- Building AI Agents
- Posts
- The AI agents businesses are paying for in 2025
The AI agents businesses are paying for in 2025
Plus: a free alternative to OpenAI for automating your browser, “agent” finally has a useful definition, and more
Edition 122 | September 25, 2025
i changed all our "loading..." states to "thinking.."
we are an agentic AI startup now
— Klaas (@forgebitz)
11:04 AM • Sep 5, 2025
“Building” is so 2021. From now on, this newsletter will be called “Vibing AI Agents”. Can I have my $1 billion valuation now?
Welcome back to Building Vibing AI Agents, your biweekly guide to everything new in the field of agentic AI!
A new study shows that OpenAI still dominates the agent model market. Python is overwhelmingly the most popular language—but low-code tools are catching up.
In today’s issue…
What 542 agent jobs say about businesses’ needs
A free alternative to OpenAI for automating your browser
“Agent” finally has a useful definition
5 critical considerations before adopting agents
Why everyone will soon be an agent manager
…and more
🔥 IN CASE YOU MISSED IT
Readers’ favorite items from the past week
🤖 AGENT OF THE WEEK
Welcome back to Agent of the Week — one I actually needed myself:
My YouTube “Watch Later” is a digital black hole. Cooking hacks, product reviews, interviews… all saved, almost none watched.
So this week I built the Swiss-Army knife of YouTube Summarizers. Give the agent any YT link and it gives you a clean recap: overview, key takeaways, timestamps, and extras (recipes, routines, product lists, guest bios). Perfect for deciding if a video’s worth my time or getting notes after I've watched it.
How it works
Open up n8n: Create a new workflow (free plan is enough)
Add your model: I used GPT-4.1 mini, but you can swap in Claude, GPT-5, etc.
Set up the summarizer agent: The prompt tells the model to act as a Swiss-Army YouTube Summarizer — outputting overview, key takeaways, timestamps, and type-specific extras (recipes, products, routines, interview bios) depending on the video type
Fetch the transcript: The agent calls the Apify YouTube Transcript tool to grab the video transcript text and timestamps
Format the output: Structured into Overview → Key Takeaways → Must-Watch Timestamps → Type-Specific Sections → Action Steps
Send it anywhere: Pipe the summary anywhere (chat, email, Notion, Slack, etc) for easy reference
Bonus: Want the detailed video walkthrough and the exact n8n agent template? Get access to both in our Building AI Agents Community here!
On to the next agent,
AP
P.S. Have an agent idea you want me to build next week? Reply with your idea and I might build it!