My husband reportedly describes me as “as stubborn as a two-year-old saying no.” Whether that is a compliment remains to be seen.

I already wrote about the 10,000 SEK experiment and about handing Grok Bot a live tape on a Hormuz day. The log since then is not funny.

Day 1 — It worked. Fine. Beginner’s luck, or the one day the universe decided to behave.

Day 2 — Offline for several hours of the trading session. Power trouble tied to Memphis, which matches a public Grok outage that morning. I was sitting here refreshing like an idiot.

Day 3 — The bot sat quietly in the boat and made no trades.

Day 4 — Same boat. Same silence.

Day 5 — Same again. Then I found the bug: the code had effectively forbidden day trading. That one is on me.

Day 6, today — First session with day-trading instructions actually in place. Now the remote machine is acting up. In hindsight that fits days 3–5 too. I have had to log back in more times than I want to admit.

Any sensible person would have killed the experiment by day 4 and gone outside. I am not doing that.

I am stubborn enough to give this two more weeks. If Grok Bot still will not get back in the game, I will try Claude instead. That is not a tantrum. It is a deadline I am writing down in public so I cannot quietly move it.

How this setup is supposed to earn a little trust

I am not a quant, not a licensed adviser, and I am not building a product. I am a mid-50s amateur with a spreadsheet habit and a 1990s story I never let go of. That is the experience part: I am doing this with my own money, in public, including the days where nothing happens.

Expertise, such as it is: I wrote the constraints, I found the day-trading bug myself, and I keep the rest of the household accounts away from the bot. Authoritativeness is the bit I refuse to fake. One working day and four dead ones do not make a track record.

Trust is the only piece I can actually design. The bot sits in its own cage — a separate ISK at an Irish-based broker, 10,000 SEK, no BankID on the family platforms, no keys to the main portfolio. If it blows up, the blast radius is that ticket. I will publish the ugly chapters as well as the pretty ones. None of this is advice. Read the disclaimer if you need it in formal language.

This is still a 10,000 SEK ticket, not a strategy. If it dies, I will write that too.

Would you have scrapped it on day 3, or are you also the two-year-old? Tell me. I will post the next chapter when the two weeks are up — or sooner, if the boat finally moves.

Disclaimer: This is a personal experiment and should not be read as financial advice. Trading involves a high risk of loss.