Futurebraining · WritingHuman reinforcement for the AI age
Writing & research

Notes on the human side of AI.

34 entries / updated Jul 8
↓ Newest first
LinkedIn commentary

Reading notes, book margins, and article responses.

Follow on LinkedIn →
LinkedIn post · Jun 25

The AI atrophy problem

LinkedIn commentary on the AI atrophy problem: the risk that people and organizations become faster with AI while weakening the human capabilities they still need most.

The Futurebraining angle is simple: AI adoption has to strengthen judgment, focus, responsibility, expertise, and connection, not quietly outsource the practice that keeps those capacities alive.

Referenced The AI atrophy problem: How CIOs fight it
AI atrophyhuman capabilityAI readiness
LinkedIn post →
LinkedIn post · Jun 5

We are in a massive psychological and social experiment.

We are in a massive psychological and social experiment.

Uses linked to learning, creativity, organizing life, and finding direction seem less prominent. Uses linked to support, advice, entertainment, companionship, troubleshooting, and delegation are moving up.

My worry is a basic rule of the brain and body: use it or lose it.

Perhaps weaker memory, less writing practice, and less independent problem-solving will simply become part of the price of convenience. Maybe AI will compensate for what we stop practicing.

But what about outsourcing the harder human capacities: difficult conversations, patience, listening, emotional regulation, and social courage?

As a father, I would not want to run that risk with children who are still building the foundation: attention, language, memory, judgment, and social skills.
They need to build those capacities before deciding what to outsource.

The article does not break the data down by age. I really hope this pattern is mostly adult behavior.

Article: HBR, “How People Are Really Using AI in 2026” by Marc Zao-Sanders.
https://lnkd.in/evJaezbg
Referenced How People Are Really Using AI in 2026
social skillscognitive erosion
LinkedIn post →