
founder https://teapaw.org/t.co/7WJPAzA8kA (in 50+ med schools) writing at https://teapaw.org/t.co/PtVSee4FRL | lived in 4 countries | ex @heyepic
Joined January 2020
- Tweets4.6K
- Following834
- Followers540
- Likes11.2K
I just implemented this to generate 3,189 customer feedbacks for Rounds.
All of that will immediately get fed into my product to improve its development roadmap and positioning.
This would have cost thousands of dollars just a few years ago.
A toothpaste company has quietly killed the entire market research industry and nobody is talking about it.
Colgate published a paper showing you can predict real purchase intent by simply asking LLMs to roleplay customers.
And this is beyond insane.
If you ask an AI, "Rate this product from 1 to 5," it gives safe, middle-of-the-road garbage.
So researchers invented a method called Semantic Similarity Rating (SSR).
Instead of asking the AI for a number, they asked it to roleplay.
They gave the LLM a demographic profile. They showed it a product concept. And they asked it to write down its raw, unfiltered thoughts.
Then, they used a semantic model to translate those written thoughts into a numerical score.
The results are staggering.
Tested against 57 real personal care product surveys and 9,300 human responses, the synthetic consumers hit 90% of human test-retest reliability.
They perfectly mirrored how different age brackets and income levels react to price changes.
And they provided detailed, qualitative feedback that was deeper and more critical than what actual humans wrote.
This destroys the economics of traditional market research.
You don't need to wait a month to see if a product will sell.
You can simulate 1,000 hyper-targeted customer interviews overnight.
You can A/B test pricing across every demographic instantly.
18
I have like 8-10 projects going at all times
Usually the vast majority are being ignored, but then I’ll have bursts of inspiration and make months worth of progress on one of them in just a couple days
I think this is why I seem both very lazy and very productive
1
49
Rounds team members now get official letterhead confirmation and appreciation of their contributions, as well as a certificate.
This helps our volunteer team members build a strong application for ERAS residency programs & demonstrate passion for medical education.
Samples below.
36
This is an example of our admin panel at Rounds
3 levels deep
Everything within reach, nothing overwhelming
Pro tip on designing apps and using interfaces: never make your app more than 3 levels deep in terms of pages.
My favorite framework is one side bar page leading to tab-based subnavigators from where any further clicks would open a sidebar panel
Everything 3 levels deep at max
1
39
Pro tip on designing apps and using interfaces: never make your app more than 3 levels deep in terms of pages.
My favorite framework is one side bar page leading to tab-based subnavigators from where any further clicks would open a sidebar panel
Everything 3 levels deep at max
Machine’s job is to work
Man’s job is to question
This has been the pattern throughout all of human history
14
This whole Nvidia thing was just exit liquidity for the largest capital allocators to buy in on Anthropic
1
98
I used to think those headlines about the market wiping out a trillion dollars were exaggerated nonsense, since it all comes back the next day.
The thing is, plenty of people—retirees especially—actually sell that day to cover their lives, so for them the drop locks in a real loss that wouldn’t have existed otherwise.
Once you stop looking at it only from the perspective of people who can wait it out, the missing value stops sounding like a made-up number.
33
I can’t even imagine what it would be like to be Jensen Huang right now. The company he started over 30 years ago as a child from an unremarkable background, with the most humble of beginnings, was ignored for 20 of its 30 years of existence.
At best, it only drew attention from nerdy gamers seeking the best graphics cards for their latest video games. Yet, it turned out to become the company single-handedly holding up the US economy in what seems to be a new industrial revolution.
Then, in a moment’s notice, someone somewhere turns on you, decides to find ways to smear you, and out of nowhere, you’ve lost your baby.
75
Initially, for the longest time, I had a simple $10 lifetime deal, for 11 sales, with zero marketing.
To me that was sign that this was a good product worth taking seriously so I upped the price to $60 lifetime.
I knew this would mean less sales, but I also knew that that meant each sale was worth six times more now.
Lo and behold, with that price tier it made 2 sales but at $120 total that was more than all my previous 11 sales combined.
This made me rethink my strategy and affirmed that fewer sales at higher price points are better than more sales at low price points.
I also noticed it forced me to make a better product genuinely worth way more than $60 as opposed to making something mediocre just because “it’s only $10”
This taught me how incentives and prices really do drive a producer to make genuinely better solution to customer problems, because truly no one will buy your product unless you fulfill some need for them
Anyways, unfortunately, I began doubting this process and decided to lower the price back down $10 per year.
My logic is simply that I would rather first see how many medical students are willing to pay a marginal amount for a medical education platform before I check to see how many medical students are willing to pay a somewhat premium price for a medical education app.
I was doubting this change recently, but the sale yesterday was a good reminder that sometimes some things just take time.
There’s no overarching lesson here no council no teaching that I’m hoping to confer. This is simply me sharing publicly and reflecting on the different decision points, and challenges that you expect to run into when you try to build a solution to consumer problem.
At the moment, my goal continues to be to improve the value of the product by making it more useful than ever and bringing on new team members to help alleviate some of the independent development burden that I have.
I was just thinking yesterday morning that wow it’s been a while since someone has paid for Rounds Pro
Woke up this morning to a new customer
There’s gotta be a word for lovely serendipitous moments like these
I’m testing a few different pricing tiers for Rounds. Need more sales to determine which one wins.
1
1
88
I was just thinking yesterday morning that wow it’s been a while since someone has paid for Rounds Pro
Woke up this morning to a new customer
There’s gotta be a word for lovely serendipitous moments like these
I’m testing a few different pricing tiers for Rounds. Need more sales to determine which one wins.
Thinking of rebranding my Substack to an extended series titled "How It's Done," where I go deep in various different things. Research and figure out how, for example, banks make their money, how hospitals make their money, and so on and so forth, or different things where people aren't familiar with the industries.
Maybe over time, get some guests in to talk about their expertise in that as well.
Kind of like a grown-up version of "How It's Made" from back in the day on the Discovery Channel.
2
93
Realizing every single thing I've ever written down and published online follows this exact pattern
lived or observed tension → overlooked distinction → mechanism/example → objection or boundary → practical implication
It's always been a winning formula for me
1
53
I am so sick and tired of politics taking up my twitter feed
how do i get back to 2024 where my whole feed was young ambitious people building businesses and just doing other cool inspiring stuff
1
51
The books I wrote and put on Amazon make a few sales every now and then and it’s always a pleasant surprise
Free coffee today i guess
4
94
Ali Mirza retweeted
Chipotle is either running at a loss or is one of the most efficiently run businesses in America
How are you possibly still only charging $10 for this bowl in this economy
Chipotle is either running at a loss or is one of the most efficiently run businesses in America
How are you possibly still only charging $10 for this bowl in this economy
Ali Mirza retweeted
Rounds #646 1/5
🟥🟥🟩🟩🟩
✅
350 pts
Can you beat my score?
getrounds.app/daily/r/eyJkIj…
Rounds #646 1/5
🟥🟥🟩🟩🟩
✅
350 pts
Can you beat my score?
getrounds.app/daily/r/eyJkIj…









