I NEVER Take Notes Anymore - My Full AI System As A Doctor

Academy: stories and events
54 followers
4 Video Views·Sep 29, 2026
I used to swear by taking zero notes in medical school, but residency changed the stakes. In this video I break down the 3-phase, AI-powered system (capture, summarize, organize) I built to put my learning on autopilot, and the lesson from neuroscientist Dr. David Eagleman on why removing all friction can hurt retention.
AI Generated·May contain errors
Video Summary
AI GeneratedThe speaker, a medical resident, shares an AI-powered \"autopilot\" note-taking system designed to combat the forgetting curve without sacrificing presence during learning. By distinguishing between \"vicious friction\" (wasteful effort) and \"virtuous friction\" (essential cognitive struggle), he balances automated capture with intentional review to ensure deep learning.
I've always been terrible at taking notes. Not just bad, genuinely embarrassingly terrible.
If you search this channel, you'll actually find a video that I made called
How I Take Zero Notes in Medical School, and for a long time I stood behind every word of it.
My argument was that when I'm writing something down, I'm not actually listening.
My attention is split, and split attention is the enemy of real learning.
I told myself that I was protecting my focus, and honestly I still believe this,
but then I got to residency. Then I entered the most knowledge-intensive training period
of my entire life, learning things that will define my career for the next 30 years,
and I was letting most of it disappear.
But I mean, when you work 80 hours a week in a hospital,
and also have a full-time job at home as a husband and a father,
finding time to take notes and actually organize them so that they can serve me in the future
seemed impossible. But in this age of AI, I knew that there was a solution to this problem.
And if you resonate with wanting your note-taking to basically happen on autopilot,
I want to show you exactly what the solution that I found is.
Now, I want to reiterate that I still believe parts of what I said in that old video.
When I'm writing something down, I'm not actually listening. My attention is split.
I absorb more when I'm fully present in the moment.
But residency changed the stakes.
The psychologist Herman Ebbinghaus documented what's now called the forgetting curve,
defining that without any attempt to review or capture new information,
we forget roughly 70% of what we learned within 24 hours. Literally one day.
That number keeps climbing over the following week.
I don't think there's a single physician, entrepreneur,
creator, or really anyone alive who will look back during periods of intense learning
and say, you know what, I'm glad I took fewer notes.
The regret almost always runs in the other direction.
The things I'm learning right now are the things that I don't just need to learn,
but I also need to remember so my future self can benefit.
So I built a system to basically make this process run on autopilot.
It's broken up into three phases.
Capture, summarize, and organize.
Each phase solves a different part of the problem,
and they're designed to connect seamlessly
so that the whole thing runs without me having to think hard about it.
The whole system starts with one question.
How do I get knowledge out of the moment and into a document
without stopping to write or type anything?
The first tool I use is called Granola.
Granola is a meeting and conference app,
but what makes it different from anything else
is that it actually understands medical terminology
and can follow everything from a quick conversation with a mentor
to a lecture or a conference.
It doesn't produce a garbled transcript full of errors.
It produces a real summary, organized, accurate, and actually useful.
What I do before any conference is simple.
Open Granola, click one button, and then I focus on what is being said.
When it's over, the summary is already there.
One day, I actually forgot to stop the recording when a conference ended,
and Granola kept running in my back pocket
all the way through the first case of the day,
which was a distal bicep repair in the operating room.
I was scrubbed in, assisting my attending,
and there was not a moment when I could have just stopped
and picked up a pen to take notes.
But when I scrubbed out and looked at my phone, everything was there.
Every teaching point that she made during the case,
every decision explained in real time.
All of it captured in a setting
where it would have been physically impossible to write a single word.
That was the moment that I understood the power
of automatically capturing moments of learning.
I'm sure there are tons of tools out there just like Granola.
Granola is just the one that works for me,
but the principle of capturing your knowledge is critical,
and finding as many opportunities to do so throughout the day
will serve your future self so well.
The second tool I use in this capture phase is Whisper.
Whisper is a dictation app,
and it's the best one that I've used by far.
It works on my computer and my phone,
and it's what I use to talk directly to AI tools like Cloud,
ChatGPT, or Gemini without touching my keyboard,
as well as my notes when I want to add something manually.
When I need to add a thought to my notes at the end of a shift,
I don't type it because I know how hard that is for me.
I just talk, and the accuracy is remarkable,
even with medical language, and the time savings are real.
If you're serious about removing the friction of typing from your workflow,
you need to get a dictation app,
and Whisper is the one that I would recommend.
Once the capture phase has done its job,
I need to actually do something with all of that content,
and that's where Cloud comes in.
I've set up a workflow where granola notes from the day
automatically get passed to Cloud,
which summarizes them into a clean, organized format,
and sends them as an email draft directly to my Gmail.
I don't have to copy anything.
I don't have to format anything.
Just by the end of the day,
there's an email sitting in my drafts
with a readable summary of everything that was captured.
I can scan through it,
and if something needs a small adjustment, I adjust it.
If it's good, I move forward.
What I love about this step
is that it creates a moment of low-stakes review.
I'm not sitting down to study.
I'm just glancing at what my day captured,
and more often than I expect,
something in that summary will catch my attention
in a way that it didn't in the moment,
and this is my initial review of that material
so that I don't fall victim
to that ebbing house forgetting curve that we talked about.
Now, this is where the system becomes something
that will serve me for the rest of my career.
I've set up my Google Drive with folders
organized by every major topic in my work,
training to become an orthopedic surgeon,
from sports to shoulder, elbow, trauma,
hand, spine, and so on,
and inside of each folder,
I have one running document per topic,
not a new note every time I learn something,
one living document
that gets richer as residency goes on,
and by the time that I'm done,
these aren't just notes.
They're a resource that I built
from real cases and real teaching,
organized and searchable.
Here's how a new note gets in.
I copy the summary from my Gmail draft,
paste it into the relevant Google document,
and then Gemini,
which is integrated directly into Google Docs,
takes over.
I tell it to find where in the existing document
these new notes belong
and incorporate them there,
and it does it within seconds.
I can talk directly to Gemini within the document,
of course, using Whisper,
make any further changes easily.
I don't have to read the entire document
to figure out where something fits.
It handles the organization
so that I can keep moving.
Now, quick pause to introduce you
to today's sponsor
who truly belongs in this video.
I just walked you through
how I use AI inside Google Docs
to reorganize my clinical notes,
but there's a whole other category of writing
that happens in residency
that's completely different from note-taking.
Case reports, research write-ups,
presentations for conference,
that kind of writing
doesn't just need organization.
It needs structure, flow, and citations,
and it takes real time.
That's where Genie.ai comes in.
Genie.ai is an AI writing assistant
built specifically for academic and research writing
and is trusted by over 6 million academics worldwide,
including people at Stanford, UPenn,
Cambridge, and Oxford.
That's not a coincidence.
General AI tools like ChatGPT
are great for a lot of things,
but when it comes to research writing,
they have real limitations.
They can hallucinate sources,
they can produce writing
that doesn't reflect your voice,
and depending on your institution's policies,
using them to draft academic work
can put you in an ethically gray area.
But Genie.ai was built with those problems in mind.
It assists your writing rather than replacing it,
which means the idea stays yours,
the voice stays yours,
and the citations it generates are real and verifiable.
The features are genuinely useful.
It helps you draft and expand your writing
in real time inside your documents.
It builds structure from a prompt,
it sharpens a weak paragraph,
and can pull in properly formatted citations
without you having to hunt them down manually.
And one of their newest features
is truly a game changer.
An AI peer review tool.
Before you submit a manuscript,
Genie.ai can score it
against an actual peer review rubric.
It can surface weak claims,
flag unsupported arguments,
and actually gives you a section-by-section feedback
the way that a reviewer would.
This kind of pre-submission pressure test
used to require finding a willing colleague
and waiting days,
but now it literally just takes seconds.
Use the link in the description below
to get 20% off because it's worth it.
Okay, let's get back to the video.
Now, I thought that the lesson
from building this system would be simple.
AI makes everything easier.
Remove all friction of the note-writing process,
problem solved.
But eventually, I noticed something
that I was not expecting.
The notes were there,
the summaries were accurate,
but I felt like I was watching my learning
happen from a distance.
I was capturing everything,
which was good, like we talked about,
but the processing wasn't there.
And then I listened to a Diary of a CEO episode
featuring Dr. David Engelman,
a neuroscientist at Stanford
who studies brain plasticity.
And something he said reframed
how I was thinking about all of this.
The idea is that not all friction is the same.
There's vicious friction
and there's virtuous friction.
Vicious friction costs you time and energy
and gives nothing back.
Having to sit down at 10 p.m. after a long shift
and manually type everything
that I learned for that day,
that is vicious friction.
It doesn't make me learn better.
It doesn't help me.
It just sucks up time.
That's the friction that this system
is designed to eliminate.
But there's also virtuous friction.
And virtuous friction is actually
where a lot of the learning lives.
It's the time it takes to sit with an idea,
to notice that you can't fully explain something,
or to fill the gap in your understanding
and then go looking for what fills it.
That process, slow and a little uncomfortable
and usually humbling, is not inefficiency.
That's actually the process of learning
and what builds real knowledge.
When I removed all of the friction
from my note-taking,
I accidentally removed some of the virtuous kind too.
I was going too fast to know what I didn't know.
So now I'm trying to add one thing back.
Every evening, I try to read through those notes
that are captured during the day.
Not to reorganize them right away
or copy them anywhere new,
just to read them with the goal
of finding gaps in my knowledge.
And whenever I find something
that I can't fully explain,
something that prompts a question,
I look it up and I add to my notes manually.
That's the principle that I want to leave you with.
The goal isn't to remove all friction from your life.
It's to become ruthless about identifying
which friction is vicious and eliminate it,
and which friction is virtuous and protect it.
AI is an extraordinary tool for the first part,
but the second part still has to come from you.
One more thing before I let you go,
the AI landscape is moving incredibly fast.
The tools that I'm using today
may look different six months from now
or six weeks from now at this rate,
and trying to track all of this on your own
is a losing battle.
This is one of the reasons I'm really glad
that our Teach As You Learn community exists.
We host regular events
and we're going to start talking a lot
about new AI tools and strategies
for studying, note-taking,
and professional development in medicine.
It's a group of people
who are all wanting to get the most
out of their professional careers
and are sharing what they learn.
If you want to stay ahead of what's available
and be in a community with others
who are doing the same,
the link to join is in the description below.
It's currently less than $5 a month
and the conversations in there
are definitely worth it.
And if you want to take advantage
of one of the most valuable,
virtuous friction activities on the planet,
reading,
and want to know the 10 books
that I recommend every human being read,
check out this video here.
I can promise you that these books
will completely change your life.
And now you even have a system
for taking notes
on the lessons that you learn.
Keep evolving
and I'll see you guys in the next one.
