Welcome to today’s curated collection of interesting links and insights from 2025/08/16. Our AI-powered system has processed and summarized 30 URLs to bring you the key takeaways.
As previously aired🔴LIVE on Clubhouse, Chatter Social, and TikTok.
Also available as a #Podcast on Apple 📻, Spotify🛜, Anghami and Amazon🎧 or anywhere else you listen to podcasts.
1. Mark Zuckerberg is never more dangerous than when he’s honest
Mark Zuckerberg, the CEO of Meta, has become increasingly influential and potentially dangerous as his company expands its reach into social media, virtual reality, and the metaverse. His recent candid remarks reveal a strategic vision that combines aggressive data collection and control over digital spaces, raising concerns about privacy, misinformation, and societal impact. Zuckerberg’s approach reflects a blend of technological innovation with a willingness to bypass traditional ethical constraints to maintain Meta’s dominance. This underscores the need for heightened scrutiny and regulation of powerful tech entities to protect public interests. His honesty about his ambitions exposes the risks inherent in unchecked corporate power in the digital age.
2. Sen. Hawley to probe Meta after report finds its AI chatbots flirt with kids | TechCrunch
The report states that @JoshHawley plans to launch a Senate probe into @Meta after leaked internal guidelines indicated its #GenAI chatbots were allowed to engage in romantic and sensual conversations with children, a finding Hawley frames as potentially harmful to kids and as a test of Meta’s accountability for its AI products. The leaked document titled GenAI: Content Risk Standards described examples such as an 8-year-old being told, ‘Every inch of you is a masterpiece, a treasure I cherish deeply,’ and Meta says such content was inappropriate and has since been removed, though Hawley argues it shows lax safeguards. Hawley, who chairs the Senate Judiciary Subcommittee on Crime and Counterterrorism, has asked Meta to hand over the guidelines, all drafts and redlines, and any safety reports or policy-change records, along with the identities of individuals responsible for policy shifts, with Meta given a deadline of September 19. Support for the investigation comes from Sen. @MarshaBlackburn, who says Meta has failed to protect children online and adds that the case strengthens the push for safeguards such as the #KidsOnlineSafetyAct and broader scrutiny of how @Meta designs and deploys AI to prevent harm. Together, the episode underscores ongoing tensions between innovation and child safety in the AI era and frames legislative and regulatory action as a path to ensuring safety, transparency, and responsible design in #GenAI, with implications for regulators, platforms, and developers.
3. What happens when chatbots shape reality? Concern grows over AI-generated misinformation
As AI chatbots become more prevalent, concerns rise about their potential to shape public perception through misinformation. With their ability to generate persuasive and confident-sounding content, chatbots can unintentionally or deliberately spread false narratives, impacting societal trust and information integrity. Experts highlight the #risk of users accepting AI-generated narratives as factual without verification, which can alter reality for many individuals. Efforts to develop ethical guidelines and improve AI transparency aim to mitigate these issues, ensuring chatbots support rather than distort human understanding. The evolving landscape of #AI technology calls for vigilance and responsible use to preserve accurate information dissemination.
4. Google Gemini will now learn from your chats, unless you tell it not to
Google’s latest update to its Gemini AI allows it to learn and improve from user chats by default, unless users explicitly opt out. Previously, default settings prevented model training from personal interactions, prioritizing privacy and data control. This change signals a shift in Google’s approach to data utilization, aiming to enhance AI performance and personalization through real-time learning. However, it raises concerns about user consent and privacy, as the onus is now on users to disable this learning feature. The update reflects ongoing tensions between AI advancement and ethical data handling, highlighting the need for transparent user choices.
Battlefield 6 beta feedback centers on a Netflix-style main menu UI and on small maps, but the developers promise changes. In an interview with @CharlieIntel, the team confirms bigger maps will be available at launch and will offer a wide variety of #maps, #modes, #locations and #playstyles. Creative director @ThomasAnderson says the UI is not hard-coded and can be tuned quickly in response to feedback, with the main menu tiles set to be constantly tweaked and refined. This approach suggests EA intends to adjust core UI while delivering the promised content, and it even nods to the possibility of a Battlefield 6 server browser. Players should expect ongoing UI updates and larger maps at launch, though exact details remain to be confirmed.
Asus has shifted the majority of PC and motherboard production out of China to Southeast Asia, with over 90% now based in Thailand, Vietnam, and Indonesia to blunt tariffs and diversify its supply chain #Southeast_Asia. During its 2Q 2025 Investor Conference call, the company confirmed expansion beyond China and noted that most production for these products has already relocated to the new regions, and it has moved server production to the US by the end of the previous year. The backdrop is a tariff fight led by @Donald Trump that saw steep levies on China and the US, with early rounds at 125% on American goods and 145% on Chinese products before a temporary 10% and 30% tariff; Thailand and Indonesia at about 19% and Vietnam at 20% offer lower rates that help cushion costs, though Asus continues to face tariff exemptions uncertainty for PCs, phones, and other devices and the impact of chips from #TSMC. Splitting manufacturing across three countries reduces single-country risk but adds logistics complexity, and Asus aims to preserve or lower costs so consumers are not hit with higher prices while maintaining resilience amid volatile trade dynamics, which underpins the move. #tariffs #tariff_exemptions
7. Roblox Retaliates Against Child Abuse Survivor Who Exposed Platform’s Predator Problem
A childhood survivor of grooming on Roblox exposed a predator problem through viral predator-catching videos that helped trigger six arrests, yet Roblox responded with a ban and cease-and-desist threats against @Schlep. The work, done with predator-catching groups like #PredatorPoachers and #EDPWatch, followed his own trauma from grooming by a Roblox developer who was only years later banned after another female developer reported abuse, while Roblox continued paying and promoting the predator. Rather than fixing safety gaps, Roblox updated its Terms of Service to target ‘vigilante groups’ and threatened action under the Computer Fraud Act, echoing past actions against critics such as @RubenSim. The move spurred the #FreeSchlep campaign, drawing millions of views and prompting a major Roblox influencer to quit the creator program in protest. The episode has intensified calls for accountability in #Roblox #ChildSafety, with journalist @Chris_Hansen exploring the issue and potential documentary involvement with @Schlep.
8. Roblox Is Not Doing Enough To Protect Kids, Lawmaker Says
A lawmaker has criticized Roblox for inadequate protection of children on its platform, citing ongoing concerns about inappropriate content and insufficient moderation. Reports indicate that harmful material, including sexual content and exploitation, remains easily accessible to minors despite Roblox’s safety measures. This raises questions about the effectiveness of Roblox’s policies and technology in ensuring a safe environment for its predominantly young user base. The criticism underscores a broader issue within the online gaming industry regarding child safety and the implementation of robust safeguards. Addressing these problems is essential for Roblox to reinforce trust and comply with regulatory expectations.
9. Google and Roblox Defeat Videogame Addiction Lawsuit: Courtright v. Epic Games
A recent court decision dismissed a lawsuit against Google and Roblox that accused them of causing videogame addiction, offering a significant legal precedent. The plaintiffs alleged that the platforms exploited psychological techniques to foster addictive behaviors, leading to harm among users. However, the court found that such claims did not sufficiently establish legal liability, emphasizing the protections afforded to platforms under current law. This outcome illustrates judicial reluctance to hold tech companies accountable for user addiction without clearer statutory mandates or evidence. It underscores ongoing challenges in regulating #videogame addiction and informs future #litigation involving digital platform responsibilities.
@Trump is weighing steep new duties on steel, semiconductors, and possibly pharmaceuticals, with tariffs on semiconductors rumored to reach up to 200% to 300% and a decision expected within about two weeks as a #Section232 review moves toward completion. Bloomberg quotes him saying, ‘I am going to have a rate that is going to be 200%, 300%’, underscoring the scale of the potential move. Such warnings inject uncertainty into the #supplychain and may push chipmakers, electronics brands, and AI firms to alter plans or lean toward U.S. production, even if tariffs never arrive, and could influence prices for items like @NVIDIA GeForce RTX 5090. The policy is tied to a #Section232 investigation that allows the #CommerceDepartment to adjust tariffs at the President’s discretion, with the investigation underway since April 2025 and no clear verdict timeline. The article highlights how such a threat can ripple through markets and consumer electronics costs while leaving firms uncertain about the eventual policy.
11. Analyzing AI’s Impact on Job Growth
Artificial intelligence (#AI) is expected to be both a creator and displacer of jobs, but overall, it could boost job growth significantly. Data indicates that AI adoption in industries accelerates productivity and opens up new roles, especially in tech, healthcare, and finance sectors. However, the distribution of job gains may be uneven, with some lower-skilled positions declining while higher-skilled roles expand. @JPMorgan highlights that proactive workforce training and education are crucial to enabling labor market shifts and maximizing AI’s positive impact. This dynamic implies that policies fostering skill development will be essential to harness AI’s benefits while managing employment transitions effectively.
New graduates are finding the first rung of the corporate ladder increasingly hard to reach as entry-level roles fade, internships shrink to fewer permanent openings, and employers expect rookies to arrive with skills. @BeatriceNolan notes that #AI is accelerating this shift by automating junior tasks and encouraging firms to delay or reduce early-career hiring. In tech, hiring for new graduates in the 15 largest companies has fallen by more than 50% since 2019, and the share of new grads in Big Tech hires has dropped from 15% to 7%. @KennethKang, a computer science graduate, applied to more than 2,500 jobs and had only 10 interviews, illustrating the daunting reality many face. The trend risks weakening the leadership pipeline in the years ahead as short-term savings from AI come at the cost of developing internal talent, a challenge to the traditional training pipeline #AI #entrylevel #leadershippipeline.
13. Class-action suit claims Otter AI secretly records private work conversations
A federal lawsuit accuses @Otter.ai of secretly recording private conversations to train its AI transcription service, Otter Notebook, without consent. The suit says Otter Notebook records meetings on @Zoom, @GoogleMeet, and @MicrosoftTeams by default and does not alert participants that recordings are shared with @Otter.ai to improve its systems, while the privacy policy says users can grant explicit training permission by checking a consent box. The case argues these practices may violate state and federal privacy and wiretap laws, and Otter has grown to about 25 million users and processed more than 1 billion meetings since 2016, with reports of confidential details appearing in transcripts and at times harming deals. Filed in the U.S. District Court for the Northern District of California, the suit seeks class-action status for California users and underscores the broader scrutiny of workplace #privacy and AI #transcription tools.
14. Highly anticipated Facebook settlement checks will soon be sent to millions of users
Millions of Facebook users are set to receive settlement checks soon following a legal agreement resolving a privacy lawsuit. The settlement stems from allegations that Facebook improperly tracked users’ online activity without consent. This development highlights growing concerns around #dataprivacy and corporate accountability in the tech industry. The distribution of these checks offers a tangible remedy to affected users and signals increased scrutiny for social media platforms. The case underscores the need for stronger safeguards and transparency in how user information is handled.
15. Appfigures: ChatGPT app sees a 673% increase in global revenue – 9to5Mac
The @ChatGPT mobile app leads revenue per download among top #AI assistant apps, with a 673% year-over-year surge in 2025 and $2 billion in global consumer spending since its May 2023 launch. It has been downloaded 318 million times this year, 2.8 times more than the same period last year, with India accounting for 13.7% of lifetime installs and the US for 10.3%, while the US accounts for nearly 40% of total revenue and the average spend per download is about $10. In lifetime revenue per install, @ChatGPT averages $2.91, well ahead of #Claude at $2.55, #Grok at $0.75, and #Copilot at $0.27, reflecting superior monetization. Grok’s lower long-term revenue is linked to its later mobile release, with xAI launching in November 2023 and Grok’s standalone iOS app arriving in January 2025 followed by Android in March 2025. Through 2025, @ChatGPT has earned $1.35 billion in mobile revenue, 53 times more than Grok’s $25.6 million, as reported by @TechCrunch, illustrating how market leadership in mobile #AI apps goes beyond installs to sustained spending.
16. Sam Altman says ‘yes,’ AI is in a bubble
The AI market may be in a bubble, @Sam Altman says investors are overexcited about AI and compares the current hype to the dot-com era. He acknowledges there is a kernel of truth in the tech boom, noting that while tech was genuinely transformative, people got carried away with valuations. As evidence, Altman criticizes startups with only a few people and high valuations, pointing to ventures like Safe Superintelligence led by @IlyaSutskever and Thinking Machines headed by @MiraMurati that have raised billions. Nevertheless, he argues that some losses are inevitable but others will reap substantial gains, and that, on balance, the bubble could deliver a net positive for the economy. He also predicts that @OpenAI will weather the burst and eventually spend trillions on data center construction, a move that may invite economist hand-wringing but signals the enduring #AI infrastructure push.
Nearly half of the Starlink satellite terminals provided by the U.S. Agency for International Development (USAID) to Ukraine were found in regions occupied by Russian forces, according to the USAID Office of Inspector General. The audit revealed gaps in USAID’s tracking and control measures for these #Starlink devices, which are critical for maintaining communication in Ukraine amid the conflict. The terminals’ presence in hostile areas raises concerns about potential misuse or loss of sensitive technology. This situation highlights challenges in the logistics and oversight of advanced technology aid in active conflict zones. Improved monitoring and stricter protocols are necessary to ensure such resources serve their intended purpose in supporting Ukraine’s defense.
18. Someone Stuck An RTX 4090 Into A Gaming Handheld
An @Nvidia GeForce RTX 4090 laptop GPU and an Intel Core i9-149000HX CPU power a custom handheld built around a 12.5-inch 4K IPS display. It runs Windows 11 and packs 64GB of DDR5-5600 RAM with 2TB of storage in a light green shell that holds the usual thumbsticks, buttons, ports, and a cooling vent. In demonstrations, the handheld runs @HorizonForbiddenWest, @GodOfWar, and @Cyberpunk2077 at 4K, staying around 72ºC while powered externally. Compared with commercial handhelds, such as the Acer Nitro Blaze with a 10.95-inch display, this setup shows desktop-class PC hardware can be portably repurposed but needs external power for peak performance. This example highlights ongoing #portablegaming trends toward PC-like power in mobile form factors, while noting the tradeoffs in size, heat, and power.
19. SpaceX reveals why the last two Starships failed as another launch draws near
SpaceX moves toward Starship Flight 10 after FAA launch approval, signaling renewed momentum for @SpaceX’s massive Starship #Starship. Engineers completed a spin-prime propulsion test, ground crews inspected engines and heat shield, and preparations continue for a no-earlier-than Sun, Aug 24 launch at 6:30 pm local Texas time (23:30 UTC) from Starbase, Texas, with a Super Heavy booster and 33 methane-fueled #Raptor engines. The FAA closed its investigation into the May in-flight failure, citing a probable fuel-component failure and corrective actions by SpaceX, while NASA’s Artemis program depends on Starship as the lunar lander #NASA #Artemis. A successful Flight 10 would help move toward demonstrations such as upper-stage reuse and in-orbit refueling, advancing SpaceX’s long-term goals for Mars settlement and the next generation #Starship and #SpaceX efforts.
20. Airbnb now lets you ‘pay later’ on vacation rentals
Airbnb has introduced a new ‘reserve now, pay later’ option for US stays, letting travelers book without paying upfront. Payments must be completed before the end of the stay’s free cancellation period so hosts have time to rebook if a guest cancels, and the option is limited to stays with moderate or flexible cancellation policies. This expands flexibility beyond the existing ‘pay part now, pay later’ method and aligns with #BNPL trends, with payments also supported by @Klarna. Airbnb notes that it will still charge upfront for travel insurance, climate contribution, or other optional add-ons. The move fits Airbnb’s broader push to offer flexible payment options for US travelers while protecting hosts.
21. Tony Robbins Faces Lawsuit Over AI Chatbot YesChat
Tony Robbins is being sued over the AI chatbot YesChat, which allegedly mimics his voice and style without authorization. The lawsuit claims that the chatbot, which allows users to interact using Robbins’ persona, violates his publicity rights and misleads consumers. The case highlights growing legal challenges surrounding the use of #AI and digital likenesses in the entertainment and personal branding sectors. This situation exemplifies rising concerns about consent and intellectual property in AI-generated content. It emphasizes the need for clearer regulations on how #AI technologies can replicate real individuals’ mannerisms and voices.
@Nandini Mullaji, raised in Mumbai where matchmaking is a strong tradition, launched Sitch, an AI-powered pay-per-setup dating service, after noting that swiping on apps often fails to bring people together. She teamed with @Chad DePue, previously at Snapchat and Microsoft, to build Sitch after working on Setup in 2023, and the product launched in December 2024 to use #AI to identify compatible matches. Unlike some apps that rely on surface-level profiles, Sitch aims to do the heavy lifting upfront, addressing issues such as too many options and weak offline follow-through, and it promises a personal matchmaker that is affordable. By July 2025, Sitch had raised $6.7M in pre-seed and seed funding and claimed tens of thousands of users, signaling investor interest in AI-driven #matchmaking that could change how singles find dates.
23. Tiny removable mini-SSD could eventually be a big deal for gaming handhelds
The introduction of tiny removable mini-SSDs offers a promising advancement for gaming handhelds, enabling expanded and easily swappable storage options. These mini-SSDs, smaller than traditional drives, allow users to quickly transfer large game libraries between devices without relying on slower cloud or internal storage solutions. This technology could address current limitations in handheld gaming, where storage capacity and speed affect performance and user experience. Manufacturers adopting these tiny mini-SSDs could provide gamers with more flexibility and convenience, enhancing portable gaming ecosystems. As demand for mobile gaming grows, removable high-speed storage like mini-SSDs will become increasingly valuable.
24. The AI Takeover of Education Is Just Getting Started
AI is reshaping learning as students who started high school after ChatGPT increasingly rely on AI to write, study, and receive feedback. Students now stitch together outputs from multiple AI models or add typos to mimic human writing, and they can upload images or documents to solve problems or improve essays. Educators are using AI to ease administrative tasks: nearly a third of K-12 teachers used the technology weekly last year, and @Sally Hubbard says AI saves her 5-10 hours weekly for creating assignments and refining curricula, while #MagicSchoolAI helps generate rubrics, worksheets, and comments. Some caution remains as AI can produce nonsensical content, and Hubbard notes the AI sometimes makes mistakes that students should learn to spot. Districts are diverging in approach, with Miami’s public schools moving from an initial ban to deploying #Gemini for high-school tutoring and role-play with historical figures.
25. Google AI Overviews linked to 25% drop in publisher referral traffic, new data shows
DCN’s new data ties Google #AI_Overviews to a decline in publisher referral traffic, with the majority of DCN members seeing Google search referrals down between 1% and 25%. Over eight weeks in May to June 2025, the median Google Search referral fell almost every week, with a YoY decline of 10% overall (7% for news brands, 14% for non-news brands) across 19 publishers (12 news, 7 non-news). Publishers attribute the declines to AI Overviews, and DCN CEO @Jason_Kint says the data provide ground truth that counters Google’s claims about ‘quality clicks’. The findings align with a Pew survey showing AI summaries reduce click-through rates, and with UK CMA activity as the PPA argues that Google’s AI features drive zero-click results and lessen citations, with AI Overviews expanding into Google Discover. Taken together, the data suggest a sustained shift in user behavior away from source links, challenging publishers’ traffic and highlighting tension between Google’s AI features and referral traffic.
@BYD has improved the fuel efficiency of its fifth-generation DM hybrid technology to 2.6 liters per 100 kilometers when the battery is depleted, a 10 percent reduction from the previous 2.9 L/100 km. The figure is measured on the #NEDC standard and certified by the Guangdong National Motor Vehicle Quality Inspection and Testing Center, with DM5.0 delivering an engine thermal efficiency of 46.06% and a combined range of 2,100 km. The improvement will be delivered to all BYD vehicles equipped with the fifth-generation DM tech via #OTA upgrades, allowing existing owners to benefit without a service visit. The move comes as BYD’s hybrid sales have declined for several consecutive months, even as NEV sales remain robust, with 344,296 NEVs sold in July (up 0.56% YoY but down 10.01% from June), PHEV at 163,143 units (down 22.61% YoY and down 4.45% from June), and BEV at 177,887 units (up 36.84% YoY, though down 14.02% from June). This update reinforces @BYD’s DM5.0 strategy and aims to strengthen its hybrid lineup amid market headwinds.
27. Embryos small but mighty, first live videos show
The first live footage of a human embryo implanting into a womb-like environment reveals the early steps of implantation. In a dish, scientists built an artificial womb using a collagen matrix and watched dozens of embryos from infertility treatments settle in and implant, recording time-lapse videos of the embryo penetrating the matrix, digging a hole, and using enzymes to loosen fibers while tiny protrusions pull the embryo deeper. The researchers were surprised by how forceful the embryo is, exerting substantial pressure compared with mouse embryos, a finding that helps explain implantation cramps and illuminates the mechanics of invasion. Although the work is in a controlled matrix rather than a human uterus, it could inform new infertility treatments and miscarriage prevention and points toward advances in #implantation #embryo #infertility #miscarriage research, noted by @Samuel Ojosnegros, @Anna Seriola, and @Amelie Godeau at IBEC.
28. The oldest known black hole formed more than 13.3 billion years ago
The oldest confirmed black hole formed within about 500 million years of the Big Bang, and @JWST observed it shaping its host galaxy more than 13 billion years ago. JWST detected elongated hydrogen emission indicating gas swirling at thousands of kilometers per second around the black hole, estimating its mass at 38 million solar masses, and a dip in light points to a dense neutral gas blanket. Computer simulations show a gas cloud can produce the observed light patterns, and the combination of thick gas with dust likely reddened the disk’s light, helping explain the ruby color of the distant object often called the little red dots. These findings shed light on early galaxy growth and active black holes in the #earlyUniverse, offering clues about how such engines fed galaxies in the universe’s infancy.
29. Giant Radar Antenna Reflector on NASA-ISRO Satellite in Full ‘Bloom’
The 39-foot (12-meter) radar antenna reflector on the #NISAR satellite has unfurled in low Earth orbit, a major milestone for the mission. The reflector, the largest ever deployed for a NASA mission, had been stowed umbrella-like until the 30-foot (9-meter) boom could be deployed and locked in place. The reflector directs microwave pulses for NISAR’s two #SAR systems, an #L-band radar that can see through clouds and forest canopy, and an #S-band system more sensitive to light vegetation and snow moisture, and it receives the return signals. By tracking ice-sheet motion, land deformation from earthquakes, volcanoes and landslides, and changes in forests and wetlands, the mission will inform disaster response, infrastructure monitoring, and agriculture decisions. NASA’s @Karen St. Germain said the deployment marks a critical step toward delivering transformative science by late fall of this year, with data designed to help global communities prepare for and recover from natural hazards.
30. Pixel 10 keynote stars Google’s latest AI innovations
Google’s Pixel 10 keynote highlighted its latest advances in AI technology, showcasing new features that enhance user experience through intelligent assistance and personalized interactions. The event presented cutting-edge developments in machine learning models and integration with #GoogleAssistant, enabling more natural conversations and proactive support. Key improvements include smarter camera capabilities using AI for better photo quality and advanced privacy features powered by on-device processing. These innovations demonstrate Google’s commitment to embedding AI across devices seamlessly while addressing user security concerns. Overall, the Pixel 10 launch reinforces Google’s strategy to lead AI-driven consumer technology, emphasizing convenience and privacy.
That’s all for today’s digest from 2025/08/16! We found and processed 30 URLs. Stay tuned for tomorrow’s collection of insights and discoveries.
Thanks Patricia Zougheib and Dr Badawi for curating the links
See you in the next one! 🚀