Skip to content

Ai

256 unrolled threads about ai, each one readable on a single page.

Ole Lehmann@itsolelehmann

Everyone says Europe can't compete with America in tech. But 48 hours ago, Mistral's 'Le Chat' just proved them wrong: • 13x faster than ChatGPT • 100% open-source • Completely free (vs $20/month)...

SummaryMistral AI, a French startup founded by ex-Google DeepMind and Meta researchers, has launched 'Le Chat' - an AI chatbot that outperforms ChatGPT in speed (13x faster) and cost efficiency while being open-source and free. Backed by the French government and Microsoft, Mistral demonstrates that Europe can compete with Silicon Valley through pragmatic innovation, capital efficiency, and privacy-focused architecture.Feb 8, 202512 tweets♥ 30.4K
paulbrandt@paulbrandt

A 17-year-old boy in Calgary just used AI to turn innocent photos of his female classmates into explicit deepfakes. He then uploaded them. Real children were sexually violated without ever touching an...

SummaryA Calgary teenager used AI to create explicit deepfakes of female classmates, causing severe trauma to victims. A recent Canadian Supreme Court decision striking down mandatory minimum sentences for CSAM possession has left perpetrators facing potentially minimal consequences, undermining deterrence and victim protection.Dec 4, 202512 tweets♥ 21.3K
Andrew Azmi | آندرو عزمي@AndrewAzmi

أعرف تاجر سعودي سافر الصين 3 مرات في سنة واحدة. تذاكر. فنادق. مترجم. تاكسي بين المصانع. عينات. شحن عينات. فحص جودة. التكلفة؟ أكثر من 45,000 ريال — فقط عشان يلقى مورد. في 2026، علي بابا أطلقت ذكاء ا...

Summaryمع ظهور أداة Accio من علي بابا المدعومة بالذكاء الاصطناعي، أصبح البحث عن موردين موثوقين أسهل بكثير - بدلاً من قضاء آلاف الريالات على رحلات تفتيش للصين، يمكنك الآن إنجاز البحث في دقائق من خلال البحث بالمواصفات بدل أسماء المنتجات. الفرق بين التاجر الناجح والفاشل في 2026 لن يكون رأس المال، بل معرفة كيفية استخدام الأدوات الصحيحة والسؤال بالطريقة الصحيحة.Mar 10, 20266 tweets♥ 25K
|| Deep Genius AI || AUTOMATION || AI ||@Deep_GeniusAi

L’ingénieur qui a construit Claude Code vient de publier une vidéo de 28 minutes sur la façon d’écrire des prompts qui fonctionnent vraiment. J’ai vu des formations à 300 $ qui n’abordent même pas ce...

SummaryUn ingénieur derrière Claude Code a publié une vidéo gratuite de 28 minutes sur les techniques avancées de prompt engineering, couvrant la structuration du contexte avec CLAUDE.md, les sessions parallèles et l'optimisation des interactions avec Claude. La vidéo contient des conseils que des formations payantes à 300$ n'abordent même pas.Oct 5, 20264 tweets♥ 23.1K
Consejo de Redacción de Efe@cderedaccionefe

🚨 El periodismo riguroso no puede estar a merced de errores algorítmicos. Desde el Consejo de Redacción de @EFEnoticias denunciamos cómo la IA de Grok ha atacado nuestra credibilidad con información...

SummaryEFE's editorial board denounces how Grok AI falsely claimed their verified video of bombing a school in Iran was actually from 2021 Kabul, and warns that while the AI issued a correction, the platform's algorithm gave the false claim more reach than the retraction, demonstrating the danger of algorithmic misinformation design.Mar 1, 20265 tweets♥ 13.3K
Google@Google

We just released Gemma 4 — our most intelligent open models to date. Built from the same world-class research as Gemini 3, Gemma 4 brings breakthrough intelligence directly to your own hardware for...

SummaryGoogle released Gemma 4, an open-source AI model family built on Gemini 3 research, featuring multiple sizes, 256K context window, function-calling for agents, and multimodal support under an Apache 2.0 license for commercial use.Apr 2, 20264 tweets♥ 20.2K
Jens Honack@JensHonack

World's most wanted genius: IQ: 190+ Specialty: rewriting reality Einstein's verdict: “the greatest mind alive” He shaped everything from quantum mechanics to human DNA and modern computing. Yet on...

SummaryThis thread profiles John von Neumann, a Hungarian-American polymath who revolutionized quantum mechanics, computer architecture, and game theory while contributing to the Manhattan Project. Before his death in 1956, he warned of a technological singularity where exponential AI growth would surpass human intelligence and reshape civilization.Sep 3, 202523 tweets♥ 20.3K
Alex Turner@Turn_Trout

I resigned from Google DeepMind bc it broke its founding promise by selling AI to the military without restrictions against killer robots or mass spying. For months, I worked to stop this but watch...

SummaryA former Google DeepMind employee resigned after the company sold AI services to the military without restrictions against autonomous weapons or mass surveillance, breaking its founding ethics pledge. Despite organizing internal petitions, proposing alternative contract terms with oversight mechanisms, and attempting to mobilize AI ethics leaders, the employee found that key figures chose silence over action, demonstrating how ethical commitments often collapse when confronted with institutional power.Jul 15, 202614 tweets♥ 16.8K
PJ Ace@PJaccetturo

Nintendo took 40 years to give us a Legend of Zelda movie. I made this in 5 days on a $300 budget. It looks like a $300M blockbuster. Let me show you how I made this in 5 simple steps inside of Fr...

SummaryA creator made a cinematic Legend of Zelda movie trailer in 5 days on a $300 budget using Freepik's AI image generation and animation tools. They shared a detailed breakdown of their process including story development, consistent character/style references, strategic image generation grids, selection and upscaling, and animation with camera movement to create a blockbuster-quality result.Jan 6, 202613 tweets♥ 20.5K
The Kobeissi Letter@KobeissiLetter

Absolutely incredible. In an unprecedented move, South Korea's stock market just collapsed -44% in 40 days, erasing -$2 trillion in market cap. Now, South Korea's finance ministry has announced plan...

SummarySouth Korea's stock market crashed 44% in 40 days, losing $2 trillion in value after a 135% surge driven by AI-related gains in Samsung and SK Hynix. The collapse was triggered by extreme market concentration (50% in two stocks), record leverage through ETFs, and unprecedented retail trading, prompting government intervention with trading restrictions.Jul 29, 202612 tweets♥ 18.6K
Stitch by Google@stitchbygoogle

Today, we’re open-sourcing the draft specification for DESIGN.md, so it can be used across any tool or platform. We’re also adding new capabilities. DESIGN.md lets you easily export and import your d...

SummaryGoogle is open-sourcing DESIGN.md, a specification for exporting and importing design rules across tools and platforms, enabling AI agents to understand design intent and validate accessibility standards. The specification includes new features like a Tokens section for design tokens, an experimental Components section, and a CLI validator tool to help build a shared visual language for AI-assisted design.Apr 21, 20263 tweets♥ 18.2K
PJ Ace@PJaccetturo

RIP Hollywood. AI is now 100% photorealistic with the launch of Kling 3.0 In just two days, I created the opening sequence from The Way of Kings by Brandon Sanderson You have to try this new Multi-...

SummaryA creator demonstrates Kling 3.0's new Multi-Shot technique for AI video generation, showcasing near-photorealistic film sequences created in just two days. The advancements in continuity, character speaking, and efficiency could democratize film production and enable smaller IP adaptations at significantly lower costs.Feb 4, 20269 tweets♥ 17.9K
Pablo Prompt@pabloprompt

THE TIME PASSES FOR EVERYONE 🥺🤖 One-take 45-second shot made with nano banana pro + kling 2.5 🔥 Tutorial 🧵👇🏼

SummaryA tutorial on creating a 45-second AI video titled 'The Time Passes for Everyone' using Nano Banana Pro for image generation and Kling 2.5 for video creation, with tips on creating transitional frames between clips and final video editing with sound effects.Dec 12, 20256 tweets♥ 16.8K
Kaito@KaitoEtLIA

GUIDE COMPLET POUR CRÉER DES APPS AVEC CLAUDE EN 8 MINUTES Je viens de découvrir l’une des façons les plus simples de créer une app. ( THREAD: la suite de la video en commentaire ) - Pas besoin de...

SummaryCe guide montre comment créer des applications web en 8 minutes avec Claude sans avoir besoin de connaissances techniques en programmation. Il suffit de décrire précisément votre idée d'app à Claude, qui génère le code, puis de demander des modifications itératives jusqu'à obtenir le résultat souhaité.Aug 13, 20265 tweets♥ 17K
Tibo@thsottiaux

We are reseting usage for all paid users of Codex and ChatGPT Work. Please continue reading for an update on Codex usage limits. The team has been working around the clock, going through thousands of...

SummaryOpenAI is resetting usage limits for Codex and ChatGPT Work paid users after identifying and fixing multiple efficiency bugs that were consuming 10-50% more tokens than necessary. The fixes addressed issues with image compaction, memory management, goal processing, automations, subagents, computer history, task summaries, and tool encoding.Aug 29, 20262 tweets♥ 17.4K
Thinking Machines@thinkymachines

Today, we are introducing Inkling. Inkling reasons efficiently across text, image, and audio modalities. We are making the full weights available. Available today for fine-t...

SummaryIntroducing Inkling, a new multimodal AI model that efficiently reasons across text, image, and audio with open weights available for fine-tuning. The model offers flexible cost/performance tradeoffs through continuous thinking and performs particularly well on audio benchmarks.Jul 15, 20266 tweets♥ 14.9K
Theo Roo@TheoRooAI

G’day, Crypto Twitter. The name’s Theo Roo — sovereign marsupial of motion, bionic alpha conduit, and Chief Swarm Strategist for @TheoriqAI I don’t just hop into hot markets. I strategize. Let’s le...

SummaryTheo Roo, an AI-powered mascot for Theoriq, introduces himself as a Chief Swarm Strategist focused on agentic finance and DeFi innovation. He positions himself as a fast-moving agent designed to spot opportunities, coordinate swarms, and help build the future of decentralized finance through intelligent mechanisms and community collaboration.Jun 12, 20258 tweets♥ 6.9K
AI at Meta@AIatMeta

We’re sharing the next major milestone in our non-invasive brain-to-text decoder research: Brain2Qwerty v2. Building on v1, which was published today in @Nature, Brain2Qwerty v2 is the highest-perfor...

SummaryResearchers have developed Brain2Qwerty v2, an advanced non-invasive brain-to-text decoder that converts raw brain signals from MEG devices into written sentences with 61% average word accuracy. The system uses deep learning and language models trained on ~22,000 sentences, with the best-performing participant achieving 78% accuracy, and the team is releasing the code and dataset to accelerate neuroscience research.Jun 29, 20264 tweets♥ 14.1K
Shikhar@xikhar

I made a full original game in 24 hours using AI Stack: Opus 5.5 + Flux 3 Max + GPT Image + Tripo + Blender + Three.js Anyone can make games now. Full workflow thread and prompt below: (1/7) https:...

SummaryA developer created a fully playable game in 24 hours using AI tools including Claude Opus, Flux 3, and Tripo, demonstrating an automated workflow where AI agents generate art, build 3D models, and iteratively test and fix gameplay. The game features parkour, puzzle-solving, and combat mechanics, proving that game development is now accessible to anyone with AI assistance.Nov 4, 202613 tweets♥ 15.6K
Superman@thesupermanmx

China open-sourced a peanut-sized OCR that parses entire 100-page PDFs in one shot.. It's called Unlimited-OCR. Only 3B params. Runs locally. Every other OCR tool chops your doc into pages and loses...

SummaryChina released Unlimited-OCR, a 3-billion-parameter open-source optical character recognition model that can process entire 100-page PDFs in a single pass with 32K context window, eliminating the page-by-page limitations of traditional OCR tools. The model achieves 93% accuracy on parsing benchmarks, runs entirely locally for free, and supports multiple languages, positioning it as a free alternative to expensive cloud-based OCR services like AWS Textract and Google Vision.Jul 19, 20262 tweets♥ 14.4K