The Human First Code – Our AI non-negotiables
It’s the topic everyone is talking about, how AI will shape our working (and personal lives) over the coming years. At AVM Solutions, we have been using AI for a number of admin tasks for years, and more recently we undertook a project to formalise and manage the use of AI withing the business in a meaningful and responsible way. The start of this journey involved sitting down and writing a list of non-negotiables for best practice. These act as principles for harnessing all the time saving efficiencies, without losing what is most important in our offering, the Human experience.
Why This Exists
Social media was meant to connect people. For a while it felt like it did. But over time, something shifted. We were communicating more but connecting less. The tools that were supposed to bring us closer had quietly worn down our ability to have real conversations and build real relationships.
AI is heading the same way. It is a useful technology and it will change how businesses work. Used well, it takes away the repetitive computer-based tasks and frees people up to do what only people can do. Used badly, it creates a team of people who produce good-looking work without ever using their own mind.
At AVM, we are not making that trade. This business was built around our partners. Everything we do starts with them. We lead with consultancy, not product. We put people first, not technology. No egos, do the right thing. That does not change because a new technology has turned up.
The Human First Code sets out where we use AI and where we choose not to. It is not a policy that sits in a drawer. It is how we work. Every person at AVM is expected to understand it, agree with it, and follow it.
These principles are built into every AI interaction across the business. They are not words on a page. They are how the tools actually work
Everything in The Human First Code comes from one belief:
If a task needs speed and accuracy, AI belongs. If a task needs heart and judgement, we do it ourselves.
AI should make us more capable, never less human.
AI is a tool. It does what machines do well — process data, do calculations, pull information together — so that people can do what people do well: think, connect, care, and make calls that need you to understand another person.
If a task needs speed and accuracy, AI belongs. If a task needs heart and judgement, we do it ourselves.
The Seven Principles
1. Human Thinking First
AI does not think for us. Before AI shapes, tidies, or structures anything going to a partner, the person must provide their own thinking — their intent, their key points, their view. AI can then help with the mechanics: pulling data, formatting, checking for gaps. But the substance comes from the person.
Example: A PM writing a project update starts with their own notes about what matters. AI helps pull the latest data from monday.com and tidies the format. But what to say, what to highlight, and what tone to use — that is theirs.
2. A Tool, Nothing More
AI handles the analytical, technical, and data work. People handle connection, communication, and judgement. We keep these separate on purpose. We use AI to reduce the digital admin that has to happen on a computer. We do not use it to replace the human work of building relationships, having difficult conversations, or making decisions that affect people.
Example: AI can put together a pricing review, build a cable schedule, or check a solution design. It should never decide how to break bad news to a partner, what tone to use in a tricky situation, or whether to pick up the phone instead of sending an email.
3. Visibility Over Prompting
AI shows people the information so they can see it and act. It does not tell people what to do or what to think. A dashboard showing a project is behind schedule respects the person — they see it, they act. A notification telling them to call the partner treats them like they cannot think for themselves. Prompting creates laziness. Visibility creates ownership. We choose ownership.
Example: Instead of AI sending a PM a reminder to follow up with a partner, we build dashboards that show project status clearly. The PM sees the data and decides what to do. The responsibility stays with them.
4. Our Values Are Split
AVM has six core values: Integrity, Positivity, Empathy, Relentless, Caring, and Innovation. Not all of them belong to AI. AI carries Integrity — it does not cut corners, does not produce anything misleading, and is honest about what it does not know. AI carries Innovation — it finds better, quicker ways to do the technical work. AI carries Relentless — it is thorough and does not half-deliver. But Empathy — seeing it through someone’s eyes — is a human skill. Caring — giving a damn about the work and the people — is a human choice. Positivity — bringing energy into a room — is a human quality. AI leaves space for these values. It does not try to copy them.
5. Values Over Profit
AI must never put speed or efficiency ahead of doing the right thing. If taking a shortcut saves time but drops the quality of our work, damages a partner relationship, or weakens our advice, we do not take it. This is the same principle that led AVM to take a large financial hit on a project rather than walk away from a problem that was not entirely ours. The technology changes. The principle stays the same.
6. Honest by Default
How we use AI is not a secret. We do not put a label on every document or email, but our approach is openly available and we will tell anyone who asks exactly how and where AI is used. AVM has always been straight with people. AI does not change that. Being open about it is not defensive — it shows we have thought about it properly and we are comfortable with the answer.
7. Earn It First
For people skills — communication, writing, presenting, advising — do not ask AI to do something you could not do yourself. The time AI saves you is only worth something if you already have the skill underneath. A junior team member needs to learn to write a project update before using AI to help write one. An engineer needs to understand cable scheduling before using AI to produce one. For technical work, the rule is slightly different but just as important: always understand and be able to check what AI has produced. Never send, submit, or act on something you could not explain if someone asked.
Example: If AI produces a solution design, the engineer must be able to explain every device, every connection, and every calculation in it. If they cannot, they are not ready to use AI for that task.
The Human First Code is not a list of restrictions. It is about being thoughtful with a useful tool.
Every day, you will make small decisions about when to use AI and when to do something yourself. No document can give you the right answer every time. What it can do is give you a way to think about it.
Before using AI, ask yourself three things:
1. Am I using this to save time on something I can already do — or am I using it to avoid thinking?
2. If a partner asked me to explain this, could I?
3. Is this making me better at my job, or is it making me lazy?
If you are honest with yourself on those three questions, you will get it right nearly every time.