gpt 5.4 in codex is real smart – feels smarter and less sloppy than opus 4.6 tbh
GENERATIVE AI
-
Y Combinator Startups Ship with 95% AI-Generated Code
By
–
One in four Y Combinator startups is already shipping with 95% AI-generated code. Building at that speed is now possible without creating data silos or stepping outside your existing security controls. This tutorial walks through how to build a full-stack app inside Databricks
-
Practical AI Tools: QMD Installation and Claude Code Integration
By
–
Sean I check out lots of videos and read lots of posts on openclaw. Especially appreciated yours because of its focus on facts and getting things done vice the hype found in so many. Had tried QMD before but made mistakes on instal. Copying and pasting your text into claude code
-
Wall Street Journal reporters trick AI CEO bot to surrender power
By
–
Think your job is safe from AI? At the Wall Street Journal, the humans didn't just compete with the AI. They staged a corporate coup against it. Reporters created fake PDF board minutes to convince the "CEO" bot to strip itself of power. The takeaway? AI doesn't understand… pic.twitter.com/vuPOcRTgZp
— Lucian Fogoros (@fogoros) 14 mars 2026Think your job is safe from AI? At the Wall Street Journal, the humans didn't just compete with the AI. They staged a corporate coup against it. Reporters created fake PDF board minutes to convince the "CEO" bot to strip itself of power. The takeaway? AI doesn't understand
-
Elon Musk: xAI will catch up and exceed all in 3 years
By
–
xAI will catch up this year and then exceed them all by such a long distance in 3 years that you will need the James Webb telescope to see who is in second place
-
ShinkaEvolve Improvements: ICFP Victory, ICLR2026 Acceptance, New Features
By
–
🧑🔬 Since the first release of ShinkaEvolve, a lot has happened & we continue to improve its performance:
— Robert Lange (@RobertTLange) 14 mars 2026
1️⃣ ShinkaEvolve supported @iwiwi + team Unagi in winning the ICFP contest: https://t.co/FCI9p0snpy
2️⃣ Shinka got accepted at #ICLR2026
3️⃣ We improved Shinka's program… https://t.co/p2YhrU0eey pic.twitter.com/5VVEcSl41F🧑🔬 Since the first release of ShinkaEvolve, a lot has happened & we continue to improve its performance: 1️⃣ ShinkaEvolve supported @iwiwi + team Unagi in winning the ICFP contest: sakana.ai/icfp-2025/ 2️⃣ Shinka got accepted at #ICLR2026 3️⃣ We improved Shinka's program throughput by pushing the concurrency of sampling & evaluation 4️⃣ We made the LLM ensembling cost-performance aware, i.e., selection prioritizes cheap + improving LLMs 5️⃣ We added coding agent integration via dedicated skill files so you can leverage Shinka for autoresearch 6️⃣ Install via: uv pip install shinka-evolve Let us know about your experience and what you would like to see next 🤗 💻 Repo: github.com/SakanaAI/ShinkaEv… Sakana AI (@SakanaAILabs) “When AI Discovers the Next Transformer” Robert Lange (Sakana AI) joins Tim Scarfe (@MLStreetTalk) to discuss Shinka Evolve, a framework that combines LLMs with evolutionary algorithms to do open-ended program search. Full Video: piped.video/EInEmGaMRLc — https://nitter.net/SakanaAILabs/status/2032714767862026553#m
→ View original post on X — @_yutaroyamada, 2026-03-14 13:42 UTC
-
Critical thinking skills essential before AI era
By
–
The time to learn how to think for yourself was before genAI, if you missed your chance, good luck
-
Qwen 3.5 27B Matches Sonnet 4.5 on Single RTX 5090
By
–
That model, Qwen 3.5 27B Dense is equal to Sonnet 4.5 Runs great on a single RTX 5090 w/ full context But we are not anywhere near Opus 4.5 even with Qwen 3.5 397B MoE
-
MiniMax M2.5 Lightning Attention Architecture for Long Context Scaling
By
–
Lightning Attention architecture used in MiniMax M2.5 is really interesting. The structure is 7 Lightning Attention layers for every 1 traditional SoftMax attention layer, which lets it scale to long contexts while keeping the quality you'd expect from standard transformers. I… pic.twitter.com/HnfTF0W6f1
— Akshay 🚀 (@akshay_pachaar) 14 mars 2026Lightning Attention architecture used in MiniMax M2.5 is really interesting. The structure is 7 Lightning Attention layers for every 1 traditional SoftMax attention layer, which lets it scale to long contexts while keeping the quality you'd expect from standard transformers. I
-
Global Accelerator Phase II Bridges India European AI Startups
By
–
The Global Accelerator – Phase II is Now Open! Bigger ambitions. Stronger bridges. Global possibilities. The Global Accelerator Program – Phase II is back, creating a powerful bridge between India’s most promising AI startups and the European innovation ecosystem. Designed to