- Modern Chaos
- Posts
- MC.91: Your AI Co-Worker Has Arrived (And It's Not Here to Replace You)
MC.91: Your AI Co-Worker Has Arrived (And It's Not Here to Replace You)
It is just a matter of time for gossip at the coffee machine to arrive.
Olivier Legris
September 18, 2025
Hello,
This week, I want to share some fascinating insights from OpenAI's massive study on how people actually use ChatGPT. The findings challenge the dominant narrative about AI replacing workers and reveal something far more nuanced: AI as a collaborative partner.
The Real Numbers Behind AI Usage
OpenAI just released comprehensive data analyzing over 2.5 billion daily messages from 700 million users, nearly 10% of the world's adult population.
Here's what caught my attention: only 27% of ChatGPT usage is work-related, and that percentage has been declining over time.
But here's the kicker, when people do use AI for work, they're not asking it to replace them. They're asking it to help them think better.
By July 2025, 18 billion messages were being sent each week by 700 million users, representing around 10% of the global adult population. For a new technology, this speed of global diffusion has no precedent
AI as Your Research Assistant, Not Your Replacement
The study reveals three dominant use cases that account for nearly 80% of all interactions:
Practical Guidance (29%) - "How should I approach this problem?"
Seeking Information (24%) - "What do I need to know about X?"
Writing (24%) - "Help me improve this draft"
Notice the pattern? These aren't "do my job for me" requests. They're "help me do my job better" conversations.
The researchers introduced a brilliant framework: Asking vs. Doing vs. Expressing. About 49% of all messages are "Asking", seeking advice and information to make better decisions.
“Only” 40% are "Doing", requesting actual task completion.
Even more telling: Asking messages consistently receive higher satisfaction ratings and are growing faster than Doing messages.
The Decision Support Revolution
What we're witnessing isn't job automation, it's decision augmentation. The study found that AI usage maps heavily to two core work activities:
Obtaining, documenting, and interpreting information
Making decisions, giving advice, solving problems, and thinking creatively
This pattern holds across virtually every profession studied, from management to engineering to healthcare.
We argue that ChatGPT likely improves worker output by providing decision support, which is especially important in knowledge-intensive jobs where better decision-making increases productivity.
The most intriguing finding? Writing dominates work usage, accounting for 42% of work-related messages. But here's the twist: about two-thirds of writing requests aren't asking AI to create from scratch, they're asking it to modify existing text.
This mirrors what we see at Alter. The real magic happens when humans and AI collaborate: humans provide context, judgment, and creativity; AI provides speed, scale, and refinement of the rough edges.
Final thoughts
Perhaps the most encouraging finding: AI excels at the collaborative aspects of work that humans value most. It's becoming the research assistant who never gets tired, the writing partner who's always available, and the brainstorming buddy who brings fresh perspectives.
The future isn't human vs. machine, it's human with machine. And the data shows that future is already here.
Cheers,
Olivier
Like this newsletter? Forward it to a friend and have them sign up here.
Until next Thursday 🎉
1 Claude does show some code but hides the bigger picture.
Reply