Jonathan Castillo

[Jensen Huang @JensenHuang]

For my first post, I’m sharing a letter @NVIDIA signed on why open models matter.

AI will transform every industry, power every company, and be built by every country.

Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.

The world needs both frontier closed models and frontier open models.

https://t.co/AUKzoQ5Ikb

[Movez @0xMovez]

Microsoft Senior AI developer just showed how they build AI agents with Claude at Microsoft.

34-minutes. free. By Microsoft team

Opus 4.7 + 1,400+ pre-built MCP tools

plug Claude into agent → give it tools → ship to production

worth more than any $500 vibe-coding course.

[Noisy @noisyb0y1]

A GERMAN DEVELOPER REPLACED HIS ENTIRE DEV TEAM WITH KIMI K2.6, VISUALIZED EVERYTHING IN OBSIDIAN AND NOW MAKES $80,000/MONTH SOLO

1 trillion parameters, 32 billion activated per token and a SWE-Bench score of 65.8 - Kimi K2.6 reads the entire client codebase, understands the architecture, writes production code and ships for $150-300 in API costs while a traditional agency pays developers $4,800 for the exact same project.

300 parallel agents per run deliver 100+ files simultaneously - search, analysis, coding and writing all in parallel - and Obsidian visualizes the entire knowledge graph in real time while the agents work.

A traditional agency with 10-15 people keeps 30% margin after salaries. He keeps 90% - $72,000 in monthly profit with $500 in overhead.

By month 10 Kimi handles 80% of the technical work and he manages only strategy and client relationships - while Obsidian maps every project, every client and every agent in one graph that updates itself.

[Movez @0xMovez]

This 1 hour podcast with the Head of Claude Code will teach you more about vibe-coding than 100 paid courses.

Bookmark this & give it 1 hour today. It's the best video about AI you'll watch this week. Then read the article below.

[Tech with Mak @techNmak]

This free CUDA course is worth more than most CS degrees.

12 hours that separate library users from GPU engineers.

I watched senior devs struggle with concepts taught in hour 3.

What makes it different:

No hand-waving. No "just use this library."

You build an MLP trainer FOUR times: → PyTorch (the easy way) → NumPy (getting harder) → C (now we're cooking) → CUDA (chef's kiss)

Same model. Same dataset. Four implementations.

By the end, you understand WHY PyTorch is fast.

The curriculum nobody else teaches:
➡️ GPU architecture (not just "it's parallel")
➡️ Writing kernels that don't suck
➡️ Profiling at kernel AND system level
➡️ When cuBLAS helps (and when it doesn't)
➡️ CUDA vs Triton (the comparison you need)
➡️ PyTorch extensions (actually useful ones)

Real talk:
➡️ After this course, you'll read PyTorch source code and understand it.
➡️ You'll optimize models other engineers can't touch.
➡️ You'll be the person teams hire to make things fast.

Created by @elliotarledge 💪

12 hours. Free. No excuses.

Who's starting this weekend?
(I will put the details in the comments.)

[Julian Goldie SEO @JulianGoldieSEO]

Claude Sonnet 4.6 + Gemini is INSANE! 🤯

Build full apps with two prompts.

No more expensive developer fees.
No more manual business workflows.
Here’s the new play 👇

→ Claude 4.6 automates backend logic
→ Gemini 3 creates visual designs
→ Process one million tokens instantly
→ Connect APIs with zero code
→ Build lead machines on autopilot

Join the AI Profit Boardroom.
Scale while others still build manually.

Want the full guide? DM me.

[Julian Goldie SEO @JulianGoldieSEO]

N8N FULL COURSE 1 HOUR (Build & Automate Anything)

[Julian Goldie SEO @JulianGoldieSEO]

GPT-5 FULL COURSE 1 HOUR (Build & Automate Anything)

[Julian Goldie SEO @JulianGoldieSEO]

Google Gemini FULL COURSE 3 HOURS (Build & Automate Anything)

[Julian Goldie SEO @JulianGoldieSEO]

NotebookLM + Google Gemini FULL COURSE 3 HOURS (Build & Automate Anything)

[Julian Goldie SEO @JulianGoldieSEO]

Google Antigravity FULL COURSE 4 HOURS (Build & Automate Anything)

[Julian Goldie SEO @JulianGoldieSEO]

Google Gemini FULL COURSE 5 HOURS (Build & Automate Anything)

[Jaynit Makwana @JaynitMakwana]

OpenClaw: FULL COURSE 3 HOURS (Build & Automate Anything)

[AI Guides @free_ai_guides]

I stopped using ChatGPT for daily tasks.

Switched to an open-source agent that:

→ Remembers every conversation I've had
→ Knows my preferences without being told
→ Adds new skills from a community library
→ Runs 24/7 without me touching it
→ Costs almost nothing

It's called OpenClaw.

And it's what AI assistants should have been from the start.

I wrote the setup guide. Free.

Comment "OpenClaw" and I'll DM it to you

[Jonathan Castillo @JonathanCJ05]

@Motion_Viz Prompt

[Prajwal Tomar @PrajwalTomar_]

https://t.co/uJeJn5Mchp

[Dialectical Thinker @DialecticalWDYT]

UI-TARS-desktop, an open-source multimodal AI agent developed by #ByteDance (the company behind #TikTok). It allows users to automate desktop tasks using natural language commands, running 100% locally without internet access after setup.

Key Details
1. Core Capabilities: The agent captures screenshots, interprets them with vision-language models, and performs precise mouse/keyboard actions. It can open apps, navigate menus, fill forms, browse websites, and handle complex workflows, all via chat-like instructions.
2. Privacy & Local Execution: Everything processes on your machine for enhanced privacy and offline use.
3. Supported Models: Primarily ByteDance's UI-TARS-1.5-7B (available on Hugging Face), with support for related Seed-VL models.
4. Platform: Cross-platform focus, with strong demos on macOS and Windows.
5. Project Status: Launched in early 2025, actively maintained, ~25.5k GitHub stars.

Resources
GitHub Repo → https://t.co/qz1RKPfOc4 (installation guides in docs).
Model → https://t.co/YhdDnFulGu.

Caution: This tool grants full desktop control, so use it carefully (e.g., in isolated environments) to avoid unintended actions. As open-source code from a major company, it can be audited, but powerful automation always carries risks.

[Jonathan Castillo @JonathanCJ05]

@PrajwalTomar_ @Lovable DECK

[Prajwal Tomar @PrajwalTomar_]

I tried using @Lovable to create slides and it ONE-SHOTTED the entire deck for my SaaS.

Took my raw notes, structured the outline, added animations, and turned it into clean, designer-level slides without me touching a single thing.

If you’re building anything in 2026, learn how to pitch FAST.

A good deck opens doors before the product even exists.

Comment "DECK" and I’ll send you the full prompt.