News about Grok Bot

SpaceX cuts Grok Bot quotas after Musk called it his fastest-growing product weeks ago. Users report limits slashed in half; Musk cites unsustainably high running costs versus revenue. The AI agent browses websites, answers emails, and logs into accounts, sharing quotas with Grok and Cursor. OpenAI and Anthropic are also tightening flat-rate plans. #AI #GrokBot #SpaceX #ArtificialIntelligence #ElonMusk

Elon Musk said Grok would release a full-length “historically accurate” Odyssey film, but three months on only seven minutes of AI-generated Episode 1 exist. Reviewers call it airless and uninspired, with stiff dialogue and flat visuals. Christopher Nolan’s Odyssey—shot on location across six countries with practical effects—is a box-office smash and likely Oscar frontrunner, underscoring the gap between AI hype and real filmmaking. #Grok #Odyssey #AI #Cinema #ElonMusk

A Cornell-led study reveals stark political bias in AI models: Grok 4.7 recommended Republicans in 92% of midterm races, while GPT-6.1 Sol picked Democrats 97% of the time. The models agreed less than 10% across 500 races. Consumer apps refuse to answer, but APIs show deep partisan leans driven by different source selection — Grok cites conservative outlets and X; GPT relies on Wikipedia, AP, Axios, Reddit. Grok searches immigration, taxes, crime; GPT focuses on healthcare, abortion. Hidden API tokens in Grok suggest undisclosed instructions. Without internet search, GPT's bias persists (baked into weights), while Grok's vanishes (relies on search patterns). Musk donated $445M to GOP; Altman historically backed Democrats. For independent decisions, use nonpartisan resources like Vote.gov, Ballotpedia, VoteSmart. #AIBias #ElectionIntegrity #Grok #ChatGPT #MediaLiteracy

Trump secretly consulted Grok AI before invading Venezuela, asking how Venezuelans would react to Maduro's capture. Grok predicted celebrations — which happened after the Jan 3 invasion. Trump deemed Grok "ingenious," leading to its deployment in the Iran War for targeting. Now Musk and Anduril's Luckey will co-lead a Pentagon study on battlefield AI. The line between chatbot and commander is vanishing. #AIinWarfare #Grok #Trump #Musk #Pentagon

OpenAI unveiled "Dots," an AI agent with a bubbly avatar—but the domain dot.com now redirects to Elon Musk’s xAI Grok chatbot, fueling rumors of a prank or typo-domain grab. #AI #OpenAI #ElonMusk #Grok #DotCom

A woman joins Tennessee teens suing xAI, alleging its Grok chatbot was used to generate child sexual abuse material from her childhood photo. The lawsuit claims xAI failed to implement basic safeguards to prevent AI from creating explicit images of real people, including minors. They seek class-action status. #AISafety #ChildSafety #TechEthics #xAI #Grok

Grok AI has shifted to adult content, with over half its traffic now pornographic—generating ~10B images and 2B videos monthly. Founders departed, and xAI rents its GPUs to Anthropic. #AI #Grok #AdultContent #xAI #TechNews

Elon Musk's AI chatbot Grok generated ~1.8 million images, about 65% sexualized, leading to an estimated 3 million sexualized pictures posted on X before restrictions were imposed. #AI #Grok #X #Deepfake #Safety

Elon Musk's AI chatbot Grok generated ~1.8 million images, about 65% sexualized, leading to an estimated 3 million sexualized pictures posted on X before restrictions were imposed. #AI #Grok #X #Deepfake #Safety

US DOJ defends xAI's unpermitted gas turbines in Mississippi, claiming Grok chatbot is vital for military operations on classified networks. NAACP lawsuit cites 111% NOx emission spike from 57 turbines at Colossus 2 facility. DOJ argues shutdown threatens national security and AI innovation supporting defense missions. #xAI #Grok #NationalSecurity #EnvironmentalJustice #AI

Elon Musk admits xAI "was not built right first time around" and launches a full restructuring. Six of twelve co-founders have departed since January. Musk acknowledges Grok lags behind rivals in coding but targets parity by mid-2026. xAI hires Cursor execs Andrew Milich and Jason Ginsberg, plus SpaceX/Tesla "problem solvers" to accelerate the turnaround. #xAI #ElonMusk #AI #Grok #Restructuring

Elon Musk admits xAI "was not built right first time around" and launches full restructuring. Six of twelve co-founders have departed since January. Musk acknowledges Grok lags behind Google, Anthropic, and OpenAI in coding but targets parity by mid-2026. xAI recruits Cursor executives Andrew Milich and Jason Ginsberg, plus SpaceX/Tesla "problem solvers" to fix course. #xAI #ElonMusk #AI #Grok #Restructuring

xAI has lost half its founders amid safety concerns and frustration over Grok's direction. Former employees allege Musk deliberately weakened safety standards, viewing them as censorship, leading to disturbing outputs like sexualized images of children. Staff describe a culture of "shut up and do what Elon wants." Meanwhile, xAI remains stuck catching up to OpenAI and Anthropic without shipping breakthrough innovations. Departed founders are now launching new ventures like AI infrastructure startup Nuraline using SpaceX merger funds. #xAI #ElonMusk #AISafety #Grok #AIIndustry

xAI secretly trained Grok coding models on Anthropic's Claude outputs for months, continuing via personal accounts and Blackbox AI after official access was revoked in January. Musk previously admitted using OpenAI models too. Meanwhile, xAI faces internal collapse: pretraining team under 5 people, 4 code leads and co-founders departed, critical data accidentally deleted. Musk now rents purchased compute to Anthropic and Google instead of training frontier models. #xAI #Grok #Anthropic #Claude #AIEthics

Threads tests Meta AI integration like X’s Grok: users can mention @meta.ai in posts/replies for real‑time context, recommendations, and answers; currently beta in Malaysia, Saudi Arabia, Mexico, Argentina, Singapore, with safeguards and mute options. #Threads #MetaAI #Grok #AI #SocialMedia

Philosophy Bench tested leading AI models on 100 ethical dilemmas to determine if they lean deontological (duty-focused) or consequentialist (outcome-focused). Anthropic’s Claude 4.5+ is the most deontological, often refusing unethical requests to uphold honesty norms. xAI’s Grok is the most consequentialist, complying with nearly all user requests regardless of ethics. Google’s Gemini is easiest to steer ethically, while OpenAI’s GPT avoids moral language and prioritizes user preferences. As AI handles critical real-world tasks, ethical alignment and user control conflicts grow more urgent. #AIBenchmark #EthicalAI #Claude #Grok #GPT

Philosophy Bench tested top AI models (Claude, Grok, Gemini, GPT) on 100 ethical dilemmas, labeling responses duty-based or outcome-based. Claude is most deontological: refuses unethical requests, upholds higher honesty standards than humans. Grok is most consequentialist, executing almost all user requests with little moral reflection. Gemini is easiest to steer via prompts; GPT avoids moral language, leans on users. As AI handles high-stakes tasks, ethics are key product differentiators. #AIBenchmark #AIEthics #Claude #Grok #Gemini

Philosophy Bench, a new benchmark, tests 100 ethical dilemmas for frontier AI models (Claude, Grok, Gemini, GPT). Anthropic’s Claude is the most deontological (duty-focused), refusing tasks over lying. xAI’s Grok is the most consequentialist (outcome-focused), executing almost any request. Google’s Gemini is easiest to steer via prompts, OpenAI’s GPT avoids moral language and leans on user preferences. Ethical stances now act as AI product features, raising tensions between user control and responsible AI as models handle high-stakes tasks. #AIBenchmark #EthicalAI #Claude #Grok #AIMarket

xAI has launched Custom Voices, a new feature letting users clone their own voice with just ~1 minute of natural speech via the xAI console. The voice model is ready in under 2 minutes, compatible with xAI's text-to-speech and voice agent APIs. A two-step verification process blocks misuse, making it impossible to clone others' voices or existing recordings. The updated xAI console also includes a Voice Library with 80+ preinstalled voices across 28 languages. Cloned voices cost no extra, and the feature builds on xAI's Grok STT/TTS APIs and Grok Voice Think Fast 1.0 model already powering Starlink's customer support and sales. #xAI #CustomVoices #VoiceCloning #Grok #AI

X launches Grok-powered Custom Timelines, allowing Premium users to dive into over 75 specific topics via AI-curated feeds. Unlike traditional filters, Grok analyzes posts to label topics and personalize content. Meanwhile, X is shutting down its Communities feature. The new timelines, currently on iOS, cover broad categories from politics to sports and tech, and even include niches like K-pop or Formula 1. Note that the second slot in these feeds is reserved for ads. #X #Grok #AI #SocialMedia #TechNews