[ “We Choose to Ascend with AGI” ] #AGIALPHA #ASIFirst
AGI
-
AI Multi-Step Planning and Context Memory Limitations
By
–
I think it's still 5-step ahead planning that's the issue. I can see from it's reasoning it has the right idea but it took a series of steps I wouldn't. Almost amnesia-like of what it should be keeping in mind for 5/10 steps ahead.
-
Fears Growing as Super AI Arrival Accelerates Faster Than Expected
By
–
La Super IA se rapproche tellement vite, que beaucoup de mes amis pourtant technophiles prennent peur. Et vous ? Vous êtes inquiets de l’arrivée beaucoup plus rapide que prévu de la Super IA ?
-
AI tool proficiency versus planning execution capabilities
By
–
Ok so they're GREAT at tool use, they know how to implement every technique and where every icon/menu option is, but they can't plan and execute a finished thing well yet. https://t.co/l6msEx5mxA pic.twitter.com/r6e6P5Qrqi
— Varun Mayya (@waitin4agi_) 25 juillet 2025Ok so they're GREAT at tool use, they know how to implement every technique and where every icon/menu option is, but they can't plan and execute a finished thing well yet.
-
Unsolved AI Problem: Three Years of Limited Progress
By
–
It's possibly the hardest problem in AI – people have been trying to solve that for nearly three years and have made hardly any progress on it at all I've been following it pretty closely!
-
Reasoning Beyond Token Generation in AI Systems
By
–
If you think reasoning is just generating intermediate tokens you don't know what reasoning is.
-
AI Safety Warning: Leading Scientists Alert Control Loss Risk
By
–
😳👇🏾🤷🏽♀️ 40+ top #AI scientists from #OpenAI, #DeepMind, #Anthropic & Meta just dropped a major warning:
— Catherine Adenle (@CatherineAdenle) 24 juillet 2025
🚨 As AI advances, we may lose control over it. Safeguards are slipping behind capability.
📹 thedylanschmidt | IG#ArtificialIntelligence #TechNews #Tech pic.twitter.com/9dcEtllROG40+ top #AI scientists from #OpenAI, #DeepMind, #Anthropic & Meta just dropped a major warning: As AI advances, we may lose control over it. Safeguards are slipping behind capability. thedylanschmidt | IG #ArtificialIntelligence #TechNews #Tech
-
Demis Demonstrates Excellence in AI Research Leadership
By
–
Demis illustrates why he runs the best AI research lab in the world.
-
ARC-AGI v1 eval sets difficulty calibration and v2 improvements
By
–
Note that the v1 semi-private eval set is about the same level of difficulty as the v1 public eval set. The fully private eval set (used for the 2020 and 2024 Kaggle competitions) is estimated to be more difficult. ARC-AGI-2 does not have this issue — all sets were calibrated