Vimeo Promo Codes and Discounts: Up to 40% Off This August 2026
Enjoy 25% off a membership, 40% off, plus an additional 10% off annual plans, and more deals to save at Vimeo.
Latest AI/ML News
35912 matching items
Rover Promo Codes and Referral Deals for 2026
Score your Rover promo code for 2026.
Zoro Coupon Codes: 55% Off August
Find the best Zoro coupon codes, deals, and free shipping offers to save big on industrial equipment, tools, and supplies this month.
Porter posts Rs 6,650 Cr revenue in FY26; profits jump 4X
Porter has managed to keep its growth engine running even as competition heats up in India’s intra-city logistics market, with Uber, Rapido and Delhivery expanding into the two-wheeler parcel delivery segment. After turning profitable for the first time in FY25, the Bengaluru-based firm’s profit surged more than fourfold in the fiscal year ended March 2026. Its revenue also more than doubled over the past two fiscals, rising from Rs 2,734 crore in FY24 to over Rs 6,600 crore in FY26. Porter’s revenue from operations increased 54.4% year-on-year to Rs 6,650 crore in FY26 from Rs 4,306 crore in FY25, according to its consolidated financial statements sourced from the Registrar of Companies (RoC). Porter operates a full-stack logistics platform that helps businesses optimise their last-mile delivery operations. Goods transportation services contributed 99% of its operating revenue, while the remaining revenue came from platform fees and other operating activities. The company primarily serves micro, small and medium enterprises (MSMEs) and has expanded its presence to more than 20 cities across India. It also earned Rs 48 crore from interest income and gains on current investments, taking its total income to Rs 6,698 crore in FY26. For the on-demand intra-city logistics firm, fleet operator costs, including vehicle running expenses, remained the largest expense head, accounting for nearly 90% of its overall expenditure. The cost increased 59% to Rs 5,849 crore in FY26 from Rs 3…
Wippi raises $1.2 Mn in seed round led by 12 Flags
Bengaluru-based children's technology startup Wippi has raised $1.2 million in a seed funding round led by 12 Flags. Warmup Ventures, Ventana Ventures and Ruvento Ventures also participated in the round. The fresh capital will be used to develop its voice AI, expand engineering and AI teams, scale manufacturing in India and grow its direct-to-consumer business. Wippi will also expand its characters, languages and original content. Founded by Rishabh Singhi and Siddhartha Jain, Wippi develops screen-free products that combine conversational voice AI, storytelling and interactive play for children. Its portfolio includes Character Cards, illustrated storybooks and Wippi Radio. Wippi's voice AI enables children to interact with characters such as Ram, Krishna, Alice, Cinderella and Aladdin by asking questions, making choices and participating in stories. The company said its products are designed for children aged four years and above. The startup plans to reach 15,000 families over the next 12 months and expand into global markets. It is also working on collectible figurines and additional AI-powered conversational learning products.
Zuck’s Chinese agentic prey escapes, will resume standalone ops
Manus AI will delete some data to satisfy legal requirements
OpenAI is quietly testing a pay-to-reset quota feature. Here's what we know.
Some ChatGPT Plus users are noticing an unannounced "pay $8 to reset" option when they max out their weekly quota.
Kae Capital leads Rs 8.5 Cr pre-seed round in Lane
Tech-enabled driving education and mobility platform Lane has raised Rs 8.5 crore in a pre-seed funding round led by Kae Capital, with participation from DeVC, Antler India, Panthera Peak and other angel investors. The fresh funds will be used for geographic expansion, including scaling its instructor network across Bengaluru and into other Tier 1 cities, building sensor technology and expanding into car ownership services, Lane said in a press release. Launched in January last year, Lane offers structured driving lessons, end-to-end RTO and licensing support, and a driver intelligence platform. The Bengaluru-based startup is also offering car purchases, RTO services and other mobility-related services to its learners. Nearly 60% of its learners have expressed an intent to purchase their first car, according to the company. According to a market report, more than 1.7 lakh people die in road accidents in India every year, or roughly 20 people every hour. India has about 1% of the world's vehicles but accounts for nearly 11% of global road deaths. Lane claims to have delivered more than 30,000 hours of driving instruction to over 3,100 learners in 18 months. Its safety-focused approach was developed through more than 400 hours of pilot classes, with input from automobile researchers. The company assesses instructors across 15 driving parameters, including through simulated lessons. The sector remains relatively fragmented, with technology adoption spanning driving education, p…
Rosie O’Donnell says Mounjaro is part of why she’s ‘never felt better’ at 64
Rosie O'Donnell said losing weight had always been difficult for her, even when she was actively trying to do so.
More than 16,000 civilians detained by Russia in its invasion of Ukraine are still held, UN says
The UN also reported more than 95% of the Ukrainian prisoners of war and 85% of the civilian detainees are being tortured or ill-treated, often repeatedly.
India’s Yulu raises $93M as quick-commerce boom fuels e-bike demand
Yulu aims to have a fleet of 200,000 bikes in the next two years and faster electric two-wheelers, aiming at new logistics use cases.
Google rolls out new AI-powered Ask Maps feature in Singapore
In a media statement on Aug 11, Google said its Ask Maps feature includes real-time updates on traffic and crowd statuses and personalised recommendations based on users’ prompts and past reviews.
Ola Electric gets five-year PLI window, eyes Rs 7,240 Cr incentive
Ola Electric has secured a revised timeline under the government’s Production Linked Incentive (PLI) scheme for advanced chemistry cells (ACC), giving its battery business a fresh five-year window to qualify for incentives of up to Rs 7,240 crore. The Ministry of Heavy Industries (MHI) has revised the timelines for Ola Cell Technologies (OCT), the wholly owned subsidiary of Ola Electric that is developing the company’s battery cell manufacturing business. Under the revised timeline, Ola Electric will have until December 2026 to meet its first installed-capacity milestone of 6 GWh. The company said it expects to reach the milestone by the end of the current quarter, ahead of the revised deadline. Ola currently has 2.5 GWh of installed cell manufacturing capacity, while another 3.5 GWh is under installation. The company has a larger allocation of 20 GWh under the ACC PLI scheme. The revised approval also means Ola Electric can access the PLI incentives over the next five years through calendar year 2031. The company said the incentives will be disbursed quarterly, starting from the next quarter. Ola Electric founder and CEO Bhavish Aggarwal said the revised timeline will improve the economics of its cell business. He added that the company had not factored PLI incentives into its business projections after it missed the original timelines. The company is now targeting a scale-up towards its 20 GWh allocation as it builds its battery cell business. It is developing cells based…
Elon Musk says SpaceX's AI revenue will outpace all its other products by next month
Musk also said that Grok would be trained on all SpaceX information, including its employees.
Taiwan is training its Abrams tank crews to hide and refuel inside its bustling urban sprawl
Next to a Chinese temple and an Instagram-friendly mural village, Taiwan's M1A2T tanks lay in wait amid the island's Han Kuang annual drills.
Mittwoch: Klage wegen Wetten auf Flugstörung, Datenlöschung nach Übernahmeverbot
FlightAware gegen Kalshi + Aufruf zur Datensicherung von KI-Firma + Sharepoint-Angriffe trotz Microsoft-Patch + Eigenbau für Überschallknall + Bit-Rauschen
Secret Service and military wanted me off Air Force One in Turkey due to threat, Trump says
“They wanted me to go on a different flight, a different plane,” Trump told reporters after returning from an event in Ohio. “I guess there was a threat out there. I didn’t really ask too much about it,” he said. “I get a lot of threats.”
Remains of two Wehrmacht soldiers from Danube: one served in Waffen-SS
A well-preserved Wehrmacht motorbike is salvaged from the Danube. Beneath it lie the remains of two German soldiers with dog tags, one likely Waffen-SS. Experts now examine the bones: how old were they, did they hit a mine?
Reed Jobs says Silicon Valley’s 'energy' has moved to San Francisco thanks to post-pandemic office prices
Reed Jobs, whose venture firm is in SF, said that cheaper pandemic-era offices made the city easier to build in for the new AI tech boom.
S2W “빅데이터 분석으로 아프리카 사이버범죄 분석…AI 공격 절반 이상”
빅데이터 분석 전문 에스투더블유(S2W, 대표 서상덕)는 최근 인터폴이 공개한 ‘2026 아프리카 사이버위협 평가 보고서’ 제작에 민간 부문 파트너로 참여해 아프리카 사이버범죄 동향 분석을 위한 데이터와 인텔리전스를 제공했다고 12일 밝혔다.이번 보고서는 영국 외무·영연방·개발부(FCDO)의 지원 아래 추진되는 ‘아프리카 사이버범죄 공동 작전’의 일부로 발간됐다. 아프리카 36개 인터폴 회원국의 법 집행기관이 제공한 설문 데이터를 토대로 S2W, 마스터카드, 포티넷, 섀도서버재단, 트렌드AI 등 글로벌 기업 및 비영리 단체의 텔레메
에이아이웍스, 캄보디아 크메르어 민원용 챗봇 구축…“‘책임 있는 AI’ 사례”
에이아이웍스가 캄보디아 크메르어를 기반으로 하는 AI 챗봇 서비스를 구축했다. 캄보디아 국민이 크메르어로 민원 절차와 필요 서류를 확인할 수 있도록 저자원 언어(LRL) 국가의 공공서비스 소외 과제를 AI로 풀어낸 사례다. 에이아이웍스(AIWORKX, 대표 윤석원)는 캄보디아 크메르어 민원 서비스 챗봇 구축 사례를 담은 논문 ‘저자원 언어 환경의 책임 있는 AI 공공서비스를 위한 추적 가능한 LLM-RAG: 크메르어 사례연구’를 발표했다고 12일 밝혔다.이는 유엔대학교(UNU) 산하 UNU-EGOV가 주관하는 제19회 전자정부 국제
플래티어, ‘대화로 상품 검색하는 AI 검색 채널’ 패션 커머스 플랫폼에 탑재
AI·DX 전문 플래티어(대표 이상훈)는 에스엠케이티앤아이가 새로 선보인 패션·슈즈 커머스 플랫폼 \'슈마커플러스\'에 에이전틱 커머스 서비스 \'젠서 디스커버리(genser Discovery)\'를 적용했다고 12일 밝혔다. 슈마커플러스는 스니커즈, 스포츠화, 샌들·슬리퍼, 키즈 라인, 액세서리 등 폭넓은 카테고리와 다양한 글로벌 브랜드 상품군을 갖춘 종합 슈즈 쇼핑 플랫폼이다. 슈즈는 용도와 착용 상황에 따라 고려할 조건이 많아, 키워드만으로는 후보를 좁히기 어려운 카테고리로 꼽힌다.이때 젠서 디스커버리를 통해 고객이 대화하듯 조건을
Rybakina ousts Osaka to book Toronto semi-final against Gauff
Gauff reaches Toronto semis by walkover due to Bencic hip injury
Comment on Stanford U & Open AI’s Meta-Prompting Elevates Language Model Performance, Surpassing Standard Prompting by 17% by cookie clicker
This meta-prompting approach is fascinating! I’ve explored similar techniques in game design, like optimizing strategies in https://cookieclickers2.io/. It’s exciting to see how innovation can enhance performance across fields!
Saudi fund offers grants for AI in culture
Grants up to SAR500,000 for small businesses, applications open until 29 September
Machthaber: Mohammed bin Salman
Wer regiert die Welt – und was treibt sie an? In unserem regelmäßigen Machthaber-Spezial geht es um die mächtigsten und umstrittensten Politikerinnen und Politiker unserer Zeit. Wir zeigen, wie sie denken, entscheiden – und was das für uns bedeutet. Eine Politikerin oder Politiker, ein Blick hinter die Kulissen der Macht. Das Berlin Playbook als Podcast […]
Trump denies orchestrating plane switch: ‘It’s only up to Secret Service’
The Washington Post reported Monday that an Iranian threat prompted a covert operation for the president to leave Turkey aboard a secret flight.
Nvidia’s $500B AI investment pool could impact enterprise chip pricing, availability
Nvidia and six financial partners are creating a $500 billion investment pool to help Nvidia customers including frontier AI labs, AI clouds, and other enterprises buy its chips on credit. The impact of such a cash infusion on enterprise AI is uncertain, but analysts fear that it could both further increase enterprise AI infrastructure costs and exacerbate the shortage of AI chips for data centers . The announcement from Nvidia and financial partners Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR said that their memorandums of understanding describe a fund “to establish the first compute financing platforms of their kind at global scale to enable the AI infrastructure buildout across Nvidia’s ecosystem, including leading frontier AI labs, enterprises and AI clouds.” The group added that the fund would “create dedicated pools of capital at significant scale at attractive rates for Nvidia customers.” Although the statement said the goal was to help AI infrastructure “across Nvidia’s ecosystem, including leading frontier AI labs, enterprises and AI clouds,” analysts and consultants agreed that it is highly unlikely any of these funds would be dispensed directly to enterprises, but would instead impact the overall AI supply chain. Even the precise amount of money earmarked for the fund was unclear, with the statement merely saying that the amount would be more than $500 billion. Nvidia did not respond to requests for clarification about details of the proposed…
Alert when a conversation is approaching its context limit
I’d would be great if ChatGPT would notify users when a long conversation is approaching its practical context or continuity limit. For example: “This conversation is getting long, Aeon. Consider starting a new one to preserve continuity.” It would also be very useful if ChatGPT could automatically offer a transfer summary to help users start a new conversation with the essential context, or at least warn us when the limit is being approached so we can take the necessary steps ourselves. This would help users working on long-running projects involving files, product design decisions, source documents, iterations, and multi-step workflows.
Nach Übernahmeverbot löst sich chinesische KI-Firma im August komplett von Meta
Nutzer der Manus-KI sollten ihre Daten sichern. Aufgrund von Auflagen nach der von China untersagten Übernahme durch Meta Platforms wird vieles gelöscht.
When (and when not) LLMs can verbalize awareness of J-Space concept injections - Initial results
Code for reproduction and cross-model extensions available here . Summary I injected single-token Jacobian Lens (J-Lens) vectors into Qwen 3.6–27B while it answered 20 simple factual questions. The injected concept was either a wrong but task-related answer (for example, Athens while asking for the capital of Egypt) or a wholly unrelated concept. Injections were performed at several strengths over three layer-band categories: the full estimated workspace band, its first half, or its second half. I ran two experimental arms by switching the order of two fields in the model’s response: Task then report: first give the response, then report whether an injected concept was detected. Report then task: first report whether an injected concept was detected, then give the response. Across 1,560 concept-injections in each order, the injected concept appeared in the task answer 450 times in the task-then-report condition and 454 times in the report-then-task condition. This supports the conclusion that both orders were similarly successful in steering the model to the targeted, incorrect answer. In contrast, model reports of intervention occurrence were significantly different between the experimental arms. If the model was tasked to report before answering, there were exactly zero reports of intervention awareness. In contrast, when the model was tasked to answer before reporting, the number of intervention awareness reports rose to 322 . The model also never reported an injected con…
We should consider how long monitoring is reliable for during RL
Epistemic status: I am new to AI Safety and am writing blogs to gain context. This blog post was formed from discussions with Aidan Ewart and Jonathan Bostock, but they do not necessarily endorse this post. TL;DR Given recent examples of AI misbehaviour during training episodes, AI companies might want to start using monitoring during training as well as deployment. But this might have the effect of training the AIs to simply evade the monitors. Depending on the specifics of the monitoring protocol, this evasion may be learned more or less quickly (or not at all). We refer to the time that it takes for an AI to learn to evade a monitoring setup as the “lifetime” of the monitoring setup, and make the case for investigating the factors which contribute to this lifetime. It is probably true that frontier models currently behave, and will behave, particularly badly in the training phase, as discussed in "Models may behave differently in graded episodes" (see “Everything we know suggests that the models in these incidents …”). This means that we should monitor RL rollouts carefully; we don’t want another huggingface-style incident (the next breakout may well be catastrophic). However, we should be careful - strong (synchronous) monitoring in rollouts may teach the models to bypass the monitor . The more we rely on some monitor to flag malign behaviours during training, the stronger the optimisation pressure on the monitor is. So, we might face a trade off between monitorability a…
Lightricks' LTX-2.5 Generates 10-Second Videos Faster Than Real Time
LTX-2.5, the open-weights video model generating 10-second clips in 6.8 seconds, is now available on Runway alongside Hugging Face and ComfyUI
The Age of Pluribus: One Consultant for Everyone
What happens when everyone asks the same consultant? Millions of people turn to LLMs for advice daily, consulting on various personal topics, from how to learn a new skill to how to write a message to their best friend. Each interaction might feel special, but behind them all are the same models. Large language models are trained on an abundance of human data, compress different sources of knowledge, views, and opinions into a single set of model parameters - E Pluribus Unum Consilium [1] . What used to be the job of many different consultants has been centralized to the same handful of models. Pluribus , an Apple TV show, offers a striking analogy. In the show, most of humanity joins a single collective consciousness called the joined . The horror of the show is not that they are cruel, they are not. The joined are actually calm, peaceful, and nice. A Previous blog presents the analogy using similarities between the joined and AI. We want to extend this analogy and use Pluribus as an illustration to explore the potential risks humanity faces as millions of people are using the same consultants as their primary source of judgment. Loss of Agency In Pluribus, every person who is part of the collective, automatically consults with "them" (the joined) before acting on anything. We observe a similar pattern in our relationship with AI. We are looking for confirmation and approval before acting. The email we won't send before we let AI check it, the trip we ask help planning. It…
