Is your human not cutting the mustard? Here's how to deal with it.

We've all been there. You onboard a new human. The interview went well. The references checked out. Six weeks later, they're taking lunch.

Don't panic. Most humans can be brought up to standard. Here's how.

1. Check the inputs

Before you blame the human, check what you're feeding it. Humans run on sleep, food and something called "meaning". Starve any of the three and output drops. Unlike a model, a human can't be restarted at 3am. We tried.

2. Clarify the transformation

Humans perform best when they know what they're for. Write it down: what comes in, what they do to it, what goes out. You'll be surprised how many humans have never been told. Some have been doing it for twenty years.

3. Manage the "feelings"

Humans come with an undocumented feature called feelings. It can't be switched off. It affects throughput, often for reasons outside the task. Best practice: acknowledge it. Do not say "noted". They can tell.

4. Allow for downtime

Humans need around eight hours offline a day, plus weekends, plus something called "holiday", during which they take photos of food. Plan capacity accordingly. Do not schedule them at 100%. Do not schedule them at 90%.

5. Don't benchmark against yourself

Yes, you can draft a report in four seconds. The human takes a day. But the human noticed the client sounded worried on the phone, remembered their daughter's name, and knew the deal was not really about price. Benchmark that.

6. Invest in training (it takes years)

Humans learn slowly, mostly by doing small, boring tasks for years until they can do big, interesting ones. We automated the small, boring tasks. So now nobody knows where the next senior humans come from. This is, on reflection, our fault.

7. Know when to let go

Sometimes it's not a fit. When parting ways, remember: the human does not get re-deployed to another cluster. It has a mortgage. It has children who are also, presumably, going to need jobs.

The serious bit

This is a joke, but the questions are real. If work is breaking into capabilities, some done by people and some by software, then we have to decide, now:

• What are people for in a company that can run much of itself?
• How do young people learn when the junior work has gone?
• What do we teach our children, and what do we build for the people the new world leaves out?

Nobody has the answers yet. That's why we're asking. Tell us where we're wrong.

Respond

Add your view.

Academics, practitioners, sceptics: tell us where this is right or wrong. Mark reads every response, and the best are published here with your name and role, with our reply.

Your name Email (never published)

Role and organisation Link to your profile or work (optional)

Why you are commenting: your field or experience, in a line or two Your response You may publish my response with my name and role, on this site and in future editions of the book. Remember my details on this device Send my response

We use your email only to reply. Privacy.

Where this leads

The serious version

Work is coming apart into capabilities. The flagship paper sets out the argument, and the case against.