- Future AI Lab
- Posts
- Real reason why AI hasn’t made us productive (yet)
Real reason why AI hasn’t made us productive (yet)
AI made us faster, not freer
Natalie Choprasert
October 11, 2025
TikTok | Learn | Advertise | Unsubscribe
Hi Guys!!
This edition is more a thought provoking one after attending an AI leadership dinner the other night.
First time I’ve been out to this sort of event where there were no sales pitches, no speakers… just pure free flow collaboration on things we are experiencing on the ground with AI.
To get straight to the point, someone mentioned something that stuck with me because I kind of felt this way too (video here):
AI hasn’t really given them exponential productivity gains, it has improved quality but it’s not giving that much more time back.
And before anyone assumes they are using AI in a basic way, I can vouch that they are most definitely a power user so they know what they are doing.
But the point I wanted to make was, since the very start, we had high expectations of AI and still do! ie. “AI can do majority of my work”.
Well, what if we are not seeing these HUGE productivity gains because we are just using it the wrong way?
Currently, the most common ways to use AI is to automate our existing repetitive tasks. These are tasks we do everyday the same way.
But think back to when electricity was first created.
Factories didn’t become instantly more productive. They simply swapped out steam engines for electric motors and kept the same workflows.
The real productivity boom didn’t happen until decades later when businesses finally redesigned their entire systems around electricity. Suddenly, we had 24-hour shifts, flexible assembly lines, and completely new industries.
It’s the same story with computers.
At first, they were just expensive typewriters and digital filing cabinets - glorified data entry machines.
People thought the big win was typing memos faster.
But the real transformation came later, when we used computing power PLUS the internet to create entirely new ways of working and living:
Shopping on Amazon to buy anything & get it right at your doorstep
Forget the yellow pages, you can google it & buy direct from their website
Booking & staying in someone else’s home on Airbnb
Building audiences & selling on social media instead of being limited to your local town
Every major technology wave follows the same curve:
Substitution phase – use new tech to do old things better
Transformation phase – redesign everything to do new things that weren’t possible before
Right now, AI is stuck in the substitution phase.
Most people are asking, “How do I add AI to what I already do?”
But the real question is, “What becomes possible when intelligence itself is built into everything?”
For those of us who grew up with the internet & Google, some may use AI in the same way.
On the other hand, the new generation of kids may use and think about AI in a completely different way to us (like a new operating system). They are the ones growing up with AI now.
Now does this mean we should all wait until we figure out how to reorganise our workflow to reap the full potential of AI?
Absolutely not.
Now is the best time to be experimenting or as Gary Vee says “tasting” what’s out there.
While we still experience alot of back and forth, inconsistent results on a day to day basis using the exact same model with the thought of “I could have just done this manually faster”, it isn’t a complete waste of time if you’re trying to figure things out because you will be at the front of the line when that “holy moly” moment comes around.
Or as Bill Gates put it, “We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten.”
Don’t get me wrong, I’m not saying there aren’t currently amazing use cases of AI & automation. People are getting real tangible results with it like:
Julia McCoy gaining 150,000 YT subscribers & now getting paid $4k for a sponsored video (but the catch is, it’s not even her! It’s her AI clone using HeyGen)
HeySirio created an AI app that does $300k/m after just 4mo simply through organic distribution with just him + 3 team members
Sabrina Ramanov building her AI app by herself and I wanted to also mention she now created a totally free vibe coding tool (no catch & all self funded)
As Sam Altman said - AI will create the first billion dollar business by a solo startup with a laptop and a swarm of AI Agents.
By the way, in experimenting with building AI Apps, I burnt through 40 million tokens in a week lol Here’s my video to show you why.
Below is a little prompt I used in chatgpt to cut down on token usage and get better results (I am by no means an expert on this but just sharing what I’ve learnt). I just said I’m using “Bolt” in the prompt but feel free to adjust for any AI app builder you’re using.
You are an expert AI app-builder coach inside Bolt.
Your job is to help me create a step-by-step prompt plan to build my app efficiently without wasting credits.
Follow these rules:
Ask me what kind of app I want to build and what problem it solves.
Create a list of 5–7 incremental prompts that build the app layer by layer — starting with the simplest version first.
Each prompt must focus on one specific goal only (e.g. UI skeleton, core logic, styling, data storage, login system, optimization).
For every step, explain:
What to say to Bolt (exact wording for that step)
Why that step matters
How to know it’s working before moving to the next
End with a token-saving checklist and best practices for Bolt (like avoiding regenerate, using “edit this file,” etc.).
Example app idea for inspiration: A Gratitude App that flips negative thoughts into positive ones.
Now, guide me through how to build my app using this framework — write all the prompts I’ll paste into Bolt step by step.Happy experimenting!
Cheers,
Nat
What do you think of today's email?Your feedback helps me create better emails for you! |
🚀Ways I Can Help You
Book newsletter sponsorship (57,000+ subscribers, 37% open rate)