AI News Archive: August 19, 2026 — Part 5
Sourced from 500+ daily AI sources, scored by relevance.
- Maximize AI Impact: More Model Choice, Smarter Routing
AI usage doesn't equal productivity. See how dynamic model routing in Cortex AI Gateway helps enterprises achieve real intelligence efficiency.
Score: 49🌐 MovesAug 19, 2026https://www.snowflake.com/content/snowflake-site/global/en/blog/ai-intelligence-efficiency-dynamic-model-routing - I Tested Google's Pixel 11 Pro for a Week, and Its AI Showed Me the Future of Smartphones
I Tested Google's Pixel 11 Pro for a Week, and Its AI Showed Me the Future of Smartphones PCMag
- Gisting: Compressing LLM Agent context to ↑ throughput and ↓ cost
A technique to reduce LLM agent context size, boosting throughput and cutting costs.
- Google Pixel 11 Pro XL Review: Unmatched Gemini AI in a Refined Design
Google Pixel 11 Pro XL Review: Unmatched Gemini AI in a Refined Design PCMag UK
Score: 49🌐 MovesAug 19, 2026https://uk.pcmag.com/mobile-phones/166817/google-pixel-11-pro-xl-review-unmatched-gemini-ai-in-a-refined-design - Alibaba Earnings Will Test Whether the Stock Is Among China’s AI Winners
Alibaba Earnings Will Test Whether the Stock Is Among China’s AI Winners Barron's
Score: 48🌐 MovesAug 19, 2026https://www.barrons.com/articles/alibaba-earnings-stock-price-0f727edc?mod - Prompts aren’t enough: Why 2 AI giants are evolving their approach
This summer, OpenAI and Anthropic both announced features that let their AI systems learn a task by watching someone do it once, instead of being told how in a prompt. In June, OpenAI launched Record & Replay , a feature that lets ChatGPT and Codex users demonstrate a workflow and turn it into a reusable skill. Weeks later, Anthropic unveiled Record a Skill inside Claude Cowork: Record your screen doing a task, narrate your reasoning, and Claude turns it into a skill it can run again. Two competitors, converging on the same solution within weeks of each other. To me, that’s an admission that prompting alone was never going to get AI where it needs to go. I’ve been thinking about and working toward this milestone since my time at Apple, where I worked for 12 years, spending my formative years building the Chinese version of Siri. As a founding engineer, I was excited to watch Siri break new ground as a voice-activated assistant, which let people use natural language instead of tapping through menus. The next challenge was context awareness: building a system that could complete the first request correctly, as well as carry that understanding into the ensuing command. The gap was never linguistic. It was that the system needed to retain the memory of someone’s preferences or habits across interactions. For example, tell a voice assistant to set an alarm for 6 every day, and it’s unclear whether the time is for morning or night. Most people don’t think of that as ambiguous, because they know their own schedule, and they expect whoever’s listening to know it too. That disconnect between what people say and what they actually mean is the same one that OpenAI and Anthropic are now aiming to close. Show, don’t tell How someone works is more idiosyncratic than what software assumes. Even if two people have the same job and responsibilities, they’ll use different tools, follow distinct sequences, and rely on a lot of unspoken context. Researchers call this tacit knowledge : what people know but can’t quite articulate, a term coined back in 1966 by the philosopher Michael Polanyi. One study estimated that 40% of a company’s valuable knowledge is inside individual employees’ heads, never written down anywhere. We often experience this limitation when using AI prompts. If you ask someone to describe how they complete an expense report, they’ll say that they upload a receipt, categorize it, and then submit it. But what they’ll likely omit is that they ask their manager to review meals over $75, or that they classify their client dinners differently from team lunches. That’s the challenge that prompt-based AI can’t address, because it can’t correct its way to context that hasn’t been stated. In contrast, when you demonstrate the same task to an AI system, it captures the sequence, the decision points, and the small judgment calls that are intrinsic to how someone gets the job done. A demonstration observes all the nuances of a workflow, because the context and the action go hand in hand from the very beginning. Build a demo library Record & Replay and Record a Skill are real progress in understanding and mimicking how we work. Paired with a scheduled task, a recorded skill doesn’t need someone to reopen it, because it can run autonomously and in the background while the user works. As more people use demonstrations to automate their workflows, the next obstacles are ensuring that people recognize what they can hand off, and making this process more intuitive. A study from the AI company Glean found that employees spend roughly 6.4 hours a week “botsitting”: babysitting AI tools, correcting their output, and reexplaining things the tool should already know. Imagine how much of that wasted time could be recouped once people get into the habit of recording what they do, so a system can learn the workflow once and then take over. There’s also an emerging opportunity for workflow mining, allowing technology to observe how work is done and extract reusable knowledge from it. Workflow data will become invaluable, not only because it automates multistep tasks from end to end, but also because it can create a library of workflows that people can select and personalize. Someone can build a workflow for closing out expense reports, while someone else adapts it for their own approval chain. As a result, the library gets sharper with every version added, the same way open-source code improves as more developers build upon it. The real test OpenAI and Anthropic’s launching demonstration features within weeks of each other is a clear signal that the biggest gains will be in systems that learn how people actually work. Most of us still default to prompting because we’ve become accustomed to it, but one demonstration will save the back-and-forth prompts and iterations every time. We’ve spent years teaching AI to understand what we say, and we can now treat a demonstration the way we’d treat training a new hire : something you do once, carefully, so you don’t have to explain it again. Judge these tools less by what they can already do, and more by whether you’ve shown them how you work.
- Tom Clancy's Rainbow Six(R) Siege Cuts Critical Toxic Voice Chat by 50% with Modulate's ToxMod
Tom Clancy's Rainbow Six(R) Siege Cuts Critical Toxic Voice Chat by 50% with Modulate's ToxMod USA Today
- Houston hospitals test patient-focused AI technology
Houston hospitals test patient-focused AI technology Houston Chronicle
- US vs China AI Race: How ChatGPT, Gemini, Deepseek, Kimi Agents Compare
A spate of new releases show Chinese competitors are closing the gap on Silicon Valley, and winning market share with lower prices.
- Salesforce expands Headless Data 360 for MCP so developers can bring insights to agents
Agentic artificial intelligence runs on data, so Salesforce Inc. today announced it’s launching Headless Data 360 for Model Context Protocol to provide that data directly to agents, allowing them access to relevant, governed customer context. Teams can do more than query data through Data 360. Now they can build, transform, map, segment and activate fields […] The post Salesforce expands Headless Data 360 for MCP so developers can bring insights to agents appeared first on SiliconANGLE .
- Bandhan Life CTO bets on AI as 80% of non-medical policies get same-day approval
Bandhan Life CTO bets on AI as 80% of non-medical policies get same-day approval Techcircle
- Flock Has a Powerful New AI Tool for Police. We Got Its Code
Flock’s surveillance cameras have already sparked outrage. WIRED reconstructed its next-generation AI system, already in use by some police, to confirm it goes much further than tracking license plates.
- Large language models in medicine: New review analyzes risks and strategies for safe use in clinical practice
Large language models in medicine: New review analyzes risks and strategies for safe use in clinical practice EurekAlert!
- Humanoid robots don’t deserve their superhuman valuations
Machines are currently far too pricey for industrial viability, let alone consumers
Score: 48🌐 MovesAug 19, 2026https://www.ft.com/content/ba2a65de-7b9a-4938-bde6-91d41722ea08?syn-25a6b1a6=1 - From spectacle to scale: why China’s robotics firms face a ‘critical juncture’
At China’s annual World Robot Conference (WRC) that kicked off on Wednesday, the hall was filled with humanoid robots wowing visitors with feats of dexterity: from playing table tennis to performing drum solos. These eye-catching displays have helped China’s robotics industry win global attention – and raise huge amounts of capital – over the past couple of years. But at this year’s summit, the overriding sentiment among executives, investors and policymakers was starkly pragmatic: the era of...
- Hesai raises non-LiDAR revenue outlook as physical AI businesses gain traction
Its Kosmo platform has begun securing commercial orders from humanoid robotics companies.
Score: 48🌐 MovesAug 19, 2026https://kr-asia.com/hesai-raises-non-lidar-revenue-outlook-as-physical-ai-businesses-gain-traction - Criminal AI tool Kriminal is mostly just Grok with a jailbreak, ThreatDown finds
ThreatDown said in new research published today that Kriminal, one of the newest and most popular tools in the criminal artificial intelligence market, owns almost nothing it sells. The service runs on SpaceXAI’s Grok, rented from the same legitimate AI industry it claims to have circumvented. Paid access starts at $12.99 a month. Kriminal is […] The post Criminal AI tool Kriminal is mostly just Grok with a jailbreak, ThreatDown finds appeared first on SiliconANGLE .
- Google built a camera-based tool to estimate body fat more accurately than wearables
Google’s PhotoScan AI turns simple phone photos into surprisingly accurate body fat readings.
Score: 48🌐 MovesAug 19, 2026https://www.androidauthority.com/google-research-photoscan-body-composition-3700687/ - AI may be learning from billions of images without copying any one of them
An MIT study finds that AI-generated images often cannot be traced to individual training images, as larger datasets cause the influence of specific examples to fade.
Score: 48🌐 MovesAug 19, 2026https://www.digitaltrends.com/cool-tech/mit-study-ai-images-training-data-attribution/ - New study reveals that AI is changing who builds software
A new study from project management platform provider Linear has revealed that AI adoption throughout organizations has grown rapidly, and that the use of AI in product development is no longer just an engineering concern. While the early adopters of AI were primarily software developers, its use has reached marketing and design, and all the... … continue reading The post New study reveals that AI is changing who builds software appeared first on SD Times .
Score: 48🌐 MovesAug 19, 2026https://sdtimes.com/ai-adoption/new-study-reveals-that-ai-is-changing-who-builds-software/ - You Already Own These AI Agents, but Are You Using Them?
The AI Investment That Keeps Paying Ask any financial services leader what’s keeping them up at night, and you’ll hear a version of the same story: customer expectations have never been higher, margins…
Score: 48🌐 MovesAug 19, 2026https://www.salesforce.com/blog/agentforce-financial-services-templates/ - HoverAir’s ingenious self-flying drone can transform into a pocket gimbal camera — and unlike DJI drones, it just got FCC approval in the US
HoverAir, makers of the original self-flying drone, have launched the Versa on Indiegogo, and it's set for US availability.
- AI labs are failing to keep their own systems in check
No AI company fully applies basic control measures to its own internal AI systems. The article AI labs are failing to keep their own systems in check appeared first on The Decoder .
Score: 48🌐 MovesAug 19, 2026https://the-decoder.com/ai-labs-are-failing-to-keep-their-own-systems-in-check/ - Fish Audio models now available on Vercel AI Gateway for free
Fish Audio 's audio models are now available on AI Gateway. To celebrate the launch, every Fish Audio model is free on AI Gateway for the next 30 days, through September 18. Capability Regular Through September 18 Text-to-speech $15.00 per million characters Free Speech-to-text $0.36 per hour of audio Free Four models from Fish Audio are available, including their latest text-to-speech model: fish-audio/s2.1-pro (text-to-speech): Built for low-latency streaming; clones a voice from a reference recording. fish-audio/transcribe-1 (transcription): Returns the text along with the duration of the audio and timestamped segments, down to individual words. fish-audio/s2-pro (text-to-speech): Covers around eighty languages and takes inline tags, plain-language directions written into the text itself, so you can change how a single word or phrase is delivered instead of setting one style for the whole request. fish-audio/s1 (text-to-speech): Reads text that can carry markers for emotion, tone, and sound effects. How to use models during the offer period Using the standard model name (i.e., fish-audio/s2.1-pro ) is free, but will automatically begin billing when the offer period ends. To ensure you aren't billed after the free period, add the -free suffix to the standard name, and the model will stop serving when the offer ends (i.e., fish-audio/s2.1-pro-free ). Speech and transcription ship in the current AI SDK 7 release. Text-to-speech Generate spoken audio from text with generateSpeech and write the result: Speech-to-text Transcribe recordings into text with transcribe . The audio can be a buffer, a base64 string, or a URL: Each segment carries the text and its start and end time in seconds, down to individual words. Playground You can also try the Fish Audio models without writing any code. Open the models list , click into a model, and send text or audio to hear or read the result in your browser. For a full overview of how to utilize audio models, refer to the speech quickstart . For more detail on Fish Audio, refer to the AI SDK docs for the provider. Read more
Score: 48🌐 MovesAug 19, 2026https://vercel.com/changelog/fish-audio-models-now-available-on-ai-gateway-for-free - National AI race narrows to three finalists
National AI race narrows to three finalists 매일경제
- Should data centre developers be more accountable? Pennsylvania's Governor thinks so
Should data centre developers be more accountable? Pennsylvania's Governor thinks so thenationalnews.com
Score: 48🌐 MovesAug 19, 2026https://www.thenationalnews.com/future/technology/2026/08/19/shapiro-data-center-ai-three-mile-island/ - Universal-3.5 Pro for pre-recorded audio: code-switching and contextual prompting in action
Showcases how Universal-3.5 Pro handles code-switching and contextual prompts in pre-recorded audio.
Score: 48🌐 MovesAug 19, 2026https://assemblyai.com/blog/universal-3-5-pro-code-switching-contextual-prompting - Teachers Warn That Students Are Losing the Ability to Think as They Lean on AI for Everything
"If writing is thinking, then any part of the struggle that is outsourced to technology amounts to relinquishing some freedom to perceive." The post Teachers Warn That Students Are Losing the Ability to Think as They Lean on AI for Everything appeared first on Futurism .
- Yotta and IntelliDB partner to build sovereign AI database infrastructure
Yotta Data Services and IntelliDB Enterprise have partnered to offer a managed, AI-ready database platform on Yotta’s sovereign cloud, targeting enterprises and government organisations that need to deploy production AI workloads while keeping data within India. The post Yotta and IntelliDB partner to build sovereign AI database infrastructure appeared first on Express Computer .
- Programmable photonic neural engine with all-optical nonlinear activation and 40,000 connections
Science Advances, Volume 12, Issue 34, August 2026.
- ChatGPT Is an Advertising Platform Now. Here’s What That Means for Your Business.
ChatGPT Is an Advertising Platform Now. Here’s What That Means for Your Business. entrepreneur.com
- Analog Devices' quarterly forecast tops estimates on AI-fueled chip demand
Analog Devices' quarterly forecast tops estimates on AI-fueled chip demand Reuters
- Inside Big Tech’s Frantic Race to Quell the Growing Backlash to AI
Tech companies are holding listening sessions, offering guaranteed jobs and writing big checks to win public support for the data centers they need.
- Exclusive: Replit taps OpenAI’s low-cost Luna model for new ‘Free Mode’
Exclusive: Replit taps OpenAI’s low-cost Luna model for new ‘Free Mode’ Fortune
- No-Filter 'Kriminal' AI Platform Raises Cybercrime Concerns
The AI company officially forbids illicit use, while offering guardrail-free social engineering, offensive cybercrime, and OSINT scanning to anyone with a bit of cryptocurrency.
Score: 48🌐 MovesAug 19, 2026https://www.darkreading.com/application-security/no-filter-kriminal-ai-platform-cybercrime-concerns - Graphwise aims to become the semantic layer for AI agents after securing major investment from Oakley Capital
Graph database startup Graphwise said today it has secured a significant, though unspecified, investment from the European private equity firm Oakley Capital as it looks to beef up its go-to-market strategy and expand globally. It will also use the funds to pursue strategic acquisitions, it said. Oakley has acquired a majority stake in the Bulgarian startup […] The post Graphwise aims to become the semantic layer for AI agents after securing major investment from Oakley Capital appeared first on SiliconANGLE .
- Running Local LLMs on the Arduino® UNO™ Q Board: a Practical Guide
Learn how to run local LLMs on Arduino UNO Q — from model selection and quantization to memory constraints and real-world edge AI use cases. The post Running Local LLMs on the Arduino® UNO™ Q Board: a Practical Guide appeared first on EE Times .
Score: 48🌐 MovesAug 19, 2026https://www.eetimes.com/running-local-llms-on-the-arduino-uno-q-board-a-practical-guide/ - How Harvey Builds AI Around Domain Expertise
Also, OpenAI disappoints some investors and shares of Unitree Robotics soar more than 600%.
Score: 48🌐 MovesAug 19, 2026https://www.wsj.com/cio-journal/how-harvey-builds-ai-around-domain-expertise-9c5bd91a?mod=rss_Technology - Opinion: AI has created a shadow medical system
Financial incentives – health care makes up nearly one-fifth of the American economy – have driven every major AI company into health care.
- Using synthetic data for AI training is 'a big mistake,' says AI pioneer Rich Sutton
Using synthetic data for AI training is 'a big mistake,' says AI pioneer Rich Sutton Business Insider
Score: 48🌐 MovesAug 19, 2026https://www.businessinsider.com/synthetic-data-for-ai-training-is-big-mistake-rich-sutton-2026-8 - Quantitative and interface-aware prediction of peptide–protein interactions by VITAL
Nature Machine Intelligence, Published online: 19 August 2026; doi:10.1038/s42256-026-01291-z Chen, Wang, Li et al. introduce VITAL, a dual-channel deep learning framework that co-learns sequence and structural contexts to quantitatively predict peptide–protein interactions, map binding interfaces and estimate affinity.
- Demand For Dead Data; ChatGPT’s Miraculous, Disappearing Advertisers
Fire Sale Firestarters One surprise beneficiary of the Spirit Airlines bankruptcy debacle: Google. Google spent $10 million to acquire “a trove of deidentified business data, software code and operations records” from the now-defunct airline, Bloomberg Law reports. The data stockpile Google bought for a song includes 100 million emails, 500 million Microsoft Teams chats and […] The post Demand For Dead Data; ChatGPT’s Miraculous, Disappearing Advertisers appeared first on AdExchanger .
- How U.S. military funding propelled China’s robot dogs
How U.S. military funding propelled China’s robot dogs The Japan Times
Score: 46🌐 MovesAug 19, 2026https://www.japantimes.co.jp/news/2026/08/19/world/us-military-china-robot-dogs/ - Why tech companies are buying up tons of rare old books to train their AI models
NPR's Juana Summers speaks with 404 Media cofounder and reporter Emanuel Maiberg about his investigation into a shipment of rare books which ended up at an Amazon AI training warehouse.
- WRC 2026 Opens: 34 Guangdong Companies March In, Showcasing a Complete Robot Supply Chain
The 11th World Robot Conference opened August 19 at Beijing E-Town with over 300 exhibitors, 150-plus new releases, and a first-ever procurement day. Guangdong sent 34 companies spanning the full robot supply chain, from EVE Energy batteries and ORBBEC 3D vision to UBTECH and LimX Dynamics full machines, as 49 central state-owned enterprises joined for the first time with 12 application scenarios.
Score: 46🌐 MovesAug 19, 2026https://pandaily.com/wrc-2026-opens-guangdong-34-companies-full-chain-robotics-supply-aug2026 - AI SRE Done Right: Why Your Data Foundation Matters
Discover why most AI SREs underperform and the three-layer architecture that helps engineering teams troubleshoot incidents faster.
- Powering AI with Salesforce Data 360 & Unstructured Data
Customer insights rarely reside solely in structured databases in rows and columns. Instead, significant enterprise value remains trapped within unstructured content —such as PDFs, knowledge base articles, email threads, customer interaction transcripts, media…
Score: 46🌐 MovesAug 19, 2026https://www.salesforce.com/blog/powering-ai-with-salesforce-data-360-unstructured-data/ - Report: Amazon Is Buying Rare Books in Bulk and Destroying Them to Train AI — A Hidden AirTag Proved It
Report: Amazon Is Buying Rare Books in Bulk and Destroying Them to Train AI — A Hidden AirTag Proved It entrepreneur.com
Score: 46🌐 MovesAug 19, 2026https://www.entrepreneur.com/business-news/amazon-is-buying-rare-books-in-bulk-and-destroying-them-to-train-ai - Agentic Engineering in Jira: A Practical Guide
Agentic Engineering in Jira: A Practical Guide Atlassian
Score: 46🌐 MovesAug 19, 2026https://www.atlassian.com/es/software/jira/guides/agentic-engineering/overview - 'We operate fleets of smaller, lower-cost assets that are centrally controlled and highly scalable': How tiny uncrewed vessels are mapping the ocean and delivering instant marine data at scale
XOCEAN is changing the way ocean data is delivered