Eureka || FreeGym

Democratize Fitness
@mutant1643
@thestrongdoc
@tanmaywizard

[Eureka || FreeGym @FreeGymAI]

Eureka Engine by FreeGym - Coming Soon

We ended up building something to coach ourselves because...

we simply couldn't keep up with our own health while developing the platform, coaching clients, staying at the bleeding edge of AI, creating content, and staying current on health, fitness, and medical research.

Not anymore.

And I have a feeling that what started as a solution for us will eventually help people all over the world take better care of their health.

Health and fitness are about to be democratized by real practitioners.

@mutant1879 @thestrongdoc

[Abhinav @mutant1879]

Saw this post on FreeGym by @ShreyasMody while literally building the Eureka Engine.

Eureka Engine is to health and fitness what Claude Code is to software development.

Except for live coaching, we will solve everything else with the Eureka Engine.

[Dr Neha Chawla || FreeGym @thestrongdoc]

Adding the part that actually explains why a "normal" LDL can be so misleading here.

The LDL-C on a standard blood panel only shows you the cholesterol mass, i.e. the cholesterol packed inside your LDL particles. It does not count how many particles you have. Those are two different things.

When your triglycerides run high, your LDL is more likely to shift into smaller, denser particles that each carry less cholesterol. Some studies show this pattern beginning around 133 mg/dL, while most clinical labs flag triglycerides at 150+ fasting or 175+ non-fasting.

Either way, the point is the same: your LDL-C can come back looking normal while the actual number of atherogenic particles in your blood is higher than you'd think.

And it's the number of those particles getting into your artery walls that drives plaque, not how much cholesterol each one carries.

ApoB counts the particles directly. When ApoB and LDL-C disagree, the particle count usually tracks your risk better.

ApoB usually isn't on a standard panel. Ask for it.

[Abhinav @mutant1879]

We don't sell devices or blood panels.
We don't partner with anyone who does.
So we can be honest about the science of health.

[Abhinav @mutant1879]

Almost every other health company has a built-in conflict of interest. When your revenue comes from a device or a blood test, you have a direct incentive to make the science say whatever sells the product - to insist that sensor's data, or that panel's markers, truly matter.

We carry no such conflict of interest. We don't sell devices or blood tests. And we maintain FreeGym-Wiki - an open-source, extremely rigorous repository on the science of health, where every claim is cited and every edit is tracked, open for anyone to audit or correct.

And there's a second reason, just as structural: we coach. A coach is judged by one thing - whether you actually get better. Distort the science and your client doesn't improve; a coach who can't deliver results has nothing. So we don't merely lack an incentive to bend the science - we have every incentive to get it exactly right.

That's why we can tell you a marker doesn't matter, a test isn't worth it, or a device won't help, because no sale rides on the answer, and our only real product is your progress. This is one of our biggest strengths: no conflict of interest pulling us to distort the science, and a coaching relationship that forces us to be honest about it - an honesty the others structurally can't match.

[Dr Neha Chawla || FreeGym @thestrongdoc]

Can't believe I have to write this, but I've heard it from more than one doctor now. "Nothing to worry about high triglycerides"

And if you're one of those people who's indifferent about high triglycerides, listen.

Worry about it.

High triglycerides can be one of the earliest warning signs of cardiometabolic dysfunction - higher risk of heart disease, type 2 diabetes, fatty liver.

They tend to travel with belly fat, high blood sugar, and high blood pressure too.

And your LDL-C can read completely normal while the number of harmful particles in your blood is actually high. So a normal LDL-C doesn't automatically clear you.

If your triglycerides are up, especially more than once, ask for ApoB and look at your non-HDL-C. Don't just relax because the LDL-C looked fine.

[Abhinav @mutant1879]

https://t.co/iZY8kx00pr

[Dr Neha Chawla || FreeGym @thestrongdoc]

Why get strong at all, if you're not an athlete or chasing a certain look?

Because strength is one of the biggest reserves your body has. In medicine we call this Physiologic reserve. Meaning - how much your body can lose and still keep working.

A fall, an illness, surgery, aging, trauma none of them ask how you looked. They ask how much your body can withstand and still recover from.

And that reserve is built years before you need it, by people who chose to get strong while they still had the choice.

Most people treat strength as optional, right up until their body makes it urgent.

So if you don't train at all, at least stop assuming the body you have today is the one that carries you through what's ahead

[Abhinav @mutant1879]

When someone says they're going to make your biomarkers and wearable data talk to each other, run. They don't know shit about health and fitness.

[Abhinav @mutant1879]

A beginner sees many separate facts. An expert sees fewer, deeper patterns.

[Dr Neha Chawla || FreeGym @thestrongdoc]

New section, same standard. I've started writing Biomarkers for FreeGym Wiki.

What a standard lipid panel actually tells you.
Whether LDL is really causal.
ApoB vs LDL-C.
Why fasting glucose misses early insulin resistance.
Why fasting insulin still isn't a standard test.

Open-source, every claim linked to its source.
No takes, no vibes. Just what the papers show.

712 citations. Every one earned.

[Abhinav @mutant1879]

Some people defend wearable scores the same way others defend homeopathy. It feels like a new-age form of pseudoscience.

[Abhinav @mutant1879]

Polymaths don't memorize fields. They model them. The trick isn't remembering more. It's forgetting well.

[Dr Neha Chawla || FreeGym @thestrongdoc]

Dips build a kind of pressing strength most machines can't touch.

You're pressing your full bodyweight through a deep range and holding it stable the whole way. That's functional pressing strength you can actually use.

But dips can also trash your shoulders if you do them wrong, and a lot of people do.

The problem is almost always the bottom, where the shoulder is most exposed, loaded deep and stretched in a position.

If you lack active shoulder extension, gravity is just going to tear through your connective tissue at the bottom of the rep.

You don't have to avoid the movement, but you do have to respect it.

Keep your shoulders down and packed. Only go as deep as you can actually control, and earn that depth over time.

And if your shoulder stability and mobility aren't there yet, build those first.

Get it right and few things build pressing strength like a clean dip. Just don't let your ego pick the depth.

[Abhinav @mutant1879]

It's pretty clear that wearable fans aren't reasoning from first principles.

They're taking a top-down approach, starting with the conclusion and working backward.

Try a bottom-up approach instead, and you'll see that many of these claims are just beliefs.

Your need to believe is what's being exploited.

[Dr Neha Chawla || FreeGym @thestrongdoc]

A client of ours had her baby last week. She got through the whole labour on just breathing and movement.

But the part I keep thinking about happened 8 months before that.

When she came to us for her pregnancy, she didn't come alone. Her husband signed up the same week. Both of them training, both taking it seriously, before the baby ever arrived.

People treat pregnancy like it's only the mother's job to prepare. She changes her food, does the classes, carries all of it, and he mostly watches from the side.

But a newborn is physical for both of you. The carrying, the broken sleep, the months on your feet.

Through her pregnancy, he was the one giving her counter pressure when her pelvis and lower back were aching late on.

And it doesn't stop at the birth. The food on the table, whether the parents move or sit still, all of it becomes the child's normal long before he/she gets a say in it.

So if you're planning a baby, both of you should be getting strong for it. Not just her. It's the first thing your child inherits.

[Dr Neha Chawla || FreeGym @thestrongdoc]

The pull-up is the one exercise I see people quit on the fastest. They try a few, can't do them, decide they're just "not a pull-up person," and move on.

But the thing is,

strength is a skill.

A lot of it is just your nervous system learning the exact movement, and your body only gets good at what you actually practise.

That's why you can lift heavy on a barbell and still struggle to pull your own bodyweight up. It's a different skill, and it takes time.

So no, you're not bad at pull-ups. You just haven't trained them yet.

Start with regressions like slow negatives or banded reps, and try grease the groove method (thread below).

That's exactly how a skill gets built. One day the bodyweight that felt impossible starts moving.

Genuinely one of the best feelings in the gym.

[Dr Neha Chawla || FreeGym @thestrongdoc]

You could be dealing with any number of health issues, and lifting weights is still one of the most powerful things you can do about it.

I see it in my clients all the time. Many come in with compromised cardio-metabolic health, and even introducing a little bit of lifting changes how they feel in a way they didn't expect.

This is one of the most underappreciated things in health, and almost nobody pays attention to it, especially for their elderly loved ones whose health is already fragile. The day they start lifting, even gently, it starts working for them.

You can follow the best nutrition and do everything else right, but if strength training is missing, you're leaving a lot on the table. It's the thing that makes everything else actually work.

[Abhinav @mutant1879]

Most people building in the health-AI consumer space are building around folklores and bro science. They are failing at defining the structure first.

And when the structure is wrong, AI will confidently answer the wrong question. A precise answer to the wrong question is worse than useless. It is misleading.

The reason they are failing at defining the structure first is that they have not seen enough recurring problems and found the recurring solutions.

In health, the hard part is not generating advice. The hard part is knowing what question should be answered in the first place.

That requires years of exposure to recurring human problems, recurring failure patterns, and recurring intervention classes.

The only way to get that is to actively coach all different kinds of people with different pain points for years, while also constantly testing AI models to see where they go wrong.

The builder of these technologies cannot just sit behind a screen and build from assumptions. They must actively coach people.

In this sense, the best builders in health-AI are also the forward deployed engineers. They are close to the user, close to the mess, close to the real constraints, and close to the repeated failure patterns.

They are not just developers. They are developers, coaches, and forward deployed engineers at the same time.

Either most builders are too young to have come across that many recurring problems, or too inexperienced to recognize the recurring solutions.

That is why so much consumer health-AI ends up becoming polished bro science.

[Abhinav @mutant1879]

Yesterday a post about someone hooking their WHOOP to their work calendar to find which coworker stresses them the most went viral.

Two things bothered me.

One - Claude Fable. For all the noise about how powerful this model is, it just did what the user asked. Reverse-engineered WHOOP, built the stress leaderboard, and shipped the chart. It didn't push back that a wrist PPG sensor cannot tell you that. A model this capable should be the one flagging it. It didn't.

Two - the comments. Almost everyone called it brilliant. People want to sell it, productize it, point it at their kids. Not a single top reply asked whether the underlying measurement supports the claim on the chart.

This is what scientific illiteracy looks like in 2026.

[Abhinav @mutant1879]

Bro thinks WHOOP can measure stress. A moment of silence.

[Dr Neha Chawla || FreeGym @thestrongdoc]

A normal fasting glucose on labs can fool a lot of people. They see that number, assume their blood sugar is sorted, and move on. But you can be insulin resistant with a perfectly normal glucose. Thanks to your pancreas, it's working overtime, making more and more insulin to keep it looking fine. Until one day it can't keep up, and the glucose finally climbs. And this can go on for 10 years before it ever shows up on a test.