I think it would be a worse puzzle. A good LLM knows its limitations and uses tools. (Raw LLMs also can’t compute SHA1, which is why I used it.) In another reply in this thread, Gemini 2.5 Pro is seen failing at this task because it thinks the fifth letter of “Madrid” is d.
CREATIVE AI
-
3rd Annual AI Film Festival Concludes Successfully
By
–
The 3rd Annual AI Film Festival has officially come to a close. Thank you to all of our incredible filmmakers, festival partners, jurors and attendees for making this our biggest year yet.
— Runway (@runwayml) 18 juin 2025
Panels, winning films and exclusive interviews coming soon. We’ll see you next year. pic.twitter.com/iakB34bJD3The 3rd Annual AI Film Festival has officially come to a close. Thank you to all of our incredible filmmakers, festival partners, jurors and attendees for making this our biggest year yet. Panels, winning films and exclusive interviews coming soon. We’ll see you next year.
-
Google Adds TikTok Format Option to VEO 3 Video Generator
By
–
Google a très vite compris ce que fessaient les gens avec VEO 3.
— VISION IA (@vision_ia) 18 juin 2025
Ils viennent d’implémenter l'option de créer des vidéos au format tiktok directement pic.twitter.com/eq94K87QwEGoogle a très vite compris ce que fessaient les gens avec VEO 3. Ils viennent d’implémenter l'option de créer des vidéos au format tiktok directement
-
ByteDance Unveils Seedance 1.0 AI Video Generator Surpassing Google VEO 3
By
–
Bytedance vient de dévoiler Seedance 1.0, générateur vidéo par IA, concurrent direct à google VEO 3 et qui sur les benchmark le dépasse même.
— VISION IA (@vision_ia) 18 juin 2025
Bientôt, on pourra affirmer, plus rien n’est réel…
Voici des exemples hallucinants : pic.twitter.com/ylYgt9DTplBytedance vient de dévoiler Seedance 1.0, générateur vidéo par IA, concurrent direct à google VEO 3 et qui sur les benchmark le dépasse même. Bientôt, on pourra affirmer, plus rien n’est réel… Voici des exemples hallucinants :
-
MiniMax’s Hailuo 02: Powerful Video AI Ranked #2 Globally
By
–
China's MiniMax continues to cook — this time delivering a powerful video AI called Hailuo 02
— Rowan Cheung (@rowancheung) 18 juin 2025
The model currently ranks #2 on the Artificial Analysis leaderboard (ahead of Veo 3) with 1080p outputs and the ability to handle extreme physicspic.twitter.com/N7WaEh8QdSChina's MiniMax continues to cook — this time delivering a powerful video AI called Hailuo 02 The model currently ranks #2 on the Artificial Analysis leaderboard (ahead of Veo 3) with 1080p outputs and the ability to handle extreme physics
-
Adobe Firefly Mobile App Launched
By
–
Adobe released Firefly mobile app on iOS and Android which supports Firefly and third party models for image and video generation.
-
Explore random style space corners with –sref prompt parameter
By
–
Wow these are quite nice! Try using –sref random at the end of your prompt to explore random corners of "style space". The word "random" will get replaced with a number (as below) which you can repeatably use for consistency or share with others around the globe.
-
Hailuo MiniMax 2 Generates Surfing Pangolin Video Model
By
–
Nuevo modelo de vídeo, nuevo pangolín surfeando!
— Carlos Santana (@DotCSV) 17 juin 2025
Esta vez cortesía del nuevo modelo Hailuo MiniMax 2 pic.twitter.com/XhrYlhSfiHNuevo modelo de vídeo, nuevo pangolín surfeando! Esta vez cortesía del nuevo modelo Hailuo MiniMax 2
-
Krea AI lance Krea 1 gratuitement
By
–
Krea AI released Krea 1, a new image generation model which you can use for free.
— 🚨 AI News | TestingCatalog (@testingcatalog) 17 juin 2025
Image gen animations are very cool 👀 https://t.co/3gQNlkHNYd pic.twitter.com/exMaftcaFyKrea AI released Krea 1, a new image generation model which you can use for free. Image gen animations are very cool
-
10 Ways to Unlock Grok 3’s Full Potential
By
–
Grok 3 is terrifyingly powerful. But most people have no idea how to unlock its full potential. Here are 10 ways to use Grok 3 for automation, research, marketing, coding, content creation and more: