The Human-AI Matrix
AI has hijacked every single conversation I have had this past week.
I met at least a dozen different people, and AI has magically managed to sneak into every single one of them.
With AI being so omnipresent, I couldn’t help but reflect on my own practice with the technology.
This article is an attempt to make sense of those thoughts and invite critique, feedback and perspectives.
The Human-AI Matrix
I arrived at a 2x2 matrix with:
X-axis:🧍🏻 Human and 🤖 AI
Y-axis: 🧠 Thinking and 🛠️ Doing
🧍🏻🧠 Human Thinking
This is the key differentiator between humans and other mammals - we have thoughts and are able to make sense of things. Some examples of this:
Interpreting social cues
Connecting dots across multiple unrelated things
Having a unique idea or insight
This muscle has become weaker for most of us (including myself!) over the past few years.
🧍🏻🛠️ Human Doing
This is us taking some action on that thinking. Some examples of this:
Reading and writing
Talking to someone
Walking, running
Unlike AI, humans procrastinate - quite a lot sometimes!
🤖🧠 AI Thinking
This is humans outsourcing the thinking to AI. Some examples of this:
“You are x, review this document and give feedback”
“You are an expert in y, help me build a study plan for it”
“Here is my task list, help me prioritise and plan my day”
This is what gives everyone nightmares and fuels those ‘AI will rule the world’ conspiracy theories.
🤖🛠️ AI Doing
This is humans executing using AI. Some examples of this:
“Here is the transcript, summarise it.”
“Make a visual on x concept”
“Help me refine this article”
This has been the entry point for a lot of people to start using AI.
Blurry Boundaries
The reality though, is hard to fit neatly in a 2x2.
Writing (🧍🏻🛠️ Human Doing) is impossible to do without connecting dots (🧍🏻🧠 Human Thinking)
Using AI to learn about a new topic requires both 🤖🧠 AI Thinking and 🤖🛠️ Doing
However, this 2x2 has given me a good framework to start thinking about my relationship with AI. I have started looking at each task through this lens and find myself asking which quadrants I covered.
My Preferred Path
I realised that if I start with 🤖🧠 AI Thinking - I tend to get lost in an AI spiral. It’s hard to be objective and figure out what my thoughts are vs what are AI thoughts, but feel like mine.
So here is what I’ve started doing instead -
I start all tasks with good old 🧍🏻🧠 Human Thinking.
I try asking the most simple questions - why, who, what, how, when, where.
I use pen and paper to make lists, mindmaps, doodles.
Once I get a clear idea on what I want, I ask -
How can AI help me?
Do I want it to Think for me or Do for me?
Is it an expert or an intern?
And unless I have a clear answer to these, I would not use AI.
This required A LOT of impulse control, the idea of getting an output quickly is so tempting. But over time, I’ve found myself less in an AI spiral and more in tune with the ‘Srishti in the loop’.
What is your workflow with AI? How do you keep ‘yourself in the loop’?
(P.S. No AI was disturbed while writing this post)
