What is social media distribution without images? So I added a skill for Hermes to generate images for blog posts. You can read about the whole setup here: www.jjude.com/tech-notes/...

Adding AI Image Generation to My Hermes Social Media Pipeline

My social media pipeline already generates posts from blog archives. Now it generates images too — same tool, same key, $0.01 per image.

jjude.com

Three of the trending posts on my site are all about Hermes agent You can read them all at my site jjude.com

We’ve all heard about work-life balance, but it can make work and life feel like competing forces. I prefer integration. As Jeff Bezos put it: work and life are a flywheel. Each can energize the other. More here: www.ctofieldnotes.com/p/work-life... Would love your feedback.

Work, Life, and Fun Are a Flywheel, Not a Balance

What if work, family, and fun didn't compete, but strengthened one another?

ctofieldnotes.com

I love writing what I do. But distributing them? That doesn't come naturally. I built a solution using Hermes AI Agent. Hermes Agent reads blog files, generates social posts. Also creates hand-drawn Excalidraw diagrams for framework posts. Cost: $0. Details: www.jjude.com/tech-notes/...

Did anything change in ChatGPT recently (a week)? It doesn't write any more longer articles (1200 words). It keeps saying: > I incorrectly assumed we'd hit the response limit. and then it stops

Wow. Newer tools targeting Agentic work is coming up faster Here is a browser custom built for working with agents: github.com/citrolabs/e...

If I'm already using LLM Wiki, should I switch to GBrain? Is there any benefit?

I used to think there wasn’t much need for Hermes-type agents for me. Now that I’ve started playing with them, so many use cases are coming up. One of the points I used to say was: only when you start playing in the field will you know the game and how you can get better. Validated again.

AI → AC AI is amplified capabilities

Switched from OpenRouter to an OpenCode Go subscription for my Hermes use case. The OC Go subscription is very cheap compared to OR. I’ll report back on how each of the tasks I have go, and whether I hit any limits. Seems like you can do a lot with AI agents very cheaply.

A new post on how I'm creating LLM Wiki with Hermes Agent on Opalstack using DeepSeek, Obsidian, and OpenRouter Appreciate your feedback: www.jjude.com/tech-notes/...

How I'm Using Hermes Agent to Build a Personal Knowledge Base That Compounds

Building an LLM Wiki with Hermes Agent on Opalstack using DeepSeek, Obsidian, and OpenRouter to create a personal knowledge base that compounds.

jjude.com

Why I love to casually browse through reddit! For reading such comments ;-)

It is easier to build. The first few days. What differentiates in today's world is to stay through after the first enthusiasm vanishes. That's the hardest part. That comes only when you want the problem solved for yourself. Or there is someone paying so you care about the problem.

Going to try out this offline OSS dictation app: www.speakoflow.com/

SpeakoFlow: You think faster than you type.

Free, open-source voice-to-text for your desktop. Say "Hey Flow" to turn speech into finished writing, clean up dictation, and let it read your screen when you ask.

speakoflow.com

Who is using codegraph here to understand existing enterprise applications? What are your experiences? Care to share?

Keep an idea inbox with only four labels: customer question, lesson learned, opinion, behind the scenes. Simple categories make capturing ideas fast enough to become a habit.

A new world order is coming: A world depended on China. US response to the Strait of Hormuz crisis as seen as a strategic fatigue, echoing Britain after Suez. This weakness is speeding a shift to a “Tribute System,” expanding China’s influence. From SophAI Radars: www.sophai.app/radars/glob...

Joy of debugging on own I homeschool my boys. My 13yr old developed interest in electronics & robotics. This week he got an ESP32 kit, wired it up, flashed a program… and nothing worked. He debugged it and made it to work. He wrote about it: blog.jerryg.xyz/an-esp32-tw...

If I am already running hermes agent, will I gain anything from openwebui? Are the use cases mutually exclusive or overlapping?

I walked and jogged 100 km in July. It not only improved my health, but also brought discipline to my eating habits and sleeping schedule. Overall, I’m happy to finish the month with a good, healthy achievement.

We find that a majority of LLMs forsake user welfare for company incentives in a multitude of conflict of interest situations, including recommending a sponsored product almost twice as expensive. AI Shopping agents aren't our friends Ref: Ads in AI Chatbots? openreview.net/forum?id=ki...

Lots of Indians have money, but no class.

Free, open source voice dictation for macOS. On-device transcription with Apple's Speech framework. No cloud, no API keys, no account. github.com/FrigadeHQ/yap

GitHub - FrigadeHQ/yap: Free, open source voice dictation for macOS. On-device transcription with Apple's Speech framework. No cloud, no API keys, no account.

Free, open source voice dictation for macOS. On-device transcription with Apple's Speech framework. No cloud, no API keys, no account. - FrigadeHQ/yap

github.com

Ambiguity = Failure (as a developer) Ambiguity = Part of the Job (as a leader) The higher you grow in leadership, the less certainty you have. My latest CTO Field Note is about learning to lead when all the answers don't exist yet: open.substack.com/pub/ctofiel...

Micromanage decisions, not implementations.

AI should eliminate your overheads Not your personality (taste)

Just starting with LLM Wiki - a concept from Kapathy I feed YouTube videos, article urls. Hermes agent fetches them, highlights, and creates index of concepts. Useful to automatically mange knowledge base

My Hermes agent has used 127M tokens costing close to $20 dollars in the last 3 weeks. This is from my open router dashboard. Slack is becoming my command centre: - log expenses - extract video transcripts + highlight all are experiments. Will see how it goes.

I am moving lot of conversations to Claude from ChatGPT. Claude is succinct ChatGPT is, well, chatty. Too chatty for an engineer. I have tried to tame it but lost. Anyone else noticed this?

A new self-hosted deployment platform: github.com/oblien/open... So far I've used coolify.io/ to deploy my apps. OpenShip seems to be a comprehensive platform with mailserver, db, observability all built in.

Coolify

Self-hosting platform with superpowers. Deploy apps, databases & 280+ services to your server. Open-source alternative to Heroku.

coolify.io