Curious to hear from you all on how you're using AI agents. Are you: > Building AI employees for your own business to help you run more efficiently? > Building and Deploying AI Employees for Other Businesses? (AI automation agency) if you could also let me know the biggest challenges you're facing in the process of all of this, that would be super useful -- based on how you're building and deploying agents, be it for yourself or for others, I'll know how best to create content to help ! Thanks for being here! Super excited to pump out more content and build together :)

I said Grok 4.5 was a bigger deal than Fable Elon even liked my post but then GPT-5.6 Sol showed up So I gave 7 of this week's newest AI models their own computer and ran the same real tasks through Hermes Final ranking: 1. GPT-5.6 Sol: best overall 2. Grok 4.5: close second, fast and relentless 3. Muse Spark 1.1: the surprise of the week Benchmarks tell you what a model might do. Watching it control a computer shows you what it can actually do Full comparison ↓

Cool video @Nick Vasilescu just released that shows a cage fight in orgo with multiplied models running their own instance of Hermes . Was pretty wild. I have to say Sol and Terra impressed me the most. Unreal how fast they worked. I have played with Sol on a personal project and it solved two issues that DeepSeek v4 flash just couldn't figure out for me. https://youtu.be/BFRmOxT8QIo?is=GEqkxLV5KPTO0eok

Sharing this because it cost me more troubleshooting time than it should have, and it might save someone else the same rabbit hole. In my setup, I’m using a local Hermes/Merlin agent as my main assistant, and an Orgo VM as a secondary remote agent/computer. I’m using Obsidian Sync so the VM agent can access the same context vault. The simple lesson: The Orgo VM is its own computer. That means if Obsidian is installed on the VM, it has its own Obsidian app, its own local copy of the vault, and its own Sync settings. It is not automatically “looking through” to the Obsidian app that is open on your laptop. What tripped me up was that everything looked connected: - same Obsidian account - same remote vault - same vault name - regular markdown notes were syncing …but some support files still weren’t showing up on the VM. There were two things to remember. First, Obsidian has to actually be running on the VM for Obsidian Sync to keep that VM’s local vault up to date. If the VM restarts and Obsidian doesn’t reopen, the vault on the VM can fall behind even though your laptop vault is perfectly current. So if your agent relies on Obsidian as a context vault, make sure Obsidian starts again after a VM reboot. Second, check this setting on the VM too: Settings → Sync → Selective Sync → Sync all other types In my case, this was turned on locally, but turned off on the VM. That meant normal notes synced, but non-note support files did not. Quick checklist before you spend hours debugging: - Are you signed into the same Obsidian account on the VM? - Is the VM connected to the correct remote vault? - Is Obsidian actually open/running on the VM? - Does Sync say fully synced? - Are any folders excluded? - Is Sync all other types turned on? - If the VM restarts, does Obsidian reopen automatically? This may not apply exactly if your whole setup lives inside the VM, but the mental model still matters: A VM is a real computer. Apps inside it need to be opened, configured, synced, and restarted just like apps on your laptop.

I just built a full AI employee from scratch in this video. Everyone’s hyped on agents but barely anyone’s actually showing how to build ones that can really work inside a business. Here’s the entire stack I spun up live (no holding anything back): • Cloud computer running Hermes on Orgo • Its own email (AgentMail) • Its own phone number (AgentPhone + iMessage) • Telegram as the main chat interface • Every tool & connector hooked up through Composio • Credit card so it can actually spend money • Obsidian vault as its knowledge base / second brain • Latitude for observability so I know when shit breaks I made one main orchestrator agent (named it Hubert) that stays in charge 99% of the time and just hands tasks off to specialized sub-agents underneath it. One giant bloated agent is a nightmare to debug. This way everything stays clean and purpose-built. Full live build, every prompt I pasted, every terminal I opened… it’s all in there. And I’m giving away the complete templates for the orchestrator + every sub-agent I build going forward. I'm giving you actual employees that can do real work. Watch the whole thing here: PS - the template is free here on Github: https://github.com/nickvasilescu/nicks-stack