU node reduced single-step training time from 850 seconds to 284 seconds.
LMSYS used Miles, SGLang, Megatron-LM, and Harbor to perform fully asynchronous agent reinforcement learning on Qwen3-30B-A3B-Instruct on AMD Instinct MI355X, raising pass@1 from 7.3% to 27.5% on a 100-task stratified subset of SWE-bench Verified. Training was conducted in two stages: first SFT on approximately 4,000 public trajectories from Open-SWE-Traces, then GRPO on approximately 1,350 SWE-smith tasks. After fixing the model and RL recipe, scaling inference from 1 node to 3 8-GPU nodes, swi
A heise article argues that as multi-agent systems and generative AI make machines into data consumers, data architectures such as Data Mesh—originally designed for human analysts—are facing challenges and require a federated Knowledge Layer to connect local context and business semantics with a global enterprise model. The article was written by Saskia Untiet-Kepp of Sopra Steria, and the full content is available only to heise-Plus subscribers.
Michele Wucker, author of "The Gray Rhino," warned at the 2026 Sustainable Global Leaders Conference that AI consumes far more energy than the human brain, and cited MIT Media Lab research saying that after people use ChatGPT they become more forgetful and their brains tend to weaken. She also noted that U.S. data centers already account for about 5% of the power grid's electricity consumption, and cited a survey saying that of more than 6,000 executives interviewed, only 7% thought AI was helpful, while about 70% thought it was not.
Michele Wucker, author of "The Gray Rhino," warned at the 2026 Sustainable Global Leaders Conference that AI consumes far more energy than the human brain, and cited MIT Media Lab research saying that after people use ChatGPT they become more forgetful and their brains tend to weaken. She also noted that U.S. data centers already account for about 5% of the power grid's electricity consumption, and cited a survey saying that of more than 6,000 executives interviewed, only 7% thought AI was helpf
Nvidia has unveiled the DSX Pod AI Factory, which packs 7 Vera Rubin AI racks into a container roughly the size of a standard shipping container and weighing about 28,000 kilograms. Cabling and liquid cooling are assembled at the factory, and the unit can be transported by truck over ordinary roads. Nvidia says the solution can cut the time for a new AI data center to start outputting tokens from 6 months to 14 days. The product is a reference design, shipped by manufacturers including Foxconn, Gigabyte, Lenovo, HPE, Quanta, and Pegatron, with deliveries expected to begin in early 2027; no price has been announced.
Nvidia has unveiled the DSX Pod AI Factory, which packs 7 Vera Rubin AI racks into a container roughly the size of a standard shipping container and weighing about 28,000 kilograms. Cabling and liquid cooling are assembled at the factory, and the unit can be transported by truck over ordinary roads. Nvidia says the solution can cut the time for a new AI data center to start outputting tokens from 6 months to 14 days. The product is a reference design, shipped by manufacturers including Foxconn,
Arda, founded by former OpenAI research lead Bob McGrew, has raised $230 million at a $1.5 billion valuation, led by Altimeter Capital, Accel, and Khosla Ventures. Arda develops AI software for automated manufacturing, and its models have already been deployed at defense technology company Anduril. McGrew describes the product as an agent for manufacturing engineers that can significantly speed up the setup of assembly lines; the next item on its roadmap is robotics.
