I build agentic products in public. The interesting part of AI is not just what a model can say. It is what a system can do reliably. My background: - PhD in Computer Science - ex-Walmart Staff Engineer Here I write about: - agents in production - system design around LLMs - local vs frontier model tradeoffs - failure modes, retries, memory, trust If you care about practical AI systems more than demo theater, welcome.

For the past year, we designed hundreds of sections for concept projects, internal training, and trend explorations. A lot of those designs were just sitting unused in our Figma files. So we thought, why not give them away for free? Today, we’re launching Blocks by Lander Studio. A free library of design blocks, website sections, app screens, dashboard designs, logos and illustrations for designers. Copy them, learn from them, repurpose them, or use them as inspiration. Beta access is now open with website sections available for everyone to copy. Other categories are launching next. Want early access to all categories? Comment below and repost this post. I’ll send you the invite link.

AI is the coworker we never had

At this point Microsoft has more opensource projects than Apple

Today I spent over $1,000 in Claude Code API costs to build this with Fable 5. In a single day. Not on prompts. Not on experiments. On production code. I built my own Agentic Development Environment - Maverick from scratch in Rust and Tauri. Here's what it looks like:

i hooked my whoop to my work calendar to find which coworker gives me the most stress 🚨 thanks to fable, I reverse engineered whoop to pull per minute heart rate. nd matched spikes with cal events and attendees I now have a leaderboard and I think about it daily. few info masked for obvious reasons ;)

I keep hearing about /loop feature in Claude. But the most dangerous AI automation is not the one that fails. It is the one that almost works. That is why agentic workflows need: - isolated sandboxes - audit trails - approval gates - evals - rollback paths - clear ownership Autonomy without verification is just technical debt with better branding.

Exciting news for iOS 27 users Apple Intelligence in Shortcuts, can now output Structured Outputs as: Text, Number, Date, Boolean, List, and Dictionary. You no longer have to parse LLM responses manually

🚀 1,000 TOKENS/S ON A 1T MODEL! 🚀 We are thrilled to release Xiaomi MiMo-V2.5-Pro-UltraSpeed in collaboration with @TileRT_AI , breaking the 1,000 tokens/s output speed on a 1 Trillion parameter model for the FIRST TIME! Not wafer-scale integration like Cerebras. Not pure on-chip SRAM chips like Groq. We achieve 1,000 tps on a 1T MoE model using just a SINGLE, STANDARD 8-GPGPU NODE. Read the full technical deep dive：mimo.xiaomi.com/blog/mimo-ti… Want to experience the future of real-time AI? 👉 Apply for UltraSpeed now: platform.xiaomimimo.com/ultr… ⏳ Limited-Time Access: Application-based · Jun 8 – Jun 23 (PDT) 💬 Chat Experience: Completely FREE for a limited time — try the blazing-fast web chat now. ⚡ UltraSpeed API: Just 3x the price for a ~10x boost in output experience. 🤝 Enterprise & Large-Scale Needs: business-mimo@xiaomi.com

what happened to all the startups building robots all of a sudden??

India tech hiring hits 28-month low 93,000 active openings. Entry-level down 44% YoY. IT services down 31% YoY. AI is eating the low end of Indian IT faster than anyone wants to admit. The tier-1 service model of TCS, Infosys is structurally breaking.

Just installed CodexPlusPlus. An enhanced wrapper for CodexApp that makes the native experience actually usable. The default Codex UI is fine. This is better. Keyboard shortcuts. Better file tree. Session management. Small improvements. Big difference.

Ping Island is the most underrated macOS utility for AI coding right now. - Dynamic Island-style menu bar. - Approve agent actions. - Jump between sessions. - Monitor multiple agents. All from the boring notch. I did not know I needed this until I tried it. Now I cannot code without it.