Absolutely.
AI is making writing code cheaper, not making engineering irrelevant.
The valuable skill is shifting from “Can you code this?” to “Do you know what should be built, how it should work, and whether the AI-generated solution is actually good?”
Claude and ChatGPT are leverage. They’re much more powerful in the hands of someone who understands software.
Hi all on @X!
I'm looking to #connect with people interested in:
- Micro-SaaS & Startups
- AI Tools & Agents
- Vibe Coding
- Building in Public
- Vibe Marketing
Let's grow together 🤝
#letsconnect#buildinpublic#founders
Builders let’s connect 🤝
I’m looking to meet people building in:
🚀 SaaS
🧠 AI & AI agents
🌐 Web apps
📱 Mobile apps
🛠️ Dev tools
📊 Data & analytics
💰 Bootstrapped startups
🤖 Automation
Drop what you’re building below 👇
I’ll check it out, and let’s connect.
5 prompts every manager should steal:
The Decision Challenger
“I’m considering [DECISION].
Act as a skeptical senior executive.
Challenge my assumptions, identify what I’m missing, and give me the 3 strongest reasons NOT to do it.
Then tell me what evidence would change your mind.”
The 1:1 Prep
“I’m having a 1:1 with [PERSON/ROLE].
Context: [CONTEXT]
Create a 30-minute agenda that helps me uncover blockers, motivation issues, career concerns, and feedback I’m not hearing.
Avoid generic questions.”
The Executive Update
“Turn these messy notes into an executive update:
[PASTE NOTES]
Structure it as:
• What changed
• Business impact
• Risks
• Decisions needed
• Next steps
Keep it under 150 words.”
The Delegation Prompt
“I need to delegate this:
[TASK]
Help me define:
• Expected outcome
• What ‘good’ looks like
• Constraints
• What they can decide independently
• When they should escalate
• Checkpoints
Make it clear without micromanaging.”
The Team Problem Solver
“My team is struggling with:
[PROBLEM]
Don’t immediately give me solutions.
First identify the likely root causes.
Then ask me the 5 questions a great leader would need answered before acting.
After that, propose 3 possible interventions ranked by impact and effort.”
Save this.
You’ll use at least one of these this week.
INSTEAD OF WATCHING AN HOUR OF NETFLIX TONIGHT.
This 60-minute Cambridge lecture by Demis Hassabis will teach you more about the future of AI than most people will learn in the next 5 years.
Bookmark it and give it an hour, no matter what.
Most AI projects don’t fail because the model isn’t smart enough.
They fail because the workflow around the model is weak.
A useful AI system needs:
→ a clear trigger
→ the right context
→ a defined action
→ a human checkpoint when risk is high
→ feedback from what happened next
The model is one component.
The workflow is the product.