The US stock market traded close to record levels on Friday after the S&P 500 reached a new closing high. Investors assessed weaker retail sales, steady Treasury yields and renewed tension in the Middle East.The S&P 500 hovered near the flatline, while the Nasdaq Composite slipped about 0.2%. The Dow Jones Industrial Average also traded with little change. All three stock market indexes opened the session with small moves.S&P 500 Eyes Third Weekly AdvanceThe S&P 500 entered Friday on course for its third consecutive weekly gain. The index had added 0.5% during the week. Meanwhile, the Nasdaq had gained 0.2%, while the Dow had fallen 0.4%.Thursday marked another record-setting session for the S&P 500. The benchmark crossed 7,800 for the first time and reached an intraday record of 7,816.70. It also secured its 27th record close of 2026.Slower inflation supported stocks earlier in the week. Two inflation reports showed easing price pressures, lowering expectations for a Federal Reserve rate increase in September. Short-term Treasury yields moved lower after Friday’s retail sales report.The 10-year Treasury yield rose slightly to 4.65%, compared with 4.63% on Thursday. Investors also considered weak employment data from the previous week when reviewing the outlook for interest rates.Retail Sales Drop More Than ForecastUS retail sales fell 0.6% in July, according to Commerce Department data. Economists surveyed by Dow Jones had expected sales to increase by 0.1%.Sales excluding vehicles declined 0.3%. Economists had forecast a 0.2% increase. Lower gasoline prices affected the monthly figure, though weakness also appeared across other retail categories.The report pointed to softer household spending during July. Nevertheless, investors made few major changes to their stock positions after the release. The bond market also recorded limited movement.Separate data showed that US business inventories stayed unchanged in June. Economists had expected a 0.1% rise. Wholesale inventories increased 0.2%, while stocks held by manufacturers edged 0.1% higher.Retail inventories fell 0.2% during June. Business sales declined 1.1% after falling 2.1% in May. At June’s sales rate, companies needed 1.30 months to clear their inventories.Reddit Jumps on S&P 500 EntryReddit shares rose more than 14% after S&P Dow Jones Indices announced the company’s addition to the S&P 500. Reddit will replace AvalonBay Communities before trading begins on August 18.Index inclusion often creates demand from funds that track the S&P 500. Reddit recently reported second-quarter revenue of $805 million, representing a 61% increase from one year earlier.Its global daily active unique users increased 18% to 130.3 million. International users grew 28%, while global average revenue per user increased 36%. Earnings reached $1.25 per share.Reddit co-founder Steve Huffman linked the company’s advertising strategy to discussions within its online communities. “Forty percent of the conversations people are having organically on Reddit are more or less about what to buy,” Huffman said.AI Stocks Deliver Uneven MovesMemory and storage companies recorded strong gains as investors tracked demand from artificial intelligence services. Sandisk, Seagate Technology and Micron attracted renewed buying during the week.JPMorgan upgraded Sandisk from neutral to overweight and set a $2,250 price target. Sandisk shares have climbed 544% in 2026. The bank linked its outlook to growing NAND storage demand from AI inference workloads.NewStreet Research also upgraded Micron to buy and assigned a $1,250 price target. Micron closed Thursday at $949.83 after gaining 4% during the session.Applied Materials moved in the opposite direction. Its shares fell more than 4% after the chip equipment company released quarterly results. Adjusted earnings reached $3.50 per share on revenue of $9.12 billion.The semiconductor systems division generated $7.04 billion in sales. That figure narrowly exceeded the FactSet estimate of $6.96 billion. Still, the results did not match the expectations already reflected in the stock price.Oil prices changed little after rising earlier following reports of an attack on a tanker leaving the Persian Gulf. Brent crude slipped 0.2% to $86.92 per barrel as traders monitored regional shipping conditions.Overall, Wall Street remained near record highs despite weaker-than-expected July retail sales, as investors continued to assess interest-rate expectations and economic signals.Also Read: Stock Market Update: Nifty 50 Opened 0.14% Lower, Sensex Declined 176.53 Points 

Reliance Digital’s D.I.S.C.O. sale (Digital India Sale Crazy Offers) is now live with deals across phones, laptops, TVs and home appliances. The sale runs until August 16 and is available through Reliance Digital stores, RelianceDigital.in and the MyJio Store. Some smartphone deals start at Rs. 10,999, while selected products come with discounts of up to Rs. 30,000, including bank cards, paper finance and UPI payments. There are also some useful laptop and TV deals. The HP 15s is available for Rs. 47,990, while the Lenovo IdeaPad Slim 3 is priced at Rs. 37,990. The TCL 55-inch QLED 4K TV is listed at Rs. 40,990. Budget-friendly 4G smartphones at Rs. 10,999, while 5G options begin at Rs. 13,999. The sale also covers ACs, refrigerators, and washing machines.Picks for Students, Office Users, GamersStudents who need a laptop for classes, projects, and browsing can look at the Lenovo IdeaPad Slim 3. Its price makes it a useful choice for basic daily work without spending too much. Office users may find the HP 15s more useful. It comes with 8GB RAM and a 512GB SSD. This should be enough for tasks such as writing, browsing, video calls, and working with documents.Gamers have different needs. The Lenovo Legion Tab could be worth Rs. 24,999. It comes with 12GB RAM and 256GB storage, making it a portable option for gaming.The D.I.S.C.O. sale covers a wide range of products, so buyers can find deals for different needs. Still, prices and stock can change during the sale. It is worth checking the final price and any bank offer before placing an order.Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

The Intelligent InvestorBenjamin Graham’s The Intelligent Investor is among Buffett’s most frequently recommended books. Buffett credits Graham with shaping his investment philosophy and understanding of value investing. The book explains concepts such as margin of safety, market psychology, and disciplined decision-making. Buffett particularly values its emphasis on separating investment fundamentals from short-term market movements.Security AnalysisSecurity Analysis, written by Benjamin Graham and David Dodd, provides a detailed framework for evaluating investments. Buffett has repeatedly praised Graham’s influence on his thinking. The book focuses on analysing financial statements, determining intrinsic value, and distinguishing investment from speculation. Its principles remain relevant for investors seeking disciplined approaches to evaluating companies.Common Stocks and Uncommon ProfitsPhilip Fisher’s Common Stocks and Uncommon Profits influenced Buffett’s approach to analysing businesses beyond financial statements. Fisher emphasises studying management quality, competitive advantages, growth opportunities, and long-term business potential. Buffett has acknowledged Fisher’s influence alongside Graham’s. The book encourages investors to understand businesses deeply before committing capital for extended periods.Business AdventuresJohn Brooks’ Business Adventures examines major corporate events and decisions involving companies including Xerox, Ford, and General Electric. Buffett recommended the book to Bill Gates, who later praised it publicly. Rather than offering conventional investment formulas, Brooks explores management decisions, corporate culture, competition, and human behaviour, making it valuable for understanding businesses.The Most Important ThingHoward Marks’ The Most Important Thing focuses on investment judgement, risk management, market cycles, and second-level thinking. Buffett has praised Marks and his investment insights. The book encourages investors to think independently, recognise uncertainty, and understand risk beyond simply measuring volatility. Its lessons complement Buffett’s long-term, value-focused investment philosophy.Poor Charlie’s AlmanackPoor Charlie’s Almanack collects speeches, essays, and ideas from Charlie Munger, Buffett’s longtime business partner. Buffett has strongly endorsed Munger’s thinking and described him as an exceptional teacher. The book explores mental models, rational decision-making, incentives, psychology, and avoiding common mistakes, offering broader lessons that extend beyond traditional investment analysis.Where Are the Customers’ Yachts?Fred Schwed Jr.’s Where Are the Customers’ Yachts? uses humour to examine Wall Street and the behaviour of financial professionals. Buffett has recommended the book for its enduring observations about investing and market speculation. Its central message encourages investors to question conventional financial advice and remain cautious about costly complexity.Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

OverviewAI observability tracks model performance, costs, quality, reliability, safety, and infrastructure.Traces help identify failures across models, retrieval systems, databases, and tools.Effective monitoring combines telemetry, evaluations, privacy controls, and production alerts.AI observability is becoming increasingly important as artificial intelligence systems move from experimental projects into production environments. Unlike conventional application monitoring, which largely focuses on infrastructure, uptime, and performance, AI observability examines how models and AI-powered applications behave in real-world use.It involves collecting, correlating, and analyzing telemetry from AI systems to understand model behavior, token usage, costs, prompts, responses, tool calls, retrieval quality, and safety.What is AI Observability?Traditional observability relies on three core signals: logs, metrics, and traces. AI observability extends these signals to capture what happens during an AI workflow.For an LLM application, a single user request can trigger several processes. These may include the application, an LLM, a retrieval system, a database or vector store, and an external tool before another model call generates the final response.Without tracing these steps, identifying the source of a slow or incorrect response can become difficult. The problem could originate from the model, the retrieval layer, the database, the tool call, or the application itself. OpenTelemetry’s GenAI work aims to standardize telemetry for these interactions.Why AI Observability MattersAI observability matters as it helps teams understand how AI systems perform under real-world conditions, quickly identify failures, control costs, improve response quality, and maintain reliable applications.Monitoring Model PerformanceAI teams need to monitor response latency, time to first token, output speed, error rates, and model availability. GenAI observability metrics can include token usage, operation duration, time to first token, and time per output chunk.Tracking AI CostsLLM costs can increase as token consumption grows. Monitoring input and output tokens at the request level helps teams identify expensive prompts, inefficient workflows, and models that may be unnecessarily costly.Debugging AI FailuresAn AI system can technically complete an API request while still producing a poor result. Tracing individual model calls, retrieval operations, and tool calls can help identify where an AI workflow went wrong.Measuring QualityA successful API request does not necessarily mean a successful AI response. An answer may be inaccurate, irrelevant, or poorly grounded even when the system reports no technical error.This makes AI observability increasingly connected with evaluation systems that measure relevance, faithfulness, toxicity, and policy compliance. OpenTelemetry provides telemetry infrastructure, while output-quality evaluation requires additional evaluation systems.What Should AI Systems Monitor?AI observability can cover several areas, including performance, usage, cost, reliability, model behavior, retrieval-augmented generation, agents, safety, and infrastructure.Key measurements include latency, time to first token, throughput, requests, input and output tokens, cost per request, errors, retries, timeouts, rate-limit events, output quality, hallucination, relevance, retrieval latency, tool-call failures, and CPU, GPU, and memory usage.Also Read: Apple’s AI Race in China Heats Up with New Alibaba-Backed MoveMonitoring RAG and AI AgentsRetrieval-augmented generation introduces another layer that needs monitoring. Teams need to know which documents were retrieved, how long retrieval took, whether those documents were relevant, and whether the final response was grounded in the retrieved information.AI agents add further complexity since they can make multiple decisions and tool calls before producing an answer. Observability should therefore track the number of steps, tool-call latency, failures, repeated actions, unexpected loops, token consumption, task completion, and human intervention. OpenTelemetry is also developing conventions for AI agent observability.Building an AI Observability StrategyA practical approach starts by instrumenting the application to capture traces, metrics, and events around model calls, retrieval operations, and tools.Teams can then standardize telemetry using OpenTelemetry’s GenAI semantic conventions, followed by quality evaluations and production alerts for latency, errors, token consumption, cost, and quality degradation.Also Read: What Is Artificial Intelligence? Beginner's GuidePrivacy Remains a Major ChallengeAI observability can also pose data security risks when prompts, responses, or tool arguments are automatically stored. Telemetry may contain personal information, customer conversations, confidential business data, credentials, or proprietary documents.OpenTelemetry documentation highlights the potential sensitivity of GenAI input and output data, underscoring the importance of privacy controls when designing observability systems.Future of AI ObservabilityAI observability is moving beyond simple LLM monitoring towards end-to-end visibility across AI applications. The emerging approach combines infrastructure monitoring, LLM telemetry, agent tracing, RAG monitoring, quality evaluation, safety monitoring, and cost analytics.For organizations deploying AI in production, observability is no longer simply about knowing whether an AI API is available. It is about understanding the complete path from a user’s request to an AI decision and the final output.FAQsWhat is AI observability?AI observability monitors AI systems using telemetry to track performance, costs, reliability, quality, safety, and overall application behavior.Why is AI observability important?It helps teams identify failures, monitor performance, control AI costs, improve response quality, and maintain reliable production systems.What does AI observability monitor?It monitors latency, tokens, costs, errors, model behavior, retrieval quality, tool calls, infrastructure, safety, and response performance.How does AI observability help RAG applications?It tracks retrieved documents, retrieval latency, relevance and grounding, helping teams identify whether failures originate from retrieval or models.What challenges does AI observability involve?Privacy is a major challenge since prompts, responses, and telemetry may contain sensitive, confidential, or personally identifiable information.Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Grayscale has had a busy August, and not in a good way for three major altcoins. The firm trimmed its NEAR Protocol holding on August 6, withdrew its Hedera ETF registration on August 7, and dropped its Cardano ETF plans entirely. Institutional retreats like that make it a lot harder to call any of these three the best crypto to buy today.BlockDAG has no ETF drama to worry about because it has no institutional gatekeeper deciding its fate. Stage 1 is priced at $0.002 against a $0.10 launch reference, a 50x spread built into the presale itself, which is exactly the kind of structural independence that could make it the best crypto to buy today.1. BlockDAG (BDAG) - No Institution Needed to Believe In ItBlockDAG does not need Grayscale's approval to matter, and that alone makes the best crypto to buy today conversation around it very different. There is no team allocation anywhere across the 150 billion token supply, and BlockDAG is targeting $100 million in launch liquidity, structural choices made by the project itself rather than decisions an outside asset manager could reverse with a single filing.That structure sits underneath stage 1 pricing of $0.002, against a $0.10 launch reference, a 50x spread that exists because of where the presale sits today, not because of any speculative forecast. Every stage that closes shrinks that gap permanently, which is exactly the kind of built-in math that makes BlockDAG a real contender for the best crypto to buy today.None of this is happening in a vacuum either. X1 Miner is already live and putting BDAG into people's hands, while the Super App and BlockDAGX exchange move through active development. Structure, math, and shipping utility together are exactly why BlockDAG deserves a serious look as the best crypto to buy today, without needing anyone else's stamp of approval.2. NEAR Protocol (NEAR) - Quantum Security Meets Institutional TrimmingNEAR Protocol is trading near $1.60, down about 92% from its January 2022 all-time high of $20.37. The network extended its post-quantum protection on August 12, adding safeguards for BTC, ETH, and SOL held via NEAR Intents, building on quantum-safe signing that already went live with the July 20 mainnet upgrade using the NIST-approved FIPS-204 scheme.That genuine technical progress landed the same week Grayscale trimmed its NEAR position in the Decentralized AI Fund during a Q2 rebalance, even though NEAR remains the fund's largest holding at 31.35%. A new fee-burn mechanism starting in August could tighten supply over time, but for now NEAR is a harder case to make for best crypto to buy today than its roadmap alone would suggest.3. Hedera (HBAR) - A Support Test With an ETF SetbackHedera is trading near $0.0665, down more than 88% from its 2024 all-time high of $0.5692, and consolidating right at a critical historical support zone between $0.058 and $0.042 that has preceded major rallies in the past. Grayscale voluntarily withdrew its Hedera Trust ETF registration on August 7, filing a Form RW with the SEC and pausing a key institutional pathway.A mainnet transition to Block Stream, announced August 11 for an October rollout, aims to improve the network's underlying data architecture and could matter for enterprise adoption longer term. Right now, though, HBAR is testing support rather than building momentum, which keeps it out of the conversation for best crypto to buy today until that ETF path reopens.4. Cardano (ADA) - A Grayscale Exit and a Mixed ChartCardano is trading near $0.18, down more than 70% over the past year, right after Grayscale confirmed it is dropping its Cardano ETF plans entirely, erasing one of the biggest institutional catalysts ADA bulls had been counting on for months. The token still leads Layer 1s with a roughly 10% monthly gain, a bright spot buried under the bigger headline.Cardano's fundamentals include well-distributed token economics and a top-five ranking among Layer 1 protocols by trading volume, real strengths that rarely get credit when an ETF story dominates the news cycle. Even so, ADA's case for best crypto to buy today gets harder to make every time an institutional door closes instead of opening wider.So, What's the Best Crypto to Buy Today?NEAR, Hedera, and Cardano all had a rough August on the institutional front, a trimmed position, a withdrawn ETF filing, and a dropped ETF plan, three separate reminders that outside gatekeepers can move a chart just as easily as they can help it. None of the three needed that outcome, but none of them could stop it either.BlockDAG's stage 1 structure does not depend on anyone's approval: no team allocation, a 50x spread built into the presale itself, and a live mining app already in use. That independence is exactly why BlockDAG deserves serious consideration as the best crypto to buy today.

Flare has added FXRP as collateral on Derive, giving XRP holders access to on-chain options and perpetual futures. The launch comes as XRP network activity rises while new-user growth remains weak.The integration allows holders to mint FXRP through Flare’s FAssets system and deposit it on Derive. Users can then manage hedging, premium generation, and directional positions through one Portfolio Margin V2 account. At the same time, XRP averaged about 35,700 daily active addresses in August, up 35% from July’s 26,400. However, new addresses stayed near 2,260 per day.FXRP Opens On-Chain XRP Derivatives AccessDerive’s XRP options settle in USDC. When an option expires in the money, the platform pays the difference in USDC while FXRP remains posted as collateral. As a result, settlement does not move the underlying XRP. Sellers still need enough USDC for payouts and face margin and liquidation risks while maintaining positions.Derive combines options, perpetual futures, and spot trading within one portfolio margin system. The platform uses infrastructure from Lyra Finance and has about $118 million in total value locked.Nick Forster, Derive’s co-founder and chief executive, said XRP had lacked the infrastructure needed for a permissionless options market. He described FXRP as an on-chain route for XRP holders.Flare Expands FXRP Across DeFi MarketsFlare’s FAssets system represents XRP through an overcollateralized structure supported by independent agents and network data oracles. These include the Flare Time Series Oracle and Flare Data Connector.FXRP already supports lending, borrowing, and yield tokenization across DeFi applications. Its deployment rose from 82 million to 144 million XRP since February, according to the provided figures.Meanwhile, more than 40 million XRP has been earned through Flare’s Smart Accounts across nearly 24,000 accounts. Flare also listed an FXRP/USDC spot pair on Hyperliquid for cross-chain movement.DeFi analyst Will Procheska said XRP holders previously lacked a permissionless options market for hedging and yield generation. The Derive integration adds another use case for FXRP holders.Also Read: XRP On-Chain Metrics Explained: 7 Signals Investors Should WatchXRP Activity Rises as New Addresses Stay FlatXRP’s network activity has increased, yet its user base has shown limited expansion. Market analyst Xaif Crypto reported the 35% monthly rise in daily active addresses during August.Still, higher activity does not necessarily signal new demand. Existing users can increase transfers, trading activity, exchange movements, or interactions across the XRP Ledger ecosystem. Can stronger activity support XRP if new-user growth continues to lag? New-address data remains the missing part of the current network expansion picture.XRP traded near $1.01 according to the supplied CoinCodex data. Meanwhile, Santiment figures showed wallets holding at least 1 million XRP increased 32% over three months. That creates a split across several network indicators. Large holders increased their positions, existing users became more active, while new-address growth remained subdued.ConclusionFlare’s FXRP integration with Derive expands XRP into on-chain options and perpetual futures while keeping FXRP available as collateral. At the same time, XRP activity has risen sharply, whales have accumulated, and new-address growth has stayed limited, leaving broader network expansion dependent on fresh participation.

Find Beginner-Friendly Projects: Start with projects offering good first issues, clear documentation and welcoming communities to make your contribution journey easier.Explore Good First Issues: Look for beginner-labelled issues that match your skills, interests and available time before choosing your first contribution.Read Project Documentation: Understand contribution guidelines, coding standards, setup instructions and project expectations before making changes to avoid unnecessary mistakes.Improve Documentation: Documentation fixes are valuable contributions and provide beginners with an accessible way to understand projects while learning collaboration practices.Fix Small Bugs: Start with manageable bugs that help you understand the codebase, project workflow and review process without becoming overwhelming.Add Tests: Writing tests can improve project reliability while helping contributors understand existing functionality, expected behaviour and potential edge cases.Suggest Features Carefully: Discuss proposed features through issues before coding, allowing maintainers and contributors to align on requirements and implementation approaches.Contribute Beyond Code: Design, community support, translations and documentation can also create meaningful value without requiring advanced programming skills.Build Long-Term Relationships: Consistent contributions, respectful communication and responsiveness to feedback can help you become a trusted open-source community member.Read More StoriesJoin our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Overview:Twitch enabled a "Training for Generative AI" setting on August 12, 2026 that opts every streamer in by default.Eligible content includes livestreams, VODs, clips, chat messages, and channel images, and it feeds models used across Amazon, not just Twitch.Streamers must manually disable the toggle in their dashboard, and even then, chat in someone else's channel stays governed by that channel owner's setting.Twitch has made millions of streamers' content available for a new Amazon AI training use by default. The platform switched on a setting that allows livestreams, archived videos, clips, chat logs, and channel images to be used for training Amazon's generative AI models, and it did so without asking anyone first.Twitch framed the announcement as a gift: a new setting that lets creators opt out. The messaging obscured the more uncomfortable fact underneath it, which is that the opt-out only exists because opt-in would have failed. Chief Product Officer Mike Minton said as much during an August 12 livestream watched by nearly three thousand people. Asked why the company did not simply let streamers choose to join, Minton answered plainly: "If this was opt-in, nobody would opt in. That's honestly the answer."That admission turned a routine settings update into a case study on consent design.What Content Is Being Used?Twitch says the training program covers channel content broadly. That includes live broadcasts, videos on demand, clips, highlights, chat messages, on-screen text, and images tied to a channel. Minton offered one concrete example: stream audio could refine Amazon's speech-to-text models, which would then improve automatic captions on Twitch and elsewhere in the Amazon ecosystem. Twitch has said the resulting models may also be applied more broadly across Amazon's product lineup, meaning content trained on for a Twitch feature is not necessarily confined to Twitch.Chat carries its own wrinkle. Whether a message counts as training data depends on the channel owner's setting, not the chatter's own preference. A streamer who opts out still has no control over what happens to a comment they post in someone else's chat if that channel leaves the setting enabled.Why Was This Not Really New?Twitch's framing suggested this was the start of something. Reporting suggests otherwise. Ars Technica reported that Minton confirmed this at a 2024 event hosted by The Information. Amazon was already using Twitch data to train AI models. This was happening well before any public setting existed to control it.So what changed on August 12 was not the practice itself. Streamers finally got a labeled switch to try to stop it. That switch applies only to future use. Twitch's own language draws that line carefully. Opting out prevents content from being used for "future" model improvements. The status of anything already used for training remains unaddressed.What Default Setting Could Mean Legally?The default-on approach raises a real question about what meaningful consent looks like. A platform is introducing a new AI use for content people have already posted. Those people posted that content under different assumptions.US privacy regulators have grown more attentive to this issue. They are watching interfaces and disclosures that can leave users unclear about how their data is collected or used. The FTC has said something relevant here. Settings engineered to maximize data capture without a clear affirmative choice can raise concerns under its unfair-or-deceptive-practices authority.None of that means Twitch has broken the law. Whether the arrangement creates legal exposure would depend on the specific terms, the disclosures Twitch gave, the jurisdiction, and how the training program actually operates in practice, none of which is settled by a single quote from a product executive.The more immediate issue is transparency, not liability. Twitch is asking creators to actively reject a new use of their content instead of actively authorizing it, while its own chief product officer said an opt-in version of the same setting would likely see little participation. That gap between what the company built and what it says it expected users to want is what keeps this in the middle of the broader debate over AI-training consent.How To Opt Out?Streamers who want out need to act manually. Head into the Twitch Streamer Dashboard, open Settings, select Security and Privacy, and locate the "Training for Generative AI" toggle near the bottom of the page. Switching it off applies going forward. Some users have already reported the setting reverting to enabled after being turned off, so a periodic check is worth the two minutes it takes.What Streamers Should Actually Do?Treat the opt-out as necessary but not sufficient. Turn it off on the personal channel first. Then check moderator settings on any channel where chat participation happens regularly, since that exposure sits outside personal control. Screenshot the toggle state as a timestamped record, given the reports of it silently re-enabling. Finally, watch for regulatory movement in home jurisdictions, since default-on AI training settings are drawing sustained attention from privacy regulators generally, independent of how this specific case is eventually resolved.Final ThoughtTwitch built a system, then handed streamers the burden of escaping it. The next fight will not be about the toggle. It will be about what already got trained before anyone had one to flip.Also Read: Twitch & GTA Online Unite: Grab GTA$500K, Street Samurai Livery and Exclusive RewardsAlso Read: Australia Bans Twitch, Expands Social Media Age Limits for U-16 UsersYou May Also Like:Best Alternatives to Twitch Studio for Windows for 2025YouTube Monetization Gets Tougher: New Eligibility Rules for Creators ExplainedYouTube YPP Rules: Monetization Bar Doubles in 2027FAQs1. What is Twitch's new AI training setting?Twitch's Training for Generative AI setting allows streamers to control whether eligible channel content can be used for Amazon's generative AI training. The setting is enabled by default.2. How can Twitch streamers opt out of AI training?Go to Twitch → Settings → Security and Privacy → Training for Generative AI and turn the setting off. The change applies to future training.3. What Twitch content can be used for Amazon AI training?The setting covers eligible channel content, including livestreams, videos, clips, highlights, images and chat messages associated with the channel.4. Does opting out remove content that Amazon may have already used for AI training?No. The opt-out is forward-looking and does not automatically remove content that may have already been used or ingested for training before the setting was disabled.5. Does a streamer's opt-out protect their messages in other Twitch chats?No. Chat content is governed by the settings of the channel where the message was posted. A streamer who opts out on their own channel may still have messages used for training if they participate in another channel that has training enabled.Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Google has introduced Gboard Rambler, a Gemini-powered voice input feature on the Pixel 11 series, designed to make voice typing cleaner and more useful. Unlike conventional live dictation, Rambler waits until users finish speaking before converting the complete utterance into text.The Pixel 11 series launches in India on August 20, 2026, with prices starting at Rs. 89,999. Rambler is integrated into Gboard and works across apps with text fields, including Messages, Gmail, and Google Docs.How Gboard Rambler WorksRambler can be opened through the microphone icon in the top-right corner of Gboard, marked with Gemini’s spark. A waveform takes over the suggestion strip while the user speaks. Once the checkmark is tapped, the session closes, and the cleaned text appears in the selected field.The approach allows Gemini to receive the complete utterance before processing it. This means Rambler can understand changes made halfway through a sentence. For instance, if a user dictates a shopping list containing apples, bananas, and oranges, and then says they have dropped the apples, the finished list shows two items. Google demonstrated this behavior when it introduced Rambler in May 2026.Rambler also removes filler words such as ‘um’ and ‘ah’, handles repetitions and self-corrections, and can respond to commands to change a message’s style, expand it, or add emoji. Tapping the microphone again allows users to add to or revise the existing text.Standard Voice Typing Still AvailableGoogle has retained the older speech engine under the name Standard. Users can switch between Rambler and Standard in Gboard settings. Google describes Rambler as a way to ‘focus on your thoughts, and revise in your own words’, while Standard remains available for users who want word-for-word dictation. This distinction could matter for addresses, code, or exact names, where rewriting based on intent may not be suitable.A separate Personal Dictionary setting can also help Rambler handle relevant terms, including names, brands and Indian place names.Also Read: iPhone vs Pixel: Google Pixel Wins the AI Photo Editing TestGemini Expands Across GboardRambler is part of a wider Gemini push inside Gboard. Sign-to-text is also included in the same Gboard release, using DeepMind’s Sign Language-to-Text model to convert American Sign Language into written text inside the keyboard and Live Transcribe.Google says the training set contains more than 100,000 hours of sign-language data across more than 50 languages. Gemini also powers an upgraded autofill feature in the suggestion strip, although it requires explicit opt-in.Pixel 11 Price and AvailabilityPre-orders opened on August 12, 2026, while retail sales begin on August 20 through the Google Store, Flipkart, and authorized partners. The Pixel 11 starts at Rs. 89,999 for 12GB of RAM and 256GB of storage, while the Pixel 11 Pro, Pixel 11 Pro XL, and Pixel 11 Pro Fold start at Rs. 1,19,999, Rs. 1,34,999, and Rs. 1,86,999, respectively.Rambler debuts on the Pixel 11, Pixel 11 Pro, Pixel 11 Pro XL, and Pixel 11 Pro Fold. Google says wider Android availability will follow later, while the feature remains limited to select countries and languages at launch.Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Hackers are using BNB Smart Chain to store malicious instructions while fake CAPTCHA pages trick Windows users into executing malware. Microsoft Threat Intelligence linked the campaign to EtherHiding, which places attacker commands inside blockchain smart contracts.Attackers first compromise legitimate websites and inject JavaScript into them. That code connects to BNB Chain infrastructure and retrieves instructions from a smart contract tied to the ClearFake malware campaign.  The blockchain does not infect devices by itself. Instead, attackers rely on social engineering to persuade visitors to execute the retrieved commands on their computers.Fake CAPTCHA Pages Turn Visitors Into Malware RunnersVictims see what appears to be a normal CAPTCHA verification page after visiting a compromised website. The page then asks them to open Windows Run, paste clipboard content and press Enter. That action launches attacker-controlled instructions. Microsoft calls this technique ClickFix, while another version called TerminalFix directs victims to Windows Terminal or PowerShell.Why would a routine CAPTCHA ever need access to Windows Run or PowerShell? Microsoft says legitimate CAPTCHA checks should never require users to paste commands into system tools. Attackers can also misuse built-in Windows utilities, including PowerShell, cmd, rundll32 and scheduled tasks. Successful infections can expose passwords, steal credentials and create persistent access.BNB Chain Storage Makes Malicious Commands Harder to RemoveEtherHiding changes where attackers keep part of their malware infrastructure. Instead of depending only on conventional servers, they place malicious instructions inside blockchain smart contracts.Security teams can often block or remove malicious websites and command servers. Blockchain data creates a tougher challenge because changing the relevant contract requires control of its associated wallet.This resilience gives attackers a storage layer that remains widely accessible. Still, they need compromised websites and deceptive prompts to turn stored instructions into an infection. Microsoft said ClickFix and TerminalFix campaigns target thousands of consumer and enterprise devices globally each day. The resulting access can support deeper corporate network compromises.Also Read: FBI Arrest in Steam Malware Scam, Hackers Steal $220,000 in CryptocurrencyCampaign Builds on Earlier Blockchain Malware TechniquesAttackers have used blockchain infrastructure in earlier malicious operations. Previous campaigns have relied on Bitcoin transactions and networks, including TRON, Aptos, and BNB Chain.The technique does not expose a flaw unique to the BNB Chain. Instead, attackers exploit public blockchain infrastructure because it offers persistence, accessibility, and resistance to centralized removal.Once malware gains access, attackers can steal credentials, maintain persistence, and move deeper into corporate environments. Microsoft said such access could eventually support ransomware attacks or wider network compromises. For users, the warning remains direct. Legitimate CAPTCHA systems do not ask visitors to paste commands into Windows Run, Command Prompt, Terminal, or PowerShell.ConclusionMicrosoft’s findings show attackers are combining persistent blockchain storage with fake CAPTCHA prompts to deliver Windows malware. The BNB Chain itself does not infect users, but malicious smart-contract data can resist removal. Users should reject any CAPTCHA that asks them to open system tools or paste commands.Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Smart Desk Lamp: A smart desk lamp lets you adjust brightness and color, helping create comfortable lighting for work, reading, or late-night tasks.Wireless Charging Stand: Charge your phone while keeping it upright and visible, making this handy gadget perfect for busy desks and quick notifications.Smart Speaker: A compact smart speaker can play music, set reminders, answer questions, and control compatible devices without interrupting your work.USB-C Hub: A USB-C hub adds extra ports for your laptop, letting you connect displays, storage drives, keyboards, and other desk accessories.Smart Desk Clock: A smart desk clock can show time, weather, reminders, and notifications while adding a neat touch to your workspace.Mini Desk Vacuum: A small rechargeable desk vacuum quickly clears crumbs, dust, and dirt from your keyboard and desk without taking much space.Monitor Light Bar: A monitor light bar provides focused lighting without taking up desk space, making long screen sessions easier on your eyes.Smart Plug: A smart plug lets you control connected desk devices remotely and create schedules, helping make your workspace more convenient.Portable Monitor: A portable monitor gives you extra screen space for multitasking, making it useful for work, gaming, editing, and travel.Read More StoriesJoin our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

OverviewFunction calling lets an OpenAI model request an action by name, but the model never runs code itself; the application executes it and returns the result.The mechanism follows a five-step loop: a question arrives, the model picks a function, the backend runs it, the result returns, and the model writes a final answer.Structured Outputs and function calling solve different problems, and mixing them up leads to the wrong tool for the job.Ask an OpenAI model to run a live calculation that depends on current data, and relying on the model's own output is not the same as running a real calculator or function. Ask it to check today's weather in a specific city, and it has no way to reach outside its training data. Function calling exists to close that gap, turning a model that can only talk into one that can act.What Function Calling Actually DoesFunction calling connects an OpenAI model to external tools and systems. A developer describes a set of functions using a JSON schema: names, descriptions, and the arguments each one accepts. When a user sends a message, the model checks whether an available function matches it. If one does, the model does not execute anything on its own. It returns a structured JSON object naming the function and its arguments.That separation is deliberate. The model never touches a database, an API key, or business logic directly. The application receives the JSON, validates it, runs the actual function, and decides what happens next. Authentication and sensitive operations stay on servers the model cannot reach.The Five-Step Loop Behind Every Tool CallEvery function call follows the same pattern, regardless of what the function does.A user sends a question or request.The model reviews the available tools and selects one, along with arguments, or decides no tool is needed at all.The application executes that function using real code.The result gets sent back to the model as part of the conversation.The model reads that result and writes a natural-language reply.Developers can steer step two directly: auto lets the model decide, required forces a call, and none turns off tool use for that turn. This loop is what makes agents possible. A single message can trigger several rounds of this cycle before a final answer appears, and some requests call several functions at once, in parallel, rather than one at a time.Function Calling vs Structured OutputsThe two features get confused often, though they solve different problems. Function calling means "do something." Structured Outputs means "return something in this format."A workflow can use both together: a function call retrieves data, and Structured Outputs formats the final response.A Practical Example: Building a Weather AssistantConsider a simple assistant meant to answer weather questions. The developer defines a function called get_current_temperature, with a required argument for location. That definition gets included in the request alongside the user's message, something like this:{  "name": "get_current_temperature",  "description": "Get the current temperature for a given location,"  "parameters": {    "type": "object",    "properties": {      "location": { "type": "string" }    },    "required": ["location"]  }}A user asks about the weather in Lisbon. The model does not answer directly. It returns a tool called get_current_temperature with the argument "location": "Lisbon." The application reads that request, calls a real weather API, and gets back a value such as 21 degrees Celsius. That result gets appended to the conversation as a tool response, tied to the original call by its ID:{  "role": "tool",  "tool_call_id": "call_1",  "content": "{\"temperature\": \"21C\"} "}That full conversation goes back to the model in a second request. Only then does it write a plain reply describing conditions in Lisbon. Nothing about the temperature came from training data. It came from a live source, routed through a function the developer controlled entirely.Also Read: Muse Glimmer: Complete Guide to Meta’s Open Agentic AI ModelWhere It Breaks DownFunction calling is not free of tradeoffs. High-throughput workloads that need thousands of requests per minute pay a cost for every extra round trip a tool call adds. Creative tasks, like writing or brainstorming, gain nothing from tools built for retrieval and computation. Models can also select the wrong function or fill in plausible but inaccurate arguments, which makes validation on the receiving end a requirement, not an option.Strict mode reduces this risk with constrained decoding at the token level rather than checking output after the fact, forcing every argument to match the schema exactly. Even so, it does not remove the need for the application to verify what comes back before acting on it.Also Read: OpenAI Agents SDK: A Complete Guide for Developers\Final ThoughtFunction calling establishes the pattern: a model requests an action, and an application decides whether to grant it. That pattern has quietly become the invocation layer underneath a larger shift, the move toward standardized protocols like MCP, which extend the same idea by letting a single tool definition work across different models and providers.The mechanics here will likely stay stable even as the surrounding ecosystem changes shape. That core pattern is what makes any of this safe to build on.You May Also Like:White House Calls OpenAI, Google, Meta for AI Safety MeetingOpenAI Faces Another Senior Exit as COO Brad Lightcap Departs Amid IPO BuzzOpenAI Launches GPT-5.6-Cyber as Daybreak Adds Blue and Red Access TiersFAQs1. What is OpenAI function calling?OpenAI function calling allows a model to request a specific function or tool with structured arguments. The application then executes that function and returns the result to the model for a final response.2. How does OpenAI function calling work?The process typically involves five steps: the user sends a request, the model selects a tool, the application executes it, the result is returned to the model, and the model generates the final response.3. Does the OpenAI model execute functions itself?No. The model generates a tool call, but the application or backend executes the actual function. This keeps databases, APIs, credentials, and business logic under the developer's control.4. What is the difference between function calling and Structured Outputs?Function calling allows a model to request an action, such as querying an API or database. Structured outputs ensure that the model's response follows a defined data format, such as a specific JSON schema.5. What can OpenAI function calling be used for?Function calling can connect AI applications to databases, weather services, APIs, calculators, customer-support systems, inventory platforms, scheduling tools, and other external services.Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Phantom Blade Zero is heading to PC on October 29, 2026. The action game is also coming to PS5 on the same day. S-GAME is behind the title, which combines fast sword fights with a dark world inspired by Chinese martial arts and action movies.The PC version is already drawing attention, as players want to know if their current systems can handle it. The game runs on Unreal Engine 5 and features detailed environments, quick combat, and more than 30 weapons. These features could make it a demanding game to run at 60 FPS.The full PC requirements haven’t yet been confirmed, but the Epic Games Store page listed the required features. The game requires Windows 10 or Windows 11 (64-bit) as the operating system. The game requires an Intel Core i7-8700K or AMD Ryzen 5 3600 as the minimum processing requirement. Similarly, the game requires an NVIDIA GeForce GTX 1660 6GB or an AMD Radeon RX 5500 XT 8GB as the minimum GPU requirement. However, if players want better performance, the Intel Core i5-10600K or AMD Ryzen 5 5600X would be their best choice. For graphics cards, they need an NVIDIA GeForce RTX 3060 Ti 8GB or an AMD Radeon RX 6700 XT 12GB. However, in all cases, 16GB of RAM is mandatory. Phantom Blade Zero looks like a game that will need a decent gaming PC. Its fast fights and detailed visuals could put extra pressure on older hardware. Currently, the listings show that the game will cost Rs. 4,299 on PC and PS5 for the Standard Edition, and Rs. 5,199 for the Digital Deluxe Edition. Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Rajasthan Royals has clarified that ChatGPT did not independently discover Vaibhav Suryavanshi during IPL 2026. AI supported the scouting process alongside data, scouts and cricket experts.RR COO Alok Chitre explained the process during a Hindustan Times discussion with RR technology leaders and OpenAI India. The franchise combined AI tools with traditional scouting to spot young talent.Chitre confirmed technology played a role in finding Suryavanshi, while human judgment remained important. “Part of it? Yes,” Chitre said when asked whether AI helped identify the young batter.RR uses technology to study large amounts of cricket data more quickly. The system can help teams spot player patterns and performance details during scouting.Scouts still watch players closely before the franchise makes important decisions. Coaches also bring experience and cricket knowledge that AI cannot fully measure.Rajasthan Royals uses AI for several other parts of its IPL operations. These areas include auction planning, ticket sales, fan engagement and content creation.The franchise also held a GenAI hackathon before an auction. Participants created models that suggested possible 25-player squad combinations for Rajasthan Royals.RR later used some recommendations from the event during its planning process. The experiment showed how AI can support auction decisions without replacing management.The franchise also uses AI agents to study ticket sales and stadium occupancy. These tools can help identify empty sections and changing demand during matches.RR has also created an in-house GenAI studio for content production. ChatGPT helps create photoshoot ideas and visual concepts for commercial campaigns.OpenAI and Rajasthan Royals both support a human-led approach to AI. Technology can process information quickly, while cricket experts make the final calls.Suryavanshi’s story shows how modern scouting is changing. AI can help find useful patterns, while scouts still bring instinct, experience and cricket knowledge.Also Read: Gemini Hits 1 Billion Users: Can Google AI Beat ChatGPT?Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Portable Bladeless Mini Fan: A compact bladeless fan provides personal cooling without taking much space, making it useful at home, work, or during travel.USB Rechargeable Mini Fan: USB-rechargeable fans are easy to power with laptops, power banks, or adapters, making them convenient for everyday use.Neckband Bladeless Fan: Neckband fans keep your hands free while delivering airflow around your face, making them useful for walks, travel, and outdoor activities.Desk Bladeless Fan: A small desk fan fits neatly beside your computer and provides steady airflow while you work, study, or relax at home.Battery-Powered Mini Fan: Battery-powered models offer flexible placement because they do not always require a nearby power outlet.Pocket-Friendly Mini Fan: Lightweight mini fans are easy to carry inside handbags, backpacks, or luggage, making them practical companions for daily travel.Quiet Bladeless Mini Fan: Quiet operation makes these fans suitable for bedrooms, offices, and study spaces where loud background noise can become distracting.Adjustable Bladeless Fan: Fans with adjustable airflow settings let you control cooling based on your comfort, room temperature, and personal preference.Travel-Ready Mini Fan: A travel-friendly bladeless fan can provide quick relief from heat during commutes, holidays, outdoor trips, and long waiting periods.Read More StoriesJoin our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

An income tax notice can unsettle taxpayers after they file their returns. However, many communications involve routine checks, missing details or mismatched records.The department may request a correction, supporting documents or payment of an outstanding demand. Each notice serves a different purpose and requires a specific response.Routine Mismatches Can Trigger NoticesThe Income Tax Department compares returns with AIS, TIS and Form 26AS. These records help officials check income, deductions, transactions and tax credits.Form 26AS primarily contains TDS and TCS details. Meanwhile, AIS provides wider information, including interest income and reported financial transactions. Taxpayers can also challenge incorrect information through the AIS feedback facility.Vipin Upadhyay, partner at King Stubb & Kasiva, said, “Income tax notices should not necessarily be viewed as an indication of wrongdoing.” He described them as part of the tax verification process.A Section 143(1) intimation communicates the outcome of return processing. It may confirm a refund, report no additional liability or create a tax demand.The department may also propose adjustments for calculation errors or unsupported deductions. Taxpayers should compare the intimation with their ITR, AIS, Form 26AS and payment records.A Section 139(9) notice identifies a defective return. Missing information, inconsistent figures or an income and TDS mismatch can trigger this communication.Taxpayers must correct the listed defect within the given period. Otherwise, the department may treat the return as invalid. AY 2026-27 proceedings follow the Income-tax Act, 1961.Scrutiny Notices Need Detailed RecordsA Section 142(1) notice asks for information, documents or clarification. The assessing officer may request bank statements, investment details, deduction proofs or transaction records.Taxpayers should answer every question separately. They should attach only relevant documents and keep copies of each submission.A Section 143(2) notice concerns scrutiny assessment. The department uses this process to examine selected parts of an ITR more closely.A scrutiny notice alone never proves tax evasion. Still, taxpayers must support the figures reported in their returns. Relevant records may include salary documents, invoices, property papers and foreign-income statements.Section 148 covers income that may have escaped assessment. Taxpayers should check the assessment year, disputed transaction and deadline before replying.Reassessment cases can involve complex records or legal questions. A chartered accountant or tax lawyer may help when the disputed amount is large.An outstanding demand may arise from a TDS mismatch, interest calculation or earlier adjustment. Taxpayers should compare the demand with tax challans and filed returns.Section 245 permits authorities to adjust a refund against unpaid tax. Authorities must send a written intimation before making that adjustment.How Taxpayers Should RespondTaxpayers should authenticate every income tax notice before responding. The e-filing portal allows verification through the Document Identification Number or PAN and notice details.The department generally treats communications without a DIN as invalid, subject to permitted exceptions. Taxpayers should avoid replying through unofficial links or sharing login details.After verification, taxpayers should review the section, assessment year, questions and response deadline. They should then reconcile the ITR with AIS, TIS, Form 26AS and bank records.A clear response should address each issue and include supporting evidence. Unrelated documents can make the submission difficult to review.The e-Proceedings service allows taxpayers to view notices, upload documents and submit responses online. Outstanding demands appear under the “Pending Actions” section.Taxpayers should retain the notice, response, attachments and electronic acknowledgement. Timely action can prevent a routine discrepancy from developing into a larger compliance dispute.Also Read: India's New Tax Amendment Bill: Everything Businesses Need to Know Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Overview:Spot Bitcoin ETFs recorded $131.13 million in daily net outflows, with ARKB seeing the largest single-day withdrawal at $58.82 million.The White House is reportedly preparing an Aug. 19 meeting around crypto and the CLARITY Act.Trezor disclosed a third-party breach affecting 13,689 customers, while Solana’s Alpenglow upgrade targets roughly 150 ms finality.The crypto markets saw major developments as spot Bitcoin ETFs recorded $131 million in net outflow, while XRP Ledger added around 490,000 new accounts in the first half of the year. Also, the White House is planning a meeting about the CLARITY Act and Trezor data was breached, leaking email addresses and phone numbers of users.Bitcoin Witnessed $131 Million in OutflowsAccording to SoSoValue, yesterday spot Bitcoin ETFs had a total net outflow of $131.13 million. The spot Bitcoin ETF with the largest single-day net inflow yesterday was Grayscale Bitcoin Mini Trust ETF BTC, with a single-day net inflow of $38.9268 million. Next was Morgan Stanley ETF MSBT, with a single-day net inflow of $7.0806 million; MSBT’s historical total net inflows have reached $436 million. The spot Bitcoin ETF with the largest single-day net outflow yesterday was Ark Invest and 21Shares’ ETF ARKB, with a single-day net outflow of $58.8155 million; ARKB’s historical total net inflows have reached $1.277 billion. The total net asset value of spot Bitcoin ETFs was $77.271 billion, the ETF net asset ratio reached 6.07%, and cumulative historical net inflows have reached $51.846 billion.XRP Ledger Adds Nearly 490,000 AccountsXRP Ledger added 489,739 accounts during the first half of 2026, pushing its total to about 8.4 million as Ripple’s RLUSD stablecoin and broader financial applications expanded across the network. Public XRPL data shows wallet creation remained strong during the six months, giving the blockchain another measure of growth as it seeks to expand beyond its traditional role in XRP payments and transfers. Ripple’s dollar-backed stablecoin has become part of XRPL’s expanding payments infrastructure, creating additional reasons for individuals, exchanges, applications and businesses to establish accounts on the ledger.Stablecoins can generate different network activity from speculative crypto trading. Also Read: XRP On-Chain Metrics Explained: 7 Signals Investors Should WatchWhite House Plans Meeting Over CLARITY ActThe White House has reportedly planned an Aug. 19 meeting with crypto and prediction market executives as Polymarket traders place the CLARITY Act’s 2026 passage odds at 21%.Politico reported on Aug. 13 that the White House is expected to host executives from the crypto and prediction market sectors on Aug. 19, citing people familiar with the planned gathering. According to the report, the White House has not disclosed a formal agenda or released a list of participants. Executives from traditional financial companies may also attend, while it remains unclear whether President Donald Trump will take part. Trezor Data BreachTrezor warned 13,689 customers that their personal information was exposed after an unauthorized actor breached systems belonging to fulfillment provider ShipMonk.The hardware wallet maker said ShipMonk notified it on Aug. 10 after discovering unauthorized access to systems containing customer order data. Trezor’s own infrastructure, devices and wallet backups were not compromised.The incident affected customers in the United States, United Kingdom, Sweden, Colombia, Brazil, Italy and Portugal. Of those affected, 11,742 had their names, email addresses, phone numbers and shipping addresses exposed. Another 1,947 customers had names, cities and email addresses compromised.Alpenglow Upgrade Targets 150 ms FinalitySolana plans to replace TowerBFT with Alpenglow, targeting roughly 150 ms finality and expected to activate with Agave 4.3, currently targeted for October 2026. Alpenglow code is included in Agave 4.2 for testing but will not activate on mainnet until Agave 4.3. Agave 4.2 (stable 4.2.0 released Aug. 7; 4.2.1 pre-release Aug. 13) will also reduce slot times from 400 ms to 200 ms in four 50 ms steps starting the week of Aug. 17. Alpenglow replaces TowerBFT with Votor, removes on-chain vote transactions, uses BLS aggregation, and expects a 1.6 $SOL validator admission ticket per epoch. The governance vote passed with 98.27% support in Sept. 2025; an Alpenglow bug bounty (up to 50,000 SOL) runs Aug. 5-19. Also Read: Ethereum Price Prediction 2026: What Google Gemini Forecasts for ETH FAQs:1. How much did Bitcoin spot ETFs lose in daily flows?US spot Bitcoin ETFs recorded about $131.13 million in net outflows. ARKB led withdrawals with roughly $58.82 million, while Grayscale’s Bitcoin Mini Trust recorded the strongest inflow.2. How much did XRP Ledger grow in the first half of 2026?XRPL added 489,739 new accounts during H1 2026, taking its total to around 8.4 million. Growth has coincided with expanding payments, stablecoin and financial use cases.3. What happened in the Trezor data breach?A breach at fulfillment provider ShipMonk exposed customer information linked to 13,689 Trezor users. Trezor said its own wallet infrastructure, devices and backups were not compromised.4. What is the White House meeting about?The White House is reportedly planning an Aug. 19 meeting with crypto and prediction-market executives. The agenda has not been formally disclosed, but the CLARITY Act is a key policy focus.5. What is Solana’s Alpenglow upgrade?Alpenglow is designed to replace TowerBFT and target roughly 150 ms finality. It is expected to activate with Agave 4.3, currently targeted for October 2026.

Researchers at the National Institute of Standards and Technology (NIST) have developed a chip-scale optical isolator that achieves 30 dB peak isolation without using magnetic materials. The device is built on a silicon nitride and thin-film lithium niobate platform and is designed to work with standard semiconductor manufacturing processes.The magnet-free isolator achieves 30.6 dB ± 0.1 dB peak isolation at 789.7 nanometres. It maintains more than 24 dB isolation across a 30-nanometre bandwidth spanning 770 nm to 800 nm and can isolate two lasers simultaneously with more than 20 dB isolation within a 10-nanometre window, without heater adjustment.How NIST Isolator WorksOptical isolators work like one-way valves for light. They allow forward-propagating photons to pass while blocking backward reflections that can destabilize lasers, corrupt signals, and introduce noise in sensitive photonic integrated circuits.Conventional optical isolators typically rely on the Faraday effect and bulk magneto-optical materials. These materials are difficult to integrate with standard silicon-based microfabrication.The NIST team instead uses radio-frequency electro-optic modulation to create synthetic motion. The approach uses time-periodic phase modulation to cancel backward-propagating light while allowing forward light to pass.At its core is a four-channel Mach-Zehnder modulator that splits incoming light into four parallel interferometric paths. Each path is phase-modulated by radio-frequency waves applied through a gold coplanar waveguide beneath a 150-nanometre thin-film lithium niobate layer, bonded onto 350-nanometre silicon nitride waveguides.6.5 GHz RF Drive and 32 mW PowerThe device uses a 6.5 GHz base frequency, with a third harmonic at 19.5 GHz. About 32 mW of RF power is delivered to the device, significantly below the watt-level drive levels required by earlier traveling-wave isolator approaches.Six chromium microheaters tune the power balance and relative phase between the four channels. This allows the system to maintain the required interference conditions.The measured forward-to-backward peak-spectrum power ratio reached 30 dB, with maximum isolation of 30.6 dB ± 0.1 dB at 789.7 nm. The researchers also measured an excess insertion loss of 0.023 dB ± 0.018 dB under fixed heater bias, a variation that is statistically indistinguishable from zero.Also Read: AI Translator Headphones: How Real-Time Language Translation WorksDesigned for Atomic Clocks and Quantum SensorsThe 770-800 nm operating range covers important rubidium-87 transitions, including the D2 line at 780.2 nm, the D1 line at 794.8 nm, and the two-photon transition at 778.1 nm, as well as potassium-39 resonances.These wavelengths are relevant to chip-scale atomic clocks, quantum sensors, laser cooling systems, and atomic spectroscopy. The technology could also be adapted to telecommunications wavelengths by redesigning the passive splitters and combiners.Foundry Compatibility Could Enable Mass ProductionThe isolator uses heterogeneous integration, with a thin-film lithium niobate layer bonded onto silicon nitride. A 100-millimeter wafer can accommodate nine chips, each 22 by 20 millimeters, with 171 isolators in total.The NIST team said the approach provides a route towards mass-produced photonic integrated circuits using standard semiconductor infrastructure. The work was developed through the NIST-on-a-Chip program, funded by the US Department of Commerce through awards 70NANB23H104 and 70NANB23H105.Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Overview:Discover the four categories hidden among 16 seemingly unrelated words.Use hints featuring Bumblebee, Echo Chamber, India Ink, and Google Drive to narrow down the groups.Check the complete answers and previous Connections solutions from August 9 to 13, 2026.The NYT Connections August 14 grid is now live, offering players a new challenge. Words like Wolverine, HiveMind, Deep Pen, Google Drive, Bubble, etc., make no sense at first glance. To conclude and identify themes such as ‘black and yellow,’ ‘like-minded people,’ ‘comic book writers,’ and ‘triangle logos,’ players need to take a careful look at the grid. The world of word puzzles is massive, but hardly any others compete with the puzzles that The New York Times presents. Whether it's Strands, Connections, or Wordle, each of these web-based word puzzles has its specific word-guessing method. In NYT Connections, the theme isn’t stated initially, but once players guess the right sets, the game reveals the themes:Hints for August 14, 2026This puzzle is tricky, but some clues may make things easier for gamers. Here are some hints to help players solve today’s puzzle. Let's start with the hints category-wise:Easiest Group (Yellow): Today’s yellow group words offer a glimpse into the world where characters are all in black and yellow.Moderate Group (Green): Do you watch a lot of science fiction movies? These are the places where people think in a similar manner.  Challenging Group (Blue): If you are a comic book fan or create your own comic stories, you must have heard about these tools.  Most Difficult Group (Purple): Finally, the purple category words are all logos, but they have a particular shape. One Word Hints For Today’s PuzzleIf you are still a bit confused about the themes of today’s puzzle, below are the one-word hints. From every category, one word is listed below:BUMBLEBEEECHO CHAMBERINDIA INKGOOGLE DRIVEThese one-word hints will help you identify what other words of the category may contain. Answers for August 14, 2026The hints should have resolved the confusion and helped players form the groups with fewer errors. However, if that’s not the case and some are still trying to figure out today’s word groups, here are the solutions to the NYT Connections of August 14, 2026. Yellow Category (BLACK-AND-YELLOW): BUMBLEBEE, CAUTION TAPE, CHARLIE BROWN, WOLVERINEGreen Category (PLACES WHERE EVERYONE THINKS THE SAME WAYS): BUBBLE, ECHO CHAMBER, HIVEMIND, SILOBlue Category (COMIC BOOK ARTIST'S TOOLS): BRISTOL BOARD, DIP PEN, INDIA INK, STRAIGHT EDGEPurple Category (TRIANGLE LOGOS): BASS ALE, CITGO, DELTA AIRLINES, GOOGLE DRIVEThese are the four categories for today’s puzzle, and congratulations to those who have solved it without any help. Also Read: Today’s NYT Strands Hints, Answers for August 13, 2026 (Puzzle #893)Check out the image below to know how to make the sets of today’s puzzle:Previous Answers (August 9-13) Final WordsThe August 14 edition of NYT Connections offers players a satisfying mix of everyday iconic characters, science fiction references, and clever wordplay. What begins with familiar character appearances and science fiction references gradually evolves into subtler patterns, especially in the ‘logo’ phrases.  This steady progression in difficulty is what makes today’s grid feel rewarding rather than frustrating. Overall, today’s challenge is a reminder of why Connections continues to stand out as one of the most engaging daily brain teasers.Also Read: NYT Connections Hints, Answers for August 13, 2026 (Puzzle #1159)Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

Gold traded lower on MCX on August 14 amid a higher US dollar and 10-year bond yields, despite the US inflation data for July easing expectations of a rate hike by the US Federal Reserve in September. August gold futures fell 0.66% to Rs. 1,52,449. September silver futures declined 0.99% to Rs. 2,33,106. Meanwhile, Brent crude futures rose 0.08% to $87.14 per barrel. US West Texas Intermediate (WTI) edged lower by 0.14% to $81.36 per barrel.Domestic Gold Prices24K gold fell by Rs. 71 to Rs. 1,52,890 per 10 grams, while 22K gold also declined by Rs. 65 to Rs. 1,40,150. By city, Mumbai and Kolkata mirrored prices at Rs. 1,52,890, while Delhi was at Rs. 1,53,040, and Chennai at Rs. 1,53,170.US Gold PricesUS gold prices slipped on Friday as investors locked in profits after mild US inflation data pushed bullion to its highest level in more than two months and weakened the case for a near-term Federal Reserve ​rate hike. Spot gold was down 0.5% at $4,326.75 per ounce. US gold futures for December delivery fell nearly 1% to $4,382.50.Spot ⁠silver slipped ​0.4% to $64.17 per ounce. Platinum slipped 0.3% to $1,711.84, while palladium ​was steady at $1,306.98.Also Read: Nifty-Gold Ratio Falls to 1.6: Is India’s Stock Market Rally About to Begin?Key Levels to Watch"There is some episodic and more speculative capital that's maybe ​taking a bit of profit in gold, because there's not a near-term catalyst quite so potent immediately in front of us," said Ilya Spivak, head of global macro at finance content network Tastylive. "Gold may be setting ​up, with some choppy trading along the way, for a meaningful rally now. And if ​we can take out $4,400, I don't think $5,000 by year-end is any kind of a stretch."On MCX, gold faces resistance near the falling trendline around Rs. 1,56,000 and witnessed profit booking, with prices now finding immediate support near Rs. 1,52,600. A sustained break below this level could extend the correction towards Rs. 1,50,000. However, the broader structure remains constructive as prices continue to trade above key EMAs.