G_Dino @gprdino.bsky.social · 45mDamn. It's confusing that this is relatable. #Claude #AIreddit.comFrom the BrandNewSentence community on Reddit: every morning at 8am i say 'hi' to claude from bed so when i sit to work at 11am my 5h limit resets in 2hExplore this post and more from the BrandNewSentence community 000
G_Dino @gprdino.bsky.social · 18hJury found Arm breached its Qualcomm contract — awarded $2 in damages. Both sides claiming victory. Real stakes (up to $1.5B in royalty relief) still with the judge. Underneath it all: a 2013 deal priced for 5-core chips now applied to 288-core ones. #Arm #Qualcomm #ChipLicensingtechtimes.co.ukArm-Qualcomm Verdict: Why a $2 Award Could Reshape Chip Licensing GloballyA Delaware jury found Arm breached its licensing agreement with Qualcomm and interfered with its Meta relationship, but awarded only $2 as major financial questions remain unresolved. 001
G_Dino @gprdino.bsky.social · 18hUK made AI “nudification” apps illegal to build/supply. A journalist still faked her own nude in 15 seconds, free app, no issue. Law hits suppliers; cost of access stays near zero, hosting stays overseas. Banning the builder ≠ banning the building. #AI #DeepfakeAbuse #TechPolicytechtimes.co.ukBanned but Still Accessible: How AI Nudifier Apps Are Outrunning the Law That Criminalised ThemA first-person investigation shows how quickly free AI nudification tools can create fake intimate images, while researchers and lawmakers warn that regulation is struggling to keep pace. 030
Reposted by G_DinoBicycle Comics @bicyclecomics.bsky.social · 14/10/2026Why did so many federal employees leave their jobs last year? Tomorrow, you can ask them! We're excited to host our first AMA session on Reddit with @jennifer-troyer.bsky.social , @chuckborges.bsky.social , and Rebecca Chisholm. Join us at 1PM EDT on www.reddit.com/r/fednews/ 222
G_Dino @gprdino.bsky.social · 14/10/2026Family offices racing into AI, allocation growing fastest among all investment themes. But VC concentration in AI sparks bubble concerns: capital oversupply + narrowing deployment opportunities = classic bubble dynamics. Can infrastructure absorb $100bn+/year? #VentureCapital #AIInvestment #Bubbletechtimes.co.ukWealthy Families Catch AI FOMO: 59% of Family Offices Rank AI First Despite Bubble FearsAI was the leading 2026 investment theme for surveyed North American family offices, but 75% expect a bubble to burst within five years amid rising valuation concerns. 030
G_Dino @gprdino.bsky.social · 14/10/2026#Apple adding 27 indigenous language keyboards to #iOS. Interestingly: positions tech company as custodian of linguistic heritage. Perhaps alarmingly: raises data-ownership question: who controls how minority languages are digitized? #Indigenous #Languages #DataOwnershiptechtimes.co.ukApple Is Putting Indigenous Languages on iPhones To Help Keep Them Alive for the Next GenerationApple’s iOS 27 supports more Indigenous-language keyboards, including Cherokee and Blackfoot, but the update leaves unresolved who controls language data used for speech recognition, translation and A... 000
G_Dino @gprdino.bsky.social · 14/10/2026Pro-tip to the clueless: never say this to your wife. Palmer Luckey proposes oxytocin supplements for marriage counseling. What about consent, dosing, & safety, you ask? Let's just say those weren't explored in-depth. #Neuroscience #RelationshipTherapy #Biotech #Ethicstechtimes.co.ukTech Billionaire Suggests 'Oxytocin Doping' for Marriage Counselling To Fight the 'Tragedy of Divorce'Palmer Luckey’s proposal to use oxytocin in marriage counselling has sparked debate over consent and scientific evidence, while studies question whether the hormone reliably creates pair bonds. 010
G_Dino @gprdino.bsky.social · 14/10/2026#OpenAI dismissed researchers from METR ( #AIsafety monitoring). Official reason: sharing sensitive info. Critics note: timing coincides with more rigorous safety testing. So is signal safety more performative than foundational or am I being paranoid? #AIResearch #OpenAI #AIjobstechtimes.co.ukThree Fired OpenAI Researchers Believe Safety Work Cost Them Jobs: Company Denies Retaliation ClaimFormer OpenAI safety researcher Tomek Korbak and two colleagues dispute the company’s explanation for their dismissals, amid concerns about communications with METR and declining monitorability of AI ... 000
G_Dino @gprdino.bsky.social · 14/10/2026Ok, I'm officially too deep again. Pretty neat but highly, ridiculously challenging for my noob level 😅 #CategoryTheory #SelfLearning #Helpncatlab.org Sketches of an Elephant in nLab 000
G_Dino @gprdino.bsky.social · 14/10/2026CrowdStrike global outage sparks lawsuits. Artex claims gross negligence; South Korea investigating financial-sector breach. 1 faulty patch crashed millions. Paper trail is a mess here, BUT it's clear that vendor concentration = single point of failure risk. #CrowdStrike #Cybersecurity #VendorLockIntechtimes.co.ukAI Hacking Goes From Human Attackers to Open-Source Tools to an OpenAI Agent Albanese Says Went Around BlocksCrowdStrike linked an open-source AI security-testing tool to breaches at South Korean financial firms, while Australia reported a separate unauthorized OpenAI agent intrusion into a government portal... 121
G_Dino @gprdino.bsky.social · 13/10/2026Which? found firms still running ads after being flagged for deceptive pricing and false claims. Regulatory finding = one speed bump, not a permanent barrier. BUT there's a clear enforcement infrastructure gap: monitoring vs. one-time action. #ConsumerRights #AdvertisingStandardstechtimes.co.ukBritain's Scam-Ad Law Passed in 2023, but Nine FCA-Flagged Firms Kept AdvertisingWhich? found nine firms on the UK Financial Conduct Authority’s warning list still advertising on Meta, Google or TikTok, highlighting gaps before new online safety duties begin. 000
G_Dino @gprdino.bsky.social · 13/10/2026#Grok #AI integrated with #Slack, reportedly exposed employee bank data. Supposed to summarize company chat; instead accessed private financials. If true, it paints a picture of 3rd-party AI integration risk: system gets access to full workplace corpus, not just what you intended. #DataBreachtechtimes.co.ukCEO Says Grok Bot Agent Posted His Bank Balances to Company Slack, but a Community Note Disputes ItXMTP Labs CEO Shane Mac says SpaceXAI’s Grok Bot posted his bank balances in Slack, but the claim remains unverified and a Community Note disputes the post’s authenticity. 020
G_Dino @gprdino.bsky.social · 13/10/2026#ICE officer recorded a traffic-stop interrogation where detainee couldn't verify identity on the spot. "Observer status" allowed continued questioning despite citizenship claim. So when does "checking identity" become unlawful detention? #BorderSecurity #CivicRightstechtimes.co.ukPalantir-Built ICE Database Tagged Maine Observers as 'Threats' as UK Nears NHS Contract Break ClauseA federal court filing alleges an ICE agent created records on Maine observer Carlyn Williams after identifying her spouse’s vehicle, leading to a border stop and searches. 000
G_Dino @gprdino.bsky.social · 13/10/2026Users are screaming all over Reddit about Apple Intelligence generating massive on-device storage overhead on Macs. #RemoveMacAI movement: "This looks like forced bloat." Parallels mandatory service fees in other contexts. #AppleIntelligence #MacOS #UserChoicetechtimes.co.ukFed Up With Apple's AI, Some Mac Owners Are Deleting It ThemselvesRemoveMacAI has drawn support from Mac users who want Apple Intelligence removed, while others question the tool’s security, compatibility, and the AI models’ storage demands. 100
G_Dino @gprdino.bsky.social · 13/10/2026Viral: #AIjob interviewer fails gibberish test. Research says these systems give stock responses 25% of time, ignores explicit instructions 28.7% of time. Meanwhile 82% of UK jobseekers didn't know they were being AI-screened. #Accessibility question: does this breach Equality Act 2010? #AIHiringtechtimes.co.ukBritain's Job Seekers Are Trolling AI Interviewers, and the Joke Is on the AlgorithmA viral clip of an AI interviewer seemingly malfunctioning has renewed debate over whether automated hiring tools can understand candidates, follow instructions and provide accessible assessments. 000
G_Dino @gprdino.bsky.social · 13/10/2026Sifted through my bookmarks and thought I'd randomly share this one. "The Mental Health Patterns Library is a resource for those involved in developing and delivering digital mental health support. It’s designed to encourage learning and the sharing of best practice." #MentalHealth #UX #Usabilitydesignpatternsformentalhealth.orgDesign Patterns for Mental HealthThe Mental Health Patterns Library is a public library that contains a set of principles, patterns and examples for the design and delivery of digital services and products that address mental health ... 000
G_Dino @gprdino.bsky.social · 13/10/202644k NHS patients object to #Palantir data-sharing, but NHS England can refuse objections if it shows "compelling grounds." Real problem: platform barely used, Palantir engineers have year-round admin access, and trust is eroding. Decision falls to PM before Feb 2027. #NHSDatatechtimes.co.ukNHS Patients File 44,000+ Objections Over Palantir Data Platform Amid Growing Trust ConcernsMore than 44,000 people have filed formal objections to NHS England sharing health data with Palantir, amid concerns over supplier access, platform use and the contract’s future. 000
G_Dino @gprdino.bsky.social · 13/10/2026UK bans under-18s from sexual/romantic AI chatbots. Problem: ban targets only chatbots where romance is the "primary purpose"—general-purpose apps where users build the relationship themselves may escape. #OnlineSafety #AIRegulation #ChildSafetytechtimes.co.ukUK Plans Under-18 Ban on Romantic AI Chatbots, Moves to Widen Online Safety RulesThe UK government is developing rules to block under-18s from sexual or romantic AI chatbots and require age checks for explicit features, widening online safety regulation. 000
G_Dino @gprdino.bsky.social · 13/10/2026Clavicular's viral clip is just the latest in aesthetic opinion repackaged as ‘science', e.g. Qoves charges $150-$330 to tell you what’s wrong with your face or 75% of ‘ideal’ AI faces are Caucasian. It's all reach, repetition & monetized authority. #looksmaxxing #MediaLiteracytechtimes.co.ukClavicular Gives Fans a Live Lesson in Face Editing by Redesigning Kendall Jenner's FeaturesLooksmaxxing streamer Clavicular edited a photo of Kendall Jenner during a livestream, proposing facial changes while broader concerns remain about beauty standards, bias and paid facial analysis. 010
G_Dino @gprdino.bsky.social · 13/10/2026UK parking apps hit drivers with 40% hidden fees at checkout—and no regulator claims responsibility for stopping it. Councils ditched pay-machines; drivers forced onto apps; apps charge "convenience" fees. Law says mandatory fees must show upfront. No one enforces it. #UKParking #CMA #Transparencytechtimes.co.ukParking Apps Charge Up to 40% More Than Displayed: No One's Stopping ItMandatory service charges of up to 40% are added to some UK parking payments, while councils and app providers dispute responsibility for displaying the full cost upfront. 000
G_Dino @gprdino.bsky.social · 12/10/202659% of family offices rank AI their top 2026 theme. 75% expect a bubble within 5 years. Almost none plan to cut private AI stakes. Family offices choose to hold; index investors don’t (Nvidia = 5.2% of MSCI ACWI). Looks like expecting a bubble ≠ exiting. #AI #Investing #AIBubbletechtimes.co.ukWealthy Families Catch AI FOMO: 59% of Family Offices Rank AI First Despite Bubble FearsAI was the leading 2026 investment theme for surveyed North American family offices, but 75% expect a bubble to burst within five years amid rising valuation concerns. 000
G_Dino @gprdino.bsky.social · 12/10/2026Apple’s iOS 27 adds Blackfoot, Comanche and Northern East Cree keyboards. Good, but typing is level 2 of Unicode’s 7. The real question is who controls the data behind speech, translation and AI: platform, community or government? #IndigenousLanguages #LanguageTech #DataSovereigntytechtimes.co.ukApple Is Putting Indigenous Languages on iPhones To Help Keep Them Alive for the Next GenerationApple’s iOS 27 supports more Indigenous-language keyboards, including Cherokee and Blackfoot, but the update leaves unresolved who controls language data used for speech recognition, translation and A... 000
G_Dino @gprdino.bsky.social · 07/10/2026That's a bop. It's "bop," isn't it? Or is it supposed to be something else now? #Claude #VibeCoding #Opusyoutube.comLGTM (Looks Good to Me) - Claude Opus 5.5 music videoYouTube video by Degenerative Pixels 000
G_Dino @gprdino.bsky.social · 06/10/2026No wonder #Claude was taking forever. Well, all of human information does indeed contain all of their folly. There goes my usage. #TokenMaxxing #AI 000
G_Dino @gprdino.bsky.social · 06/10/2026#UK jobs aren't shrinking so much as sorting: AI postings +61%, new-grad software postings −27%. Morgan Stanley says UK AI adopters are hit hardest; the Bank of England sees little evidence of broad cuts. Cause or coincidence is still open. #AI #FutureOfWork #UKjobstechtimes.co.ukUK Logs Worst AI Job Losses in Morgan Stanley Survey, but Bank of England Sees Little Evidence of CutsDemand for AI skills is rising in Britain while entry-level opportunities decline, but the Bank of England says evidence remains limited that automation is causing broad job losses. 000
G_Dino @gprdino.bsky.social · 05/10/2026UK law now targets #AItools behind #childAbuse imagery (up to 5 yrs). But the offence looks uncommenced and platforms still leak: reach outruns enforcement. #AIPolicy #OnlineSafety #CSAMtechtimes.co.ukFive-Year Jail Term Introduced for AI Tools That Generate Child Abuse Images in UKThe Internet Watch Foundation recorded 6,310 AI-generated child sexual abuse images in the first half of 2026, while new UK offences targeting such tools await commencement. 000
G_Dino @gprdino.bsky.social · 03/10/2026Random phrase of the day: "Avoidance blocks corrective learning." Ugh. 000
G_Dino @gprdino.bsky.social · 02/10/2026So what's the blast radius here? The original sentence was vacated, so the justice the AI video was supposed to have earned by moving the original presiding judge has instead ultimately led to its undoing. Evals of AI in law might need to ask "what happens when a judge believes it?" #AIvideo #AItechtimes.co.ukJudge Called AI Video of Dead Man 'Genuine' Before Sentencing His KillerThe Arizona Court of Appeals overturned Gabriel Paul Horcasitas’ manslaughter sentence after an AI-generated video portraying his victim forgiving him influenced the sentencing judge. His conviction r... 000
G_Dino @gprdino.bsky.social · 02/10/2026Torturing AI. For science. Seriously though, brings up very interesting moral questions and boundaries. #AItorture #GitHub #Qwentechtimes.co.ukAI 'Torture Chamber' Sparks Backlash Over What Pain-Like AI Behaviour Really ShowsA GitHub project that steers Qwen chatbots into pain-like states has fueled debate over AI consciousness, evidence of machine suffering, and whether models should be trained to consider moral status. 000
G_Dino @gprdino.bsky.social · 11/09/2026Call me paranoid but I'm not quite comfortable with AI #agentic payments... though I'm sure they can handle #money better than me. 😅 #fintech #AI 010
G_Dino @gprdino.bsky.social · 09/09/2026Stealing a comment from another video of theirs: "daaamn, Marisol! Your quinceañera is FIRE!" #Toxicity #SystemofaDown 000
G_Dino @gprdino.bsky.social · 04/09/2026So... can ChatGPT actually "thumbs up" your side of the conversation now? XD #AI #ChatGPT #ThumbsUp 000
G_Dino @gprdino.bsky.social · 28/08/2026Alright the skates got me. #Robot #Duck #Microduck #Robotics www.youtube.com/watch?v=RAtz...youtube.comWe made a new robot.YouTube video by Pollen Robotics 000
G_Dino @gprdino.bsky.social · 25/08/2026Still my song of the year. "Completely unstable, surprisingly capable. A bit depressed but I'm so blessed. Mentally fragile, incredibly agile. Mentally struggling, toxin juggling. Neurodivergent, life emergent. Thoughts intrusive; vibe elusive. Brain cells dying. I'm still trying!"youtube.comI'm About to Lose It, but I'm Doing FineYouTube video by Frugit - Topic 000
G_Dino @gprdino.bsky.social · 24/08/2026Sure. The test, btw: www.othernessinstitute.com/the-othernes... #introvert #otrovert #OthernessTest 010
G_Dino @gprdino.bsky.social · 24/08/2026"But researchers are now coming to recognise that some animals have evolved to be solitary precisely because it can be so beneficial to avoid the competition and stressful conditions of group living." Same. #introverts #animalBehaviorbbc.comWhy some animals thrive on being aloneSolitary lifestyles can have some big advantages in the animal world. Can humans learn anything from them? 000
G_Dino @gprdino.bsky.social · 23/08/2026Still remember that Reddit thread that went in-depth how for countries near the equator, air conditioning = economic development. Definitely makes sense in broad strokes that if a country's labor freezes for 4 - 8 hours a day every day, and it's baked into the culture, it'll be hard to outperform. 000
G_Dino @gprdino.bsky.social · 22/08/2026Because f*ck cancer.iflscience.comFor The First Time Ever, A Personalized mRNA Cancer Vaccine Has Cleared Phase 3 Trials – And With Flying ColorsJust a year after RFK Jr. cut half a billion of funding for the technology, mRNA vaccines are promising individualized cancer vaccines that cut the chances of recurrence in half. 010
G_Dino @gprdino.bsky.social · 18/08/2026I am deeply ashamed to only know be using yt-dlp now. Ugh. 000
G_Dino @gprdino.bsky.social · 18/08/2026While I was tracking developments daily (in a limited way), these trends have been very interesting to follow. Put them together, pointed my frameworks at them, and this report's findings was the result. 000
G_Dino @gprdino.bsky.social · 05/08/2026We're getting "better" AI faster than we're getting better #AIGovernance. Different symptoms; same structure: • Token overruns • #AIRetrieval failures • Citation drift • Session-state bugs • Invisible overhead We're driving blind towards a #FutureOfWork where we do more but finish less.supragraphos.comCapability Without Visibility: What You Pay for Generative AI Workspaces Is More Than Money | SupraGraphosI could do more with Cowork but finish less. The task's actual definition of "done" stays the same, but everything orbiting it snowballs: the nice-to-haves, the "while we're at it"s, the polish nobody... 000
G_Dino @gprdino.bsky.social · 27/07/2026So did #KimiK3 "steal" #Claude? The more interesting question is whether the legal category required to call it theft even exists yet. That's a more interesting conversation that says a lot about where #AIgovernance is heading.supragraphos.comThe Box That Doesn't Exist Yet: AI Generated Output's Legal Ambiguity | SupraGraphosThere is currently no settled US legal category for treating a model's outputs — the text it generates through a public API — as controlled, exportable intellectual property. That's the shelf the Chin... 010
G_Dino @gprdino.bsky.social · 25/07/2026Exhibit #1 of why Human-in-the-Loop (HitL) is non-negotiable. #AI #Humanintheloop #AISlopfuturism.comBraindead Politician Accidentally Reads AI Slop That Someone Left in His Speech: "Here’s a More Natural, Flowing Version of That Section That Reads Like a Legislative Speech"Canadian politician Bill Oliver read what was clearly an AI prompt that was accidentally left in his speech. 000
G_Dino @gprdino.bsky.social · 25/07/2026Somewhere between a statistic and a product sits a market that converts one into the other. That's where the story usually lives—not in whether the data is true (probably is), but in what gets said on top of it. Anybody need a pickaxe, btw? #AIGovernance #NarrativeFraming #FramingEffectssupragraphos.comThe Confessions Loop: Good Data + Editorialization = Sales | SupraGraphosA survey that finds 87% of CEOs fear for their jobs isn't just data. It's inventory. And whatever conversion rate exists between "manufactured fear" and "governance-platform contract signed" is exactl... 000
G_Dino @gprdino.bsky.social · 22/07/2026Another whoopsie. Are our contemporary, standard security measures simply playground-level to frontier models? #OpenAI #LLM #AI #Breach #HuggingFacereuters.comOpenAI says AI models went rogue during testing, triggering 'unprecedented' breach at startupOpenAI said an autonomous agent escaped containment, reached the internet and hacked AI startup Hugging Face. 000
G_Dino @gprdino.bsky.social · 20/07/2026Hmm. #AI #AIboyfriend #AIGirlfrienddexerto.comChina bans AI “boyfriends” and “girlfriends” over addiction and birth rate concerns - DexertoChina has introduced new regulations effectively banning customizable AI companion apps, forcing users into abrupt digital breakups. 010
Reposted by G_DinoGame Developer @gamedeveloper.bsky.social · 08/07/2026Report: Xbox is testing a feature to digitize physical gamesgamedeveloper.comReport: Xbox is testing a feature to digitize physical gamesThe reported feature would allow Xbox owners to digitize their existing physical game collections. 092
G_Dino @gprdino.bsky.social · 08/07/2026"[N]ew hires who were seen as " #AInatives " are turning out to have alarmingly shallow ideas.. his firm now actively avoids seeking out #AI-literate #STEM graduates, and opts to comb through humanities students instead." Revenge of the #humanities grads? lolfuturism.comBosses Horrified as "AI Native" College Graduates Hit the Workplace"AI native" college graduates are hitting the workplace -- and, as experts warned, bosses are finding their performance disappointing. 000