A place to learn and a network to build with.

Maurice Hickman, “Mo” · Business & Leadership

AI is changing the work your people do. It is also changing the preparation, speed, and range of options your customers bring into the room. Leadership has to watch both movements at once.

AI does not arrive with a strategy in the box.

A model can help your team hold more context, connect signals, explore options, and produce a useful first pass. It still needs an outcome, trusted inputs, and someone who can judge the result. Buying a license does not answer any of those questions.

Your customer is gaining capability too.

If customers can do more preparation themselves, what will they expect you to contribute next? Product leaders, operators, and commercial teams have to connect those views before the market does it for them.

Choose an opportunity the business can carry.

Map possible workflows, then compare likely value with time, resources, risk, and the people required. A small change that becomes part of the work can be more valuable than an ambitious system nobody owns.

• What repeated work consumes judgment without creating much value?
• How might AI change what customers bring to us?
• Which improvement has an owner and a route to adoption?

Zain Haseeb, “Big Zain” · Product & Adoption

Leadership says the company needs to “use AI more.” A few people open a chatbot. Activity rises. Then the quarter ends and nobody can explain what changed. That is not resistance. It is a design failure.

Usage is a weak proxy for adoption.

Counting logins does not tell you whether a decision improved, a handoff became clearer, or a customer received something better. Pick one recurring piece of work with a real owner, a visible output, and a problem the people doing it recognize.

The workflow carries hidden context.

Experienced teams know which source can be trusted, when an exception matters, who approves a change, and what a strong answer looks like. Every correction can become a clearer instruction, better source, approval boundary, or test.

Build the learning loop into the job.

Run one live, reversible workflow. Compare it with the current way of working. Ask what saved time, what lost context, and where judgment still had to enter. Then adjust and run it again.

• Name one recurring output, not a tool.
• Write down sources, judgment calls, and exceptions.
• Run one reversible test.

Read Zain’s related public writing

Zain Merchant, “Lil’ Zain” · Engineering & Governance

Vibe coding lets people who understand a problem bring an idea to life. But a prototype earns a different question the moment a business starts depending on it: can we trust it?

What is the source of truth?

A business tool needs to handle incomplete, outdated, duplicated, or conflicting information. Results should carry the evidence that lets a person understand why they surfaced.

What is the system allowed to touch?

Reading a copied file differs from changing the original. Drafting a message differs from sending it. Define access, decision rights, and approval points before convenience turns into authority.

How will we know the answer is right?

Decide what quality means for the job, then test awkward examples, not only the one that inspired the prototype.

How do we recover?

Logs, versioning, backups, and a human escalation path let a team use the tool with confidence.

• Can we trace the answer to trusted inputs?
• Does the system have the least authority it needs?
• Can a reviewer test quality and reverse the result?