My Week #219 - Life with an assistant

Automation is the future that has already come

I set up my personal assistant with OpenClaw a week ago. After the first week, I can already see that, on the one hand, this is not a revolution—but on the other, it clearly points toward the future. It finally enables building human-like automations without strict rules or the need for sophisticated software to gather structured inputs. Now, with LLMs, we can provide inputs in a very flexible form, and the system can generate outputs based on that input and its context.

To give an example—my health assistant asks me every day about the meals I ate the previous day. I simply type them in whatever format is convenient for me. Based on this, it analyzes the data and provides feedback. It’s like having a personal diet assistant.

As another example, I have an assistant responsible for real estate research. We want to build a house, so the first step is buying a plot of land. Previously, I would scan listings manually or rely on basic notification systems, often missing good opportunities. Now, my assistant performs searches daily and returns with a report.

These are just two basic examples, but they clearly show the direction:

• find a task you perform on a regular basis
• check if you can automate it
• create a dedicated assistant to handle this task
• test the initial results and refine it over the first few iterations
• let it run and focus only on the outputs

I have plenty of other tasks that could potentially be automated, but I’m cautious about granting access to sensitive data. Because of this, my automation won’t go beyond a certain point—but I’m fine with that.

Summary

✅ Continue physical training

✅ Continue setting up OpenClaw

Plan for week #220

💻 Start building the new mobile app

💪 Continue physical training

🪏 Complete the counstruction of fence

Weekly reminder

🏆 Completed marathon

🥈 Released game remaster