This open source setup beats all productivity apps



Thank you to Hostinger for sponsoring this video! I reached out to them because they have a super simple way to copy this open source setup without having to do anything technical (Literally takes one click after purchasing). You can check it out here:
– Link: https://hostinger.com/REYSUHERMES
– Coupon Code: REYSU (for an additional 10% off)

Outline:
0:00 Productivity assistant setup (What it can do)
0:48 Every AI Assistant and Agent explained
2:15 Why I use this open source AI assistant (Hermes)
4:00 Hermes full setup guide using Hostinger (in 2 mins)
6:13 How I use my AI assistant for work
9:51 Using my AI assistant to fill out forms
10:41 Using my AI Assistant to help me reorder supplies
12:05 Control an Android Tablet (and Track macros)
14:49 Building custom apps on the go
17:00 Two mistakes with setting up an AI Assistant

Playlists:
📱 Tech and Gadgets: https://www.youtube.com/watch?v=EoyX8rzEopA&list=PLacX82b_pofx–V2uqJmJVMGOfsJDBSm-
📖 Journaling guides: https://www.youtube.com/watch?v=OmyfB513E1s&list=PLacX82b_pofztutMWary6SRslMTvYVzsi&index=1&t=229s&pp=iAQBsAgC
🧑‍💻 Productivity Optimization: https://www.youtube.com/playlist?list=PLacX82b_pofx0W3JjHZzi795ocRUZrSfT
🎤 Outliers Podcast: https://www.youtube.com/watch?v=qoUP3SY-Gy4&list=PLacX82b_pofx4w0E-5QdA-L_VUTHYeO3l

source

35 Comments

  1. I sent you a DM on Instagram, but I guess your system didn’t detect it as important. I’ll wait for your response.

    I’m going to drop some buzzwords here just to see if they trigger the alarm: AI, Hermes, Grok, automations, money, Japan, business.

    Hopefully this one gets through. 😅

  2. I like the idea of using apps you enjoy like Macrofactor. What’s the process for getting Hermes hooked up to the spare Android tablet. Does that speak with the MacBook Air

  3. This is the kind of AI I actually find useful. I don’t need it to run my whole computer, but I do like having something that helps me decide what to focus on. I use Actimate and their AI assistant helps organize my goals into a few clear priorities so I’m not wasting energy figuring out what to do next.

  4. If you’re still using an online model, there is no benefit regarding your data. All the model touches still leaves your control and the model provider may do anything with it.

  5. Hey Eric, I’m that freelance logo designer who reached out to you, had you have clear direction for your channel yet? I have a coding themed merch graphic I sent to you to have a look, would you like to have it as merch on your channel? 🙂

  6. Great video! I’m running pretty much the same exact setup, but on a Mac mini M4 that I managed to grab really cheap. Running Hermes paired with Codex, hooked up to my standard $20 ChatGPT subscription. Having a dedicated, always-on M4 Mac mini headless in a corner is genuinely the ultimate low-power, high-efficiency setup for this. Keep up the awesome content!

  7. the architecture behind it is still folder indexed just like any other claude + obsidian setup right? Only addition/ improvement is dedicating a machine to have it run 24/7 and using hermes for telegram integration? just trying to understand how to set it up myself

  8. Bro, genuinely, thank you for changing my life.

    I bought an iPhone 13 Mini because of one of your videos. I started journaling because of you. I started using a notepad for my to-do list because of you. All of it has genuinely helped me use my phone less and become way more organised and intentional with my life.

    And now seeing what you’re doing with AI… this feels like the next step.

    As someone who’s grown up surrounded by technology, social media and constantly having the newest stuff, AI can honestly feel quite overwhelming. Your videos make it feel practical, simple and actually useful rather than just another thing I need to keep up with.

    I’m only halfway through this video and I’m already blown away 😂. I’d genuinely love to see a more detailed video on your whole setup and how everything connects, even if that’s something you do as a members only video.

    You probably don’t realise how much impact your content has had on people. You’re genuinely like a work coach/life coach without even knowing it. You’ve helped me become more organised, more planned and more intentional with my time.

    So yeah bro, just wanted to say thank you. Love you, love the content, and keep it up. ❤️

  9. my dog you are using opus 5 tokens for an agent? what does that run you monthly? thats the whole reason i stopped using assistants, without a harness/monthly subscription its way too expensive to get a model thats actually capable enough to do my work

  10. Just have it code your own to do and productivity app

    Mine opens in browser, but its hosted on a mini pc and accessed off network using tailscale

    You can add it as an app link on your phone, it will still open in browser but in terms of friction its just as easy as any other app

    I had it merge like 3 different apps I liked into one, and its completely customisable this way, only include the features you like and will use

  11. Great video, love the ideas shared here to use AI to create a unified experience between devices. I think this is the future we've been waiting on, but isn't quite there yet (on average at least). What did you use to edit this video? Ton of really thoughtful and great animations, especially curious about the Calendar part 8:09.
    The one thing I disagree with is using a cloud model for delegating my personal life. I know they're much more capable, and economically-speaking, spending $20 for those capabilities is great… but then you're providing a ton of personal information to OpenAI or Anthropic. Coding, and impersonal things, sure—but I'd draw a line in the sand for handing off personal tasks like that, even if its way better, more capable, and cheaper. What you're giving up in your data isn't worth the time savings or value it brings. Cannot delete the data, and once its created, you'll never know where or how it's stored/used. Local models on small compute like M1 or mobile devices aren't there yet, but if you have the horsepower, local models are quite capable, and can remain private.
    Thank you for sharing, and love to see the journey here.

  12. I tried a lot of apps and combinations and in the end I came up with de following setup: Open WebUI+Ollama (first local for intentional and deep work, my private ChatGPT) & Hermes (more like a personal assistant, it's actually called Jarvis). Also you can use cloud ZDR models if needed, powered by OpenRouter. Great video as always, thanks for sharing.

    Greets from Spain.

Leave a Reply

Your email address will not be published. Required fields are marked *