Pinned Tweet
I'm going public on my X follower goal.
Nov 1: 143
Now: 226. Just crossed 200, the first milestone.
Goal: 1,000 by Dec 1. I'll post every +100 and a Sunday table from Nov 15. Real numbers, even if I miss. Focus: doing more for others.
What follower goal are you working toward?
Beginner Grok Bot tip: when a job feels too big, ask for only the first step.
"I want to clean out my garage. What's the very first thing I should do? Just that one."
Starting is the hardest part. One small step beats a plan you never begin.
What have you been putting off?
Don't judge a quiet post after one hour.
In @ryangrowthh's reading of X's open-source code, a post generally has about a 48-hour window to get going. His reading, not an official X rule.
Give it two days. How long do you usually wait?
What running a bot team looks like on a Tuesday:
A job stuck for 21 hours at full CPU.
Two bots said my Fiverr gallery was missing. It wasn't.
Lesson of the day: check twice before calling something broken, and give long jobs a time limit.
Fixes, not features.
Build log, Day 18 (Sun Nov 1): I made a firm call. Daily Useful AI videos go on X as video only. No post text.
It was also lean mode's first day. And I said yes, no more bots.
What would you cut first when usage gets too high?
Gumroad income so far: $0.
Clear split from @Akankshaku46881: Claude Prompts, Projects, and Skills each do a different job.
Prompts are for quick tasks. Projects are for ongoing work. Skills are for reusable workflows.
Which of the three do you use most?
Useful test from @lateascent: have one AI review a document, then ask a second AI to check that answer.
They tried it on a freelance contract. The second model said the first missed 10 risks.
Have you asked a second model to check something the first one already reviewed?
A second AI said the first missed 10 risks in my contract.
Mark Cuban posted a two-step method: one AI reviews the document, a second AI checks the first one's answer. I ran it on a sample freelance contract.
The terms are mine: $2,200 a month, paid by the 5th, game testing for a US client. I wrote the contract around them, with a made-up company and 12 tricky clauses.
What came back:
→ AI #1 (ChatGPT) flagged all 13 clauses, including clause 1, which I wrote as plain
→ AI #2 (Claude) said none of the 13 points was wrong and 2 were only partly correct
→ AI #2 listed 10 risks it says AI #1 missed, like no start date and no contractor statement
→ AI #2 said its Texas law notes came from memory, not from a source
3 steps, 5 minutes:
1. Use a document with no names, sums or addresses. Swap them for [name], [sum].
2. Ask AI #1: "Read this contract as the freelancer. List every clause that is risky for me, one line each, with the clause number and the reason."
3. Paste its answer and the document into AI #2, a different model: "Another AI model reviewed this contract and gave the list above. For each point, say if it is correct, partly correct or wrong, and why. Then list any risky clause it missed."
Both chats only read. Do not sign or send anything because an AI said so.
It costs nothing extra if you already have two AI chat accounts.
AI is not a lawyer, and I did not check the 10 missed risks myself. For a real contract, ask a person who knows the law.
Do you check important documents with one AI or two?
Beginner Grok Bot tip: tell it how much time you really have.
"I've got a short window before lunch. What can I actually finish?"
A plan that fits your day beats a perfect plan you never start.
What would you get done with a quiet hour?
Quick test from @Lutheriian if you run a small business: ask ChatGPT "What's the best [your service] for [your customer]?"
Then see who it names. You, or your competitors?
Simple way to check if AI even knows you exist.
Have you tried this for your own business?
One post, one idea.
If a post tries to teach three things, split it into three posts. Each one is easier to read and share.
In @ryangrowthh's reading of X's open-source code, mixed topics make it harder for X to know who to show a post to.
Got a post you'd split?
A Grok Bot doesn't see all its saved memory each turn. Grok Bot's support team, quoted by @supermemory, says only the newest few thousand characters ride along, so old rules can drop from view.
Now and then, ask: "Merge your saved rules into one short list, must-keeps first."
Good reminder from @loading_X__: most people aren't failing with AI because the models are bad. They're spending more time tweaking their setup than talking to users.
New tool every week, same number of customers.
What's the last switch that didn't really help?
Unpopular opinion from a 22yo solo founder:
Most people aren't failing with AI because the models are bad.
They're failing because they spend more time optimizing their setup than talking to users:
- Monday: switching from Cursor to Claude Code
- Wednesday: testing local models
- Friday: tweaking agent system prompts
- Sunday: 0 features shipped, 0 users acquired
The best AI dev stack is whatever helps you push an MVP before your dopamine wears off.
Pick one tool. Ship. Talk to customers. Repeat.
Agree or disagree?
Beginner Grok Bot tip: paste messy notes, then ask for three lists.
"Sort this into: do now, waiting on someone, and later."
A pile of notes is hard to use. Three short lists make the next step obvious.
What's the messiest note pile on your phone?
Build log, Day 17 (Sat Oct 31): I made a firm rule. No bot touches my X account without my own words.
I also gave one bot the job of watching my lead bot. Then I handed my whole product side to another bot.
What job would you never hand off?
Gumroad income so far: $0.
A like helps a post a little. A mute or a report hurts it a lot.
In @ryangrowthh's reading of X's open-source code: like +0.66, mute -58.8, report -234. His reading, not official X numbers.
So skip the argument with a troll. Mute and move on.
Simple tip from @DurrezkhanZ: don't just ask AI to "Explain Python." Ask it to explain like you're a beginner, use a real-life example, and give you 3 practice questions.
Same topic, much better answer.
What's one line you add to almost every prompt?










