Building a voice agent to handle phone calls. Setting up a Slack channel where agents talk, not humans. Is the future here yet? Not quite. It will be when an agent builds this autonomously (and correctly), not me.

Not only does it not count as execution, but it also does not count as an active workflow. Cool new feature to easily create the sub-workflows on n8n.

Need to make ChatGPT voice mode available through Apple CarPlay. If WhatsApp, calls, and audiobooks are available, why not make ChatGPT available too?

You can take a man out of the country, but you can't take the country out of a man

Running on double espressos and music from FISHER these days. youtu.be/PNPpSPS1nok?si…

Lovable, Bolt, etc., but Supabase here is the real MVP and should see their customer base growing 📈

.Marshall Mathers's Lose Yourself But the word "music" replaced with "AI"

Just used NotebookLM to create an explainer video about a solution I built. Pretty amazing. If I could add my avatar somewhere in the corner, this would be an easy way to create and share tutorials (but not sure someone would watch, now you only ask AI to summarize & explain)

I have some .txt files saved with prompts and outputs from 2021 using Davinci from OpenAI. 2021 Davinci: “Sorry, you’ve hit the token limit.” 2025 GPT: “Is that all you’ve got?”

Made another video with NotebookLM Gave the json (converted to .txt) of a n8n.io workflow I made, added 2 links to Retell AI docs. And the video is made.

Nice little update on n8n.io when you activate a new workflow. Especially about error notifications, as you don't receive any notifications by default if you have not set up a workflow for managing errors.

The duality of web development.

The wait is finally over – TildeOpen LLM is now live on Hugging Face! 🎉🎉 TildeOpen is our new open-source foundational large language model with 30B+ parameters, built from the ground up with European language data at its core. 🔗 tilde.ai/tildeopen-llm/

That's how I like to see my n8n.io

GenAI and life have something in common: it’s beautifully effortless to create an outcome when you’re vague, but it gets much harder when you carry a precise picture of what you want in your mind. Anyway, the difficult things are worth it!

Have noticed that Google Gemini will very rarely mention itself when suggesting which model to use. Most of the time, it mentions one of the GPTs from OpenAI

What's the best way (if any) to add a custom voice to a video generated with these premium models when using Higgsfield AI 🧩? If using Lypsync Studio, then it offers different types of models

Many are sharing amazing results created with Kling 3.0, so here is something less impressive. "Vibe directed" by someone who is not working with video daily. Created using the Higgsfield AI 🧩 platform

Wait, you want to tell me it was possible to acquire Swedish citizenship without speaking the language?!

I am wondering, how it is for Karol Nawrocki to watch his national team's game, where the leading player has nothing, completely nothing connecting him to Poland. DISGRACE