SCROLL NEWS / DISCOVERY
Search headlines
Find the stories shaping the conversation.
Results for " GOOGLE " 3,759 found 🔀 AI-powered Shuffle
Asus Googlebook shows off a big trackpad and plenty of ports in leaked images [Gallery]
New images of a Googlebook laptop from Asus have surfaced, showing off of selection of ports and a big ol’...
Samsung’s battery upgrades for Galaxy S27 Ultra and Pro leak
A new leak reveals the battery sizes of Samsung’s upcoming Galaxy S27 Ultra and S27 Pro including some long overdue...
Why AI is speeding up scientific research but not lab experiments
A new report by Google, Google DeepMind and MIT finds that researchers are using AI…a lot. It’s just not helping as much in the lab
Google Adds Local Businesses To EEA Search Result Units
Google updated its Search documentation to add local business queries to the two result units rolled out in the EEA for directories and direct providers.
Google's new 'CC' is an AI agent that helps families run their households
Google is refocusing its CC AI agent on household coordination, letting families share emails, schedules, and tasks so the AI can manage calendars, fill out forms, make shopping lists, plan meals, and more.
Google’s Pixel VIPs redesign starts rolling out to users
Pixel VIPs is getting a visual refresh with swipe navigation, redesigned profile pages, and the removal of activity recommendations.
Engrams Embedding Entendre: Codesign for Efficient DRAM/SSD Offloading
Engram extends standard token embeddings with learned multi-token lookups. Recurring local patterns retrieve vectors directly, reducing the need to reconstruct them through attention and feed-forward layers. With Engram model architecture optimization, it allows for lower HBM capacity to be needed for models at the same quality. This does not mean there won’t be an insane demand for HBM but it just means that model architecture will continue to innovate around constraints. This model architecture design is naturally codesigned for parameter offloading: each token accesses a few embedding rows whose addresses depend on token IDs, not hidden states. The runtime can prefetch those rows from host DRAM while earlier layers compute, keeping the table outside HBM without transferring entire weight matrices. Our Memory model contains our latest estimates of quarter by quarter HBM, DRAM, & NAND supply and demand. Offloading frees HBM for model weights and KV cache, potentially supporting larger batches or more concurrent sessions. When DRAM becomes the next constraint, NVMe offers another tier. Recommendation systems already cache frequently or recently accessed embedding rows in faster memory while backing colder rows with SSDs. After NVIDIA roadmap had to change due to massively despec’ing Rubin Ultra from 1024GB to now ~200GB of HBM per chip, model architecture optimizations like emgram maybe helpful. Our DeepSeek-V4.1-Flash configuration uses roughly 189 GiB of memory for Engram. We replace it with a memory-mapped (mmap) file and measure serving performance with offload to SSD. Later on in our report, we will show our Engram offloading experiments along with the official InferenceX agentic inference serving results on Engram models like DeepSeekv4.1 Flash across all 6 NVIDIA GPU SKUs along with MI355X. Unsurprisingly, the CUDA Moat is still mogging MI355X on the ultra popular DeepSeekV4.1 Flash model. We also show how even on high capacity HBM SKUs, offloading emgrams to DRAM could result in even better performance for most of the pareto than keeping the emgram in HBM. Our benchmark has been widely reproduced, validated and/or supported by almost every major buyer of compute from Google Cloud to Microsoft Azure to Oracle, to Meta and many more. Furthermore, it has the support of the ML community including from vLLM, LMCache, SGLang, PyTorch, Huggingface and the support of major labs like OpenAI, MiniMax, ZAI, Qwen, Moonshot Kimi, etc. Star the InferenceX GitHub repository if you find the open-source benchmark and data useful!. InferenceX is the only inference benchmark in the world to have TPUv7, Jalapeño, Nvidia Rubin NVL72, AMD, and soon, SambaNova and Trainium. Due to how realistic AgentX scenario is to real world agentic inference workloads, AMD has committed to collaborating on MI455X UALoE72 too. DeepSeek did not release the original paper’s two trained Engram models. We replicated its setup on fineweb-edu using the released code and training hyperparameters, at an estimated 6E18 FLOPs per run. We observed the same U-shape scaling: Engram improved performance over pure MoE baselines. We also reproduced DeepSeek’s results, where earlier-layer representations with Engram resembled those of later layers. Like the original Engram paper, it is possible to probe Engram’s gate scores to see what n-grams DeepSeek-V4.1-Flash makes the most use of. Our gate scan finds names, code fragments, relational phrasing, and boilerplate. These examples prioritize interesting-ness over gate strength. One unexpected result was Wright : Ace Attorney. These examples suggest learned memory optimizes the training objective, not a judgment of which facts deserve storage. Licenses, bibliography fragments, API scaffolding, and website furniture can provide prediction shortcuts, so the value of additional Engram capacity may depend on what survives data preparation. This does not show that table capacity is “wasted”: the evaluation-corpus scan establishes neither training exposure nor the capacity occupied by each category. For offloading, strong gates do not identify cache-hot rows. Low gates do not automatically save reads either: computing the gate requires the retrieved key and undoes the performance gain of a fused kernel. Skipping reads would require a separate usefulness predictor before retrieval. In the original paper’s inference-time ablation, factual-knowledge benchmarks retained just 29–44% of their original performance, while reading comprehension retained 81–93%. This is due to the training–inference mismatch. The resulting degradation therefore measures this trained model’s dependence on Engram, not the performance difference between models trained with and without it. In our ablations, suppressing Engram worsens token likelihood across all evaluated domains, especially encyclopedia text and several code corpora. Surprisingly, GSM8K accuracy stays within measured run-to-run variation and removing Engram has no effect. Engram is not a detachable dictionary beside an unchanged MoE. Removing it changes downstream features and expert selection. We tested whether rerouting hurts or compensates by holding tokens fixed in a teacher-forced experiment on CRUXEval, a code-reasoning benchmark of small Python functions where the model predicts a function’s output from its code and an input, and scoring the reference answer. Removing Engram raised answer loss from 0.2848 to 0.3093 bits/token. Forcing the ablated model to use the original Engram-on expert choices made it worse still, at 0.3375 bits/token. Rerouting partially compensates for the missing memory. Memory features and expert selection work together, rather than following a clean “memory stores facts; experts reason” division. On the same CRUXeval, removing Engram during either phase reduced accuracy and increased generated tokens; removing it throughout produced the largest changes. Keeping Engram for prefill leads to more correct answers than keeping for decode likely due to semantically richer KV cache transferred to decode workers, allowing it to mitigate some of the performance loss. Engram’s table is large, but each lookup is small. DeepSeek-V4.1-Flash requests 24 rows at each of two Engram layers, about 12.4 KiB per processed token position across the model, or 3.1 KiB per GPU when split across four GPUs. Currently as of Day 7 since Model Release, MI355X is still 2-4x worse performance per dollar compared to B200 even when normalized by Mi355X’s lower TCO. Our full total cost of ownership breakdown comes from our AI Cloud TCO Model along with monthly market surveys of over 100+ gpu clouds & gpu cloud customers. On the Day 0 release of DeepSeekv4.1 Flash, NVIDIA vLLM works out of the box with zero issues across all 6 SKUs: H100, H200, B200, B300, GB200, GB300! This was thanks to the amazing work by the NVIDIA & Interact teams! In comparison, AMD vLLM did not work on day 0 for DeepSeekv4.1 Flash.Source: SemiAnalysis InferenceX AMD’s vLLM documentation points to using vllm/vllm-openai-rocm:deepseekv41-flash-0909, but from hour 0 of the model release to hour 23, AMD has not publicly released the image. AMD claims, “SPEED IS THE MOAT,” yet it has still not released it by the 23rd hour. We wish that, going forward, the AMD team has a better process for hour 0 model releases. Eventually when they did publicly release for “day 0” image support, performance-wise, it is currently up to 14.8x worse perf per dollar than H200 and up to 42x worse perf per dollar than B200/B300. The power of the CUDA MOAT is NVIDIA’s collaboration with its massive 6 million-developer community ecosystem including most of the vLLM & SGLang & Tokenspeed maintainers which means that CUDA is optimized on day 0. Overall, AMD did make significant improvements but the performance per dollar is still currently 2-4x worse than B200. AgentX Engram DRAM Offloading Improving Performance The HBM and DRAM offload use the same GPU kernel to select and dequantize rows. With HBM, it reads GPU memory; with Unified Virtual Addressing (UVA), it reads pinned host memory directly. Both support full decode graphs. Moving the table into HBM accelerates only the sparse lookup, leaving decoder computation and communication unchanged, resulting in little overall benefit while consuming memory otherwise available to KV cache. Another benefit of Engram offloaded to DRAM which means you can reduce the communication overhead by using less HBM GPUs per replica. For example, when enabling Engram offloading on B300, we are able to switch from TP4 to now TP2 which improves the pareto curve by up to 1.6x. When iso-model quality, less HBM is required as the engrams could be offloaded to host DRAM. Thus HBM bandwidth matters way more than HBM capacity. For inference workloads where memory bandwidth matters the most, 4-hi HBM provides the best $/bandwidth and therefore lowest cost per token. If China continues to make more and more revolutionary model architecture innovations, soon it could potentially 0Hi HBM stacks. [ ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) [ Long Live the Short King: Why 4-hi HBM Wins ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) Myron Xie, Bryan Shan, and 3 others · Sep 13 [ Read full story ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) Moreover, on B300 and week-0 stack, moving the Engram table back to HBM did not improve results and stay within run-to-run variance. This is a result of the work optimizing DRAM offload, such as async, and overlap. SSD Offloading On B200, we created a unoptimized vLLM fork and stored the Engram tables in memory-mapped files on local SSDs. File backing lets the OS reclaim table pages when other applications need RAM. Pages already cached in memory can be served without reading the SSD again. Furthermore, note that we were unable to turn on GDS The unoptimized SSD implementation changes how rows reach the GPU. It copies row IDs to the CPU, deduplicates them, gathers the requested rows into pinned buffers, copies those rows back to the GPU and dequantizes them. This work runs between segments of the GPU execution graph. Native UVA performs row selection and dequantization directly on the GPU, avoiding the CPU round trip. A warm filesystem cache removes physical SSD reads, but leaves the coordination, row gathering and transfers. This is why a file already cached in RAM can still perform worse than a pinned DRAM table. The comparison measures the whole serving path; it does not separate the time spent on each of these operations. B200 DRAM dominates both measured SSD serving curves in total tokens per dollar and P90 interactivity. Near 125 tokens/s/user, DRAM delivers 121 million total tokens per dollar versus 52 million for SSD. For production serving, SSD offloading is likely not worth the tradeoff. On the B200 configurations we measured, SSD offloading loses on both measures: every observed SSD point has a DRAM alternative that delivers higher P90 interactivity and more total tokens per dollar. The only points where Cheaper storage does not automatically produce a cheaper inference service. Moving Engram to SSD leaves the same four expensive GPUs and the rest of the server in place. Reclaiming RAM only creates an economic benefit if it enables a cheaper server configuration or additional useful capacity. The current unoptimized path provides neither benefit, and the filesystem cache still consumes RAM when table pages are resident. Next we will look at the mechanisms and specific implementation of ngrams from DeepSeekv4.1 Flash, LongCat, Qwen3.8 Flash Next.
The Pixel Weather widget now lets you catch up on the forecast with a single tap
Google's latest Pixel Weather update brings a new refresh button lets the home screen widget grab a fresh forecast in a single tap.
Google Moving Into Former Science Lab Space in Watertown
A life science lab project permitted to be built at the Arsenal on the Charles complex has a tenant listed as “Google.” It is not clear what the tech giant plans to do with the space, according to the report by Watertown’s Sam Ghilardi. The space is located at 200 Talcott Ave. formerly housed Werewolf […]
WHO flags growing health workforce gap as workers and patients age
GENEVA, Sept 18 (Reuters) - The World Health Organization warned on Friday of a gap in the medical workforce around the world in the 2030s, with a quarter of doctors set to retire in the next decade. "In many countries both our populations and health...
Fulton Fall Festival to offer variety of attractions, entertainment
FULTON — The Fulton Fall Festival, presented by the Richard S. Shineman Foundation, will bring a full day of family fun Sept. 26.The free event will be held from 10 a.m. until 4 p.m. throughout the Fulton War Memorial complex. × This page requires Javascript. Javascript is required for you to be able to read premium content. Please enable it in your browser settings. kAmxE H:== :?4=F56 3@F?46 9@FD6D AC@G:565 3J t!xr tG6?ED !2CEJ #6?E2=[ EC242? u@F?52E:@?[ 7246 A2:?E:?8 3J uF=E@? w:89 $49@@= pCED[ 32==@@? 2?:>2=D 3J $E2CCJ $Ajv p?:>2=D[ 2?5 |28:4 |2EE 3J E96 ~DH68@ r@F?EJ *@FE9 qFC62F]k^Am#placement_726068_0_i{width:100%;margin:0 auto;}var rnd = window.rnd || Math.floor(Math.random()*10e6);var pid726068 = window.pid726068 || rnd;var plc726068 = window.plc726068 || 0;var abkw = window.abkw || '';var absrc = 'https://ads.empowerlocal.co/adserve/;ID=181918;size=0x0;setID=726068;type=js;sw='+screen.width+';sh='+screen.height+';spr='+window.devicePixelRatio+';kw='+abkw+';pid='+pid726068+';place='+(plc726068++)+';rnd='+rnd+';click=CLICK_MACRO_PLACEHOLDER';var _absrc = absrc.split("type=js"); absrc = _absrc[0] + 'type=js;referrer=' + encodeURIComponent(document.location.href) + _absrc[1];document.write('');kAmp=D@[ |C] $@?87=@H6C[ A6C7@C>2?46D 3J q6==6 #@J2=6 s2?46 p4256>J 2?5 ~77 qC@25H2J s2?46 r6?E6C[ 2 56>@?DEC2E:@? 3J |2DE6C !CJ@C’D %266C:42[ J2C5 82>6D[ E6>A@C2CJ E2EE@@D[ 2?5 >@C6]k^AmkAm%96C6 H:== 36 5@K6?D @7 G6?5@CD[ =@ED @7 7@@5[ 2?5 =:G6 >FD:4 2?5 6?E6CE2:?>6?E E9C@F89@FE E96 52J]k^Am googletag.cmd.push(function() { googletag.display('yup-rectangle-one'); });
That security patch date on your Android phone is no longer the full story
Google has released new tools allowing Android apps to get a much more accurate picture of how secure your phone really is.
One of the Pixel 11’s exclusive features is already showing up on Pixel 10
Google is testing the Pixel 11’s Proactive Assistance feature on the Pixel 10 with the latest Android 17 QPR2 Beta 5.
Nestle weighing all options after Kremlin moves to seize Russia business
By John Revill ZURICH, Sept 18 (Reuters) - Nestle said on Friday it was weighing all options to protect its rights after the Kremlin's decision to take control of the Swiss food maker's Russian business, the latest move against a Western company since ...
LinkDaddy LLC Expands AI Verified Business Passport for AI Visibility
LinkDaddy LLC is expanding AI Verified, a business identity platform that brings public business information and supporting sources into one pro...
Billy Joel's rep denies death speculation after reported Google error
A week after Billy Joel revealed he had undergone brain surgery, his publicist addressed an error that reportedly led Google users to believe the singer had died.
Google's Revamped CC Is An AI Agent For Families And Groups
CC gives an AI agent its own Google account in your family or group and sends all members a daily briefing.
Disney names insider Adam Smith as chairman of streaming business
Sept 17 (Reuters) - Walt Disney on Thursday named insider Adam Smith chairman of its streaming business, as the company seeks to sharpen its focus across its direct-to-consumer platforms such as Disney+ and Hulu. • The company also named Joe Earley...
Disney names insider Adam Smith as chairman of streaming business
Sept 17 (Reuters) - Walt Disney on Thursday named insider Adam Smith chairman of its streaming business, as the company seeks to sharpen its focus across its direct-to-consumer platforms such as Disney+ and Hulu. • The company also named Joe Earley...
Google announces new experimental "CC" AI agent for families
Multiple family members can share data to help the agent make plans and complete tasks.
The king and AI: UK monarch Charles meets artificial intelligence leaders as safety concerns swirl
LONDON (AP) — King Charles III on Thursday urged senior leaders from OpenAI, Anthropic, Google DeepMind and Nvidia to make sure that artificial intelligence remains under control and in the service of people and the planet. The gathering in Scotland was convened to discuss how AI can benefit society, and comes at a pivotal moment […]
Android Bench 2.0 focuses on long-horizon tasks, agent evaluations
Google’s development of Android Bench continues today with a version 2.0 that reflects how AI can handle more complex development tasks.
US Health Secretary Kennedy adds eight new members to key preventive health panel
Sept 17 (Reuters) - US Health Secretary Robert F. Kennedy Jr. on Thursday named eight new members to the U.S. Preventive Services Task Force, the committee that decides which preventive medical care should be provided to patients at no cost. The Prev...
Vitality and Google Expand Technology Partnership to Bring AI-Driven Health Platform to the United States to Incentivize Healthier Behaviors
Vitality AI, built on Google Cloud and using Gemini Enterprise, expands into the U.S. market — bringing proven personalized health recommendations, coaching, and incentives to health plans and...
Google Home just became a power user’s paradise
Google Home MCP opens the smart home to every AI agent; Android 17 QPR1 and One UI 9 land the same week; Trash: the 39 toll booth for your light bulbs.
Across Minnesota, opposition to data centers is a political campaign season rallying cry
It's campaign season, and one issue is taking center stage: data centers. In Pine Island, a proposed hyperscale data center prompted a slate of first-time candidates to run for local office.
Morocco’s political parties battle youth disenchantment ahead of election
By Ahmed Eljechtimi and Alexander Dziadosz RABAT/CAIRO, Sept 17 (Reuters) - Morocco’s political parties are trying to entice young voters to the polls next week with promises of jobs and affordable housing after violent protests last year and a border ...
BlitzReels Announces AI Clipping Workflow That Turns Business Expertise Into Polished, Editable Short Videos
BlitzReels is announcing an AI clipping workflow that transforms podcasts, webinars and B2B recordings into polished short videos while giving b...
Here’s everything new in Android 17 QPR2 Beta 5 [Gallery]
Android 17 QPR2 Beta 5 is rolling out today to continue the development cycle ahead of the stable December launch.
Meta is navigating its 'Google Glass' moment
Google Glass taught Google that camera glasses make people uncomfortable. Now, Meta is reportedly developing a camera-free version of its AI glasses.
Google doubles down on today's preview builds with Android 17 QPR2 Beta 5
Android 17 QPR1 just had its stable release earlier this week, but development continues on QPR2, with Beta 5 now landing.
Google rolling out Android 17 QPR2 Beta 5 for Pixel
The two-week update cycle continues with Google rolling out Android 17 QPR2 Beta 5 for Pixel devices today...
Google Home app brings battery-efficient widget, reduces gray device tiles
Google today is also rolling out Home app (especially for the homescreen widget) and Gemini for Home voice assistant updates.
Google confirms a dangerous Pixel security flaw was exploited by hackers
Pixel owners were hit by zero-day attacks
Goldman CEO says fixed income, currencies, commodities business slightly softer in Q3
NEW YORK, Sept 16 (Reuters) - Goldman Sachs expects its fixed-income, currencies and commodities (FICC) business will be slightly softer in the third quarter compared to very strong performance for its equities business, Chief Executive David Solomon ...
The September Pixel Drop improves the VIP contacts widget, adds Harry Potter themes
Audible just released Harry Potter: The Full-Cast Audio Editions audiobooks and Google is offering a free two-month trial to eligible users.
Google Home MCP lets Antigravity, Claude, OpenClaw, & more control your smart home
Google today announced “Home MCP” to let your personal AI agent interact with your smart home devices via the Model Context Protocol.
Google Pixel phones pwned in zero-click attacks
CISA gives federal agencies just 3 days to patch
Your AI agents can now control your Google Home devices
Google is launching early access to a new MCP server for Google Home, allowing AI agents like Claude, ChatGPT and others to control connected devices, review camera summaries and access smart home activity using natural language.
Google Home Is Going Agentic Via Integration With The MCP Standard
Premium Advanced users will get an early access build "in the coming weeks."
Claude comes for Gemini with its own take on Docs and Slides
Claude’s getting its own version of Google’s productivity tools.
Google confirms Pixel phones were exploited in ‘targeted’ attack related to modem
Google has confirmed that a “limited” number of Pixel phones were attacked by a vulnerability related to the modem on...
US business inventories rise more than expected in July
WASHINGTON, Sept 16 (Reuters) - U.S. business inventories increased more than expected in July as retailers and wholesalers restocked to meet robust demand, suggesting that inventories could add to economic growth in the third quarter. Inventories ju...
Google says some Pixel phone owners were hacked in zero-day attacks
The Pixel phone maker said there are indications that a bug in the phone's modem "may be under limited, targeted exploitation."
Samsung starts rolling out Android 17 with One UI 9, confirms initial list of Galaxy devices
After months in beta, Samsung is now rolling out One UI 9, based on Android 17, to select Galaxy devices....
Fresh Googlebook leak offers an early glimpse at Acer, ASUS, and Lenovo models
Leaked renders reveal upcoming Googlebooks from Acer, ASUS, and Lenovo featuring slim builds, a bunch of ports, and a retro Glowbar accent.
R/GA hires former Google exec Sadie Thoma as global chief business officer
In a newly created role, Thoma returns to the agency side after 14 years at Google to steer R/GA’s AI commercialization and global growth.
N0va Phishkit Targets US and EU Businesses: A New Challenge for Identity Security
N0va phishing abuses legitimate authentication flows to capture access and refresh tokens and establish SSO access to corporate resources.
Google Patches Pixel Modem Flaw Amid Signs of Limited Targeted Exploitation
Google patches CVE-2026-58704, a Pixel modem privilege escalation flaw that may be under limited, targeted exploitation.
Public invited to meet author at Oswego library
OSWEGO — Just in time for spooky season, author Savannah Stephens will visit the Oswego Public Library Oct. 3 from 1-2 p.m. to talk about her book, “Witch Queen Rising.”Step into gothic kingdoms, cursed forests, and haunted palaces where every bargain with magic demands blood, loyalty, or your soul. × This page requires Javascript. Javascript is required for you to be able to read premium content. Please enable it in your browser settings. kAm$E6A96?D :D 2 52C< 72?E2DJ 2FE9@C H9@ AC@F5=J 4C62E6D E96 C6AC6D6?E2E:@? D96 =24A=6I AC@E28@?:DED 2C6 H2CC:@CD[ BF66?D[ 2DD2DD:?D[ H:E496D[ 2?5 >@?DE6CD H9@ 2== 6>3@5J q=24< v:C= |28:4] w6C 8@2= :D ?@E @?=J E@ 8:G6 C6256CD 72?E2DE:4 H@C=5D E@ 6D42A6 :?E@[ 3FE =6DD@?D 2?5 E@@=D E@ DFCG:G6 E96 C62= @?6 H96? E96 A286D DE@A EFC?:?8]k^AmkAmw6C $@FE96C? v@E9:4 72?E2DJ 563FE[ “(:E49 "F66? #:D:?8[” H2D 2? p>2K@? t5:E@CD’ !:4< 7@C q6DE $4:\u: U2>Aj u2?E2DJ WpAC:= a_aeX]k^Am#placement_726068_0_i{width:100%;margin:0 auto;}var rnd = window.rnd || Math.floor(Math.random()*10e6);var pid726068 = window.pid726068 || rnd;var plc726068 = window.plc726068 || 0;var abkw = window.abkw || '';var absrc = 'https://ads.empowerlocal.co/adserve/;ID=181918;size=0x0;setID=726068;type=js;sw='+screen.width+';sh='+screen.height+';spr='+window.devicePixelRatio+';kw='+abkw+';pid='+pid726068+';place='+(plc726068++)+';rnd='+rnd+';click=CLICK_MACRO_PLACEHOLDER';var _absrc = absrc.split("type=js"); absrc = _absrc[0] + 'type=js;referrer=' + encodeURIComponent(document.location.href) + _absrc[1];document.write('');kAmq@@>F?:EJ C@@> @? E96 =@H6C =6G6= @7 E96 =:3C2CJ[ =@42E65 2E `a_ t] $64@?5 $E][ ~DH68@] %9:D AC@8C2> :D 7C66 2?5 @A6? E@ E96 AF3=:4]k^AmkAmu@C 2?J BF6DE:@?D[ 42== b`d\bc`\dgef @C 6>2:= k2 9C67lQ>2:=E@i@DH=:3o?4=D]@C8Qm@DH=:3o?4=D]@C8k^2m]k^Am googletag.cmd.push(function() { googletag.display('ad-2978731'); });
Deep in Trump country, a revolt against corporate money could reshape political spending
By Brad Brooks COLUMBIA FALLS, Montana, Sept 16 (Reuters) - More than 15 years after the U.S. Supreme Court reshaped American politics with its Citizens United campaign finance ruling, Montana voters will decide this November whether to make their stat...
Google’s Thompson Center debut delayed to 2028
The tech giant dropped the news in a blog post revealing new details of public spaces planned for the revamped former state office building.
Watercolor artist shares how she does and doesn't use AI in her work
Artist Sara Wilczynska, who previously was a software engineer at Google, reflects on the strict boundaries she has for AI usage in her work.
The new Boox e-ink devices get Android 16 out of the box, including Palma 3
Boox just announced three new e-ink devices on Android 16; the largest tablet is available now, but you’ll have to...
Google rolling out September 2026 update for Pixel Watch 2, 3, & 4
As part of the September 2026 Feature Drop, Google today is rolling out a Wear OS update to the Pixel Watch 2, 3, and 4.
Gemini 3.8 Live Extended Thinking powers Gemini Live, Gmail, & Keep
Google today announced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking as its “most advanced live dialogue models yet.”
Google announces $10 million investment in energy affordability in Nevada
Google announced an eight-figure investment to “support energy innovation and affordability in Nevada,” according to a press release.
Mark your calendars: A Googlebook 'celebration' is happening next month
Google is reportedly sending out invites for an event where it will debut "the first laptop designed for Gemini Intelligence"
The Gemini 3.8 family is getting a little bigger with two new additions
Google is introducing to new models to the Gemini 3.8 family: Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking.
September's Pixel update squashes a ton of bugs
In addition to all the new Android 17 QPR1 features, Pixel users have a ton of patches and bug fixes coming their way.
Samsung Gallery is adding Google Photos backup to replace OneDrive
New updates are starting to roll out that will let Samsung Gallery replace its OneDrive integration with Google Photos support....
OpenAI says it’s working with Anthropic, Google on AI safety
OpenAI is working on steps to address artificial intelligence safety issues with its top competitors Anthropic PBC and Google DeepMind, escalating industry efforts to respond to a groundswell of concern
Google’s September Pixel Drop upgrades VIP contacts, scam detection, and Pixel Watch
Google's September Pixel Drop makes VIP contacts one tap away.
Google's Latest Pixel Drop Will Keep You More Connected To Your VIPs
Google has revealed the updates coming with the latest Pixel drop.
Google rolling out Android 17 QPR1 for Pixel
Android 17 QPR1 is rolling out to Google Pixel devices with the September 2026 Feature Drop and security patch.
Google just shared these fun Pixel 11 blueprint wallpapers [Gallery]
Google has shared some new blueprint-style wallpapers for the Pixel 11 series, and they look pretty awesome. Shared through the...
Looks like Samsung Gallery has already found its OneDrive replacement
As Samsung's partnership photo storage partnership with Microsoft winds down, the company is moving to Google to fill the gap.
Google Photos backup is coming to Samsung Gallery
A few weeks ago, Samsung announced that it would end support for Microsoft OneDrive integration with the Gallery app on Galaxy phones and tablets. Now, it has b
Amazon says Fire TV devices will be able to sideload Android apps again in upcoming update
Fire TV hardware will be able to sideload apps again with an upcoming Android software update for Amazon’s streaming hardware...
Community businesses need to show up in AI search, not just Google
Community businesses need to adapt to AI search to improve local visibility. It helps them connect with nearby customers and stay competitive as search behavior evolves.
China-Linked Hackers Exploit Chrome-Windows Zero-Day Chain to Deploy GRIMWEDGE
UTA0560 exploited a Chrome-Windows zero-day chain against NGOs to deploy GRIMWEDGE; APT31 used the same chain to install LONGTALE.
Meteor Pulse Launches Hiiiilo, an AI WhatsApp Receptionist
Kuala Lumpur agency Meteor Pulse has launched Hiiiilo, an AI-powered WhatsApp receptionist that answers, books and suggests add-ons for Malaysia...
Will Netflix’s Latest Horror Film Save Its Summer? And Introducing a New Data Source...
(Welcome to my weekly streaming ratings report, the single best guide to what’s popular in streaming TV and what isn’t. I’m the Entertainment Strategy Guy, a former streaming executive who now analyzes business strategy in the entertainment industry. If you were forwarded this email, please subscribe to get these insights each week.) All month, I’ve been polling you, my readers, on how we think the summer’s biggest blockbusters will do when they eventually land on streaming, and I wanted to share the results. First up, how readers think The Odyssey will perform when it comes to Peacock... People are pretty bullish that The Odyssey will smash it for Peacock. I don’t know that it will have as much rewatchability as Wicked did for kids, but we’ll see. It is now Universal’s highest-grossing film of all time. One note: almost half of Christopher Nolan’s films weren’t available for streaming the week The Odyssey came to theaters, including some of his biggest films like Inception, Dunkirk, Tenet, Oppenheimer and Insomnia. If you did want to watch the others, Prime Video had Memento and The Dark Knight trilogy, HBO Max only had The Dark Knight trilogy (even though Warner Bros. produced most of Nolan’s films), Hulu had The Prestige and Interstellar, and Paramount+ had Interstellar. So you needed at least two streaming subscriptions to catch the rest. Warner Bros. whiffed right here. I know they have long term licensing deals, but they should have tried to get as many Nolan films on HBO Max for launch as possible. Here are the results for Spider-Man: Brand New Day: The audience is a bit more mixed on this one, with everyone agreeing Spider-Man will swing to a big opening, but not necessarily elite or best all time. On to this week’s issue. Netflix continues searching for a hit straight-to-streaming summer film and Lioness and Ted Lasso allow me to look at the difference between a “hit” and a ‘hit for them”. All that, plus new YA/teen melodramas on Prime Video and Netflix, the House of the Dragon finale, USA Network claims a big opening, more big NFL ratings, all the flops, bombs and misses, and a whole lot more. But we start with a fun new data announcement. Let’s dive right in! (Reminder: The streaming ratings report focuses on the U.S. market and compiles data from Nielsen’s weekly top ten viewership ranks, Luminate’s Top Ten Data, JustWatch and Reelgood interest data, Samba TV household viewership, company datecdotes, Netflix hours viewed data, Google Trends, and IMDb to determine the most popular content. While most data points are current, Nielsen’s data covers the weeks of August 3rd to August 9th, 2026. You can find a link to my terminology here.) Subscribe A New Data Source: Reelgood’s Data on Content Libraries Here’s a fun new treat for you. Reelgood—an online service and app that lets users track where shows or films are streaming—is providing data every month for the Streaming Ratings Report on the size of each streamer’s catalogue, along with the new shows and films for each streamer. Yay! Previously, I had pulled “new shows on streamers premiering on the first of the month”: While I liked that data cut, it was a big wonky since it didn’t show all of the new shows or movies that came to a streamer each month. Instead, I would have preferred something like this from Reelgood’s annual report: And now we have it! I’m going to feature Reelgood’s data monthly, showing both total catalogue size and new shows and films appearing on the charts. That’s going to make these two cool new charts. First, here’s the size of various streamers’ film catalogues… (By the way, Reelgood offered to make the charts for me, but I like doing that task by hand to understand the data better.) They also provided data back to last year. Here’s the number of films timeline: And here are the new films on each streamer for the month prior to Aug: As I get used to this data, I plan to keep tinkering with how I display it, but I’m excited to feature it monthly. Today, I’m sharing the film charts; next issue, I’m going to feature TV shows by streamer. Now, in full disclosure, Reelgood is a sponsor of this little-newsletter-that-could. But in my constant quest to figure out what streamer various shows and films are on, I was already a Reelgood user before this sponsorship. I go to their website all the time, and their data is very accurate on where shows are (or have been). Plus, I’ve been sharing their weekly top ten lists for years now, pre-sponsorship. For me, this is a great win-win, since I get to amplify a service I love, and I get a cool new data source to play with. Now, normally folks don’t usually mention their sponsors directly, but as I wrote before, as I start featuring ads in my newsletter, I plan to be extremely open about who is sponsoring my writing so you can trust what I write. Television - Two Big Returning Shows Look Set to Take Over The Charts One key question when it comes to judging streaming shows is whether a show is a “hit” or a “hit for them”. To be clear, the former means a show that’s one of the top shows on streaming, no questions asked. The latter means a show is one of the biggest, for a given streamer. That distinction makes a big difference! It’s really good to be one of the biggest shows on a streamer; it’s much better to be the biggest on streaming, despite the size of a streamer. Because, as we’ve seen now, nearly every streamer—while lacking the number of subscribers or usage of Netflix—can make hit TV shows: Paramount released Landman, whose first and second seasons were among the top two TV shows the year they came out. HBO Max’s The Pitt’s second season is the biggest show of the year so far. Love Island USA was the top show in 2025 for Peacock. And, crucially for today’s discussion: Wayyyyyyyyyyyyyyyyyyyyyyyyyy back in 2023, Ted Lasso was actually the top streaming show of the year according to Nielsen! So even Apple TV (formerly +) can make hit shows. It just doesn’t make them that often. And looking at the Nielsen charts this week, we have two prime candidates for this discussion: That’s right: Ted Lasso returned to Apple TV for its fourth season. Did it return in all its expected glory? Sure, though right now it’s still just a “hit for Apple TV”, but we’ll see how long it lasts on the charts. Its ten episodes are coming out weekly. (Smart choice, Apple!) Ted Lasso started out strong, actually making the Nielsen charts the week before new episodes dropped, which is a rare feat. (Shows like Stranger Things or Bridgerton have pulled that off. That’s terrific company to join!) On Nielsen, it opened to 14.4 million hours in its first week, and 16.6 million the week of 10-Aug. Here’s how that compares to previous seasons: Apple put out that this is their biggest TV launch through two days ever, and looking at the chart and Apple TV’s viewership history, I’d believe it. Here’s Apple’s top shows for season two and beyond: Other data tells a (mostly) similar story. Ted Lasso made Samba TV and it has elite IMDb scores, an insane 8.7 on 465K reviews. The only outlier is Luminate, where the latest season only had 7.4 million hours, and no other seasons made the top ten. As you’ll see below, that’s well below some other solid, but not spectacular shows like Sterling Point, Ride or Die or Furious. I’d be surprised if Ted Lasso isn’t bigger than those other shows, but it’s fine for different data sources to occasionally disagree. And while it’s sucky to bring this up, given that Warner Bros. produced this show, I do wonder how the streaming wars may have been different if Ted Lasso landed on HBO Max initially instead of Apple. I don’t think one show can make or break a streamer, but given HBO’s brand, it could have helped bring people to that streamer much earlier in its lifespan. (Though yeah, I don’t love calling out executives for passing on hit shows, since that’s just the nature of the job. Even the absolute best execs will pass on lots of successful shows. The key is not greenlighting the under-performers/misses.) The other show I’m thinking about in the “hit or hit for them” category is Lioness from Paramount+. Like Ted Lasso, its first season missed the charts, but it’s grown its audience since then. Here’s how season three opened: We’re just getting started with this issue, but the rest is for paid subscribers of the Entertainment Strategy Guy, so if you’d like to find out… How Lioness did on streaming... Two new melodramas for Netflix and Prime Video... Whether Netflix’s latest horror film saved its summer... Hulu’s Furious and Prime Video’s Ride or Die continuing their strong runs... The latest high profile showrunner to flop on streaming... All the flops, bombs and misses... And a whole lot more... ...please subscribe! We can only keep doing this great work with your support. Next week, we’ve got a double issue—since the week of 17-Aug had very few notable new shows or films—so I’ll look at HBO’s Lanterns, Prime Video’s Reacher, Netflix’s Tires, Outer Banks (which posted up huge numbers four years ago), an Aussie romance, and another baseball event (that didn’t resonate). And a bunch of movies like the latest dramedy from an Adam Sandler daughter, Don’t Say Good Luck, Camp Rock 3 on Disney Channel and Disney+, Michael on Starz, and more. Later this month, we’ve got a ton of shows and films to analyze! Movies will be the big story, featuring a box office bomb (Supergirl on HBO Max) and an “underperformer” (The Mandalorian and Grogu on Disney+) contending with a bunch of star-studded straight-to-streaming films like Netflix’s The Whisper Man (Adam Scott, Michelle Monaghan and De Niro!), Mayday on Apple TV+ (Ryan Reynolds and Kenneth Branagh!), The Runner (Gal Gadot) and The Last Sunrise (Eva Longoria and Maia Reficco) on Prime Video. Plus, a ton of TV shows have sophomore seasons coming out. Will The Gentleman put up big numbers on Netflix or suffer from a sophomore slump? Leanne and Beauty in Black on Netflix seem small, but they both might be bigger than you think. Hulu has a bunch of shows: a new season of The Secret Lives of Mormon Wives (the first after all the controversy last year) and the second season of Chad Powers (from Peyton Manning’s production company and starring Glen Powell). This one flopped last year; can it score a touchdown this year? Same goes for FX’s Adults, which I didn’t track last year; will it hit this year? Peacock’s The Office spinoff, The Paper and Apple TV’s Dark Matter also return for second seasons. And Netflix has more Untold specials featuring Vince Young, Raygun (that horrible breakdancer) and Mr. T. Long term, Paramount+ is rebooting yet another franchise, Clueless, and I’m somewhat bullish on this new show. The talent attached (Alicia Silverstone) and creative Josh Schwartz and Stephanie Savage (of The O.C. and Gossip Girl) also feel solid...if they make it funny like the original.
Google finally lets all engineers use Anthropic's Claude
Google engineers can finally use one of Anthropic's strongest coding models for internal work.
Googlebook pre-orders finally have a date: Here's when you can buy one
Google's first Googlebooks will be available to pre-order on September 21.
Google finally sets a date for Googlebook pre-orders
Planning on getting a Googlebook? Google has officially revealed when you'll be able to place your pre-order.
Google Maps on Android Auto is finally getting a speedometer that’s always on
Last week's bug was actually a feature
Samsung launches ‘Re-Newed’ Galaxy S26 series, $1,099 for Ultra
Samsung is now selling its Galaxy S26 series flagships at a discount with refurbished models available for around $200 off....
Samsung pokes fun at iPhone to Android switching drama in reality star-packed ad [Video]
Switching from iPhone to Android is a process that’s getting continually better and, in a new ad, Samsung pokes fun...
Pre-Orders For The Android-Based Googlebook Laptops Open On September 21
Gemini Intelligence is at the heart of the operating system powering Google's upcoming laptops.
Samsung reportedly considers early ‘Galaxy Z TriFold 2’ release in wake of iPhone Duo
In the wake of iPhone Duo, Samsung is reportedly considering an early release for a Galaxy Z TriFold sequel. The...
Googlebook pre-orders start on September 21
After previously teasing a fall timeframe, Google is set to open pre-orders for Googlebooks on Monday, September 21.
Google’s official Pixel 11 series cases are already on sale
Following last month’s launch, Google’s official Pixelsnap cases for the Pixel 11 series are already on discount at Amazon.
Google rolls out Motion Assist bubbles to Pixel, Samsung
After announcing earlier this month, Google has rolled out Motion Assist to Pixel and Samsung phones running Android 17.
What we know about Lenovo’s mysterious third Googlebook ahead of tomorrow’s NYC media event
Tomorrow morning in New York City, Google is finally taking the stage to pull back the curtain on Googlebook - well, for invited media, that is. Over the last few weeks, the leaks have been flowing...
Google Maps rolling out Android Auto speed limits and speedometer that show even without navigation
While the speedometer has been hit or miss for some Android Auto users as of late, Google Maps is now...
Google Wallet for Android getting vertically stacked card list
Following the homepage redesign and fullscreen passes revamp this year, Google Wallet is trialing a vertically stacked card list.
Top AI companies have discussed creating their own standards body
Anthropic, Google and OpenAI have discussed creating an AI industry standards body, people familiar with the situation told CNN.
Kansas Faces Growing Demand for Nurses as Workforce Trends Reshape Healthcare Needs
Aging populations, rural care gaps, and workforce shortages drive demand for nursing education and career advancement across the state.Kansas Ci...
5 Budget Android Phones You Should Consider Instead Of The Google Pixel 11
Considering a Google Pixel 11 smartphone, but feeling put off by the $900 price tag? Why not consider these five much cheaper Android phone alternatives.
Anthropic selects Nasdaq for IPO, Business Insider reports
Sept 13 (Reuters) - U.S. tech startup Anthropic has selected Nasdaq for its potential initial public offering, Business Insider reported on Sunday, citing a person familiar with the company's plans. (Reporting by Ananya Palyekar in B...
City Hall Construction Is Squeezing Old Town Businesses. Alexandria Is Looking for Ways to Help.
Businesses beside Alexandria City Hall say the years-long renovation is cutting foot traffic and parking. The City is now considering parking help, promotion and other relief.
Movies & Series Launches Expanded Arabic-First Entertainment Discovery Platform for Egypt and MENA
Movies & Series has launched an expanded Arabic-first entertainment discovery experience that brings films, television, cinema releases, str...
Broadcom’s Anthropic Ramp Puts Google’s TPU Business in a Different Light
Broadcom’s September 2 results show custom AI hardware turning into substantial sales. On the accompanying call, management said it expected to accelerate Ironwood shipments to Anthropic in the fourth quarter. That puts Google’s TPU technology at the center of a relationship whose economics differ for Broadcom Inc. (NASDAQ:AVGO) and Alphabet Inc. (NASDAQ:GOOGL). Broadcom can sell […]
Google Health's mistakes are a grave warning for Apple’s new AI-powered Health app
Fitbit users know how this turns out
Video: Rare Mastodon Tooth Found in California Creek
Brigid Lynch, a geologist, was conducting a field survey in San Mateo County, Calif., when she saw a strange object sticking out from the creek. After a reverse Google image search, and an expert’s confirmation, she learned she had found a Pacific mastodon molar.
Botpool Create Launches AI Website Builder Backed by Human SEO Experts for Small Businesses
The AI website builder lets small business owners describe their business in plain language, publish a live site the same day, and start capturi...
Google Search Redesign, Business Profile Post Views
Google redesigned Search results across the EEA, Business Profile posts get view counts again, and ChatGPT Shopping now leans harder on product feeds.