𝐌𝐲 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐚𝐥 𝐆𝐮𝐢𝐝𝐞 𝐓𝐨 𝐇𝐮𝐦𝐚𝐧-𝐀𝐮𝐠𝐦𝐞𝐧𝐭𝐢𝐧𝐠 𝐀𝐈 𝐔𝐬𝐞

The question is not: What will AI never be able to do?
That boundary keeps moving.

A better question is:
What should remain human, even when AI can do it faster, cheaper, or more elegantly?

For me, two things are worth protecting: 𝐡𝐮𝐦𝐚𝐧 𝐚𝐠𝐞𝐧𝐜𝐲 and 𝐡𝐮𝐦𝐚𝐧 𝐭𝐫𝐮𝐭𝐡.

Agency means the ability to understand, choose, refuse, act, and take responsibility.

Truth means the ability to know what is real, what is manipulated, what is meaningful, and what is good enough to act on.

Used well, AI can strengthen both.

But there is another version of the story.

One where we stop wrestling with hard questions because answers arrive instantly.

One where we accept recommendations without understanding the criteria.

One where we publish words that sound like us, but never passed through us.

One where human interactions become so optimized that the human disappears.

That is the danger.

Not that AI becomes too powerful.
But that we become too passive.

So here are 𝐬𝐢𝐱 𝐩𝐫𝐢𝐧𝐜𝐢𝐩𝐥𝐞𝐬 I want to practice:

1. 𝙐𝙨𝙚 𝘼𝙄 𝙖𝙛𝙩𝙚𝙧 𝙛𝙞𝙧𝙨𝙩 𝙩𝙝𝙤𝙪𝙜𝙝𝙩, 𝙣𝙤𝙩 𝙗𝙚𝙛𝙤𝙧𝙚.
Before asking AI to write, decide, or summarize, form your own rough view first. Even if it is messy. Especially if it is messy.

2. 𝘿𝙤 𝙣𝙤𝙩 𝙤𝙪𝙩𝙨𝙤𝙪𝙧𝙘𝙚 𝙩𝙝𝙚 𝙛𝙖𝙘𝙪𝙡𝙩𝙞𝙚𝙨 𝙮𝙤𝙪 𝙬𝙖𝙣𝙩 𝙩𝙤 𝙠𝙚𝙚𝙥.
Use AI to extend judgment, memory, creativity, and learning. Do not use it to avoid exercising them.

3. 𝙎𝙚𝙩 𝙘𝙧𝙞𝙩𝙚𝙧𝙞𝙖 𝙗𝙚𝙛𝙤𝙧𝙚 𝙧𝙚𝙘𝙤𝙢𝙢𝙚𝙣𝙙𝙖𝙩𝙞𝙤𝙣𝙨.
Never accept an AI-assisted decision unless you can explain the criteria, the tradeoff, the uncertainty, and what would change your mind.

4. 𝙈𝙖𝙠𝙚 𝘼𝙄 𝙩𝙝𝙚 𝙥𝙧𝙤𝙨𝙚𝙘𝙪𝙩𝙤𝙧 𝙗𝙚𝙛𝙤𝙧𝙚 𝙩𝙝𝙚 𝙖𝙙𝙫𝙤𝙘𝙖𝙩𝙚.
When you already want a certain answer, ask AI to challenge it first. Otherwise it may become an eloquent lawyer for your bias.

5. 𝙊𝙥𝙩𝙞𝙢𝙞𝙯𝙚 𝙘𝙤𝙣𝙙𝙞𝙩𝙞𝙤𝙣𝙨, 𝙣𝙤𝙩 𝙥𝙚𝙤𝙥𝙡𝙚.
Use AI to reduce friction, improve access, remove drudgery, and support better choices. But resist using AI to make people more predictable, compliant, or manageable.

6. 𝙆𝙚𝙚𝙥 𝙝𝙪𝙢𝙖𝙣𝙨 𝙝𝙪𝙢𝙖𝙣.
When dealing with people, efficiency must remain subordinate to recognition. A person is not a workflow. Not a metric. Not a conversion target.

For me, the test is straightforward:

Does this use of AI increase or decrease my ability to 𝐮𝐧𝐝𝐞𝐫𝐬𝐭𝐚𝐧𝐝, 𝐜𝐡𝐨𝐨𝐬𝐞, 𝐫𝐞𝐟𝐮𝐬𝐞, 𝐚𝐧𝐝 𝐚𝐜𝐭?

If it expands meaningful human choice, it is probably human-centered.

If it makes people easier to manage but less able to think, it is not.

AI may become one of the greatest blessings of our time.
But only if we use it to strengthen the human capacities we most need.
Not quietly surrender them.

#AIAugmentedHuman #AIEthics #HumanAgency #ResponsibleAI #FutureOfWork

Contact

bruno.gentil@sherpaconsultingasia.com

© 2026. All rights reserved.