Are you driving a car or riding a train?

Sometimes we have flexibility: we can choose paths, stop, explore, go slow or fast as needed.

Sometimes we have speed and low overhead: things run on their own, progress is virtually automatic, and before we know it we’re at our destination, ready to go to the next station.

I’ve been thinking about the car and the train as metaphors for the tradeoff between flexibility and low-maintenance speed.

And we often can’t have a train that’s flexible and can easily leave its tracks or a car that can travel at 300km/h on a straight line while we read a book.

We HAVE to choose.

How to .. do it all?

4th of July weekend, and I traveled to meet my mentee in person. I meet my mentees every other week over Zoom, but it’s always nice to hang out live and talk about life. I had just visited one of my mentees in person for his wedding in France, and now I was hanging out with the other in a private indoor pool while watching the World Cup. Can’t complain.

He leads a busy life: exercise and dance classes, new job, AI study, reading, and being a pillar for his family while leading a social project with over a hundred volunteers and hundreds of people in need as beneficiaries.

We were talking about the challenge of excelling at many things at once: the lack of time, the struggle of preserving focus with the myriad calls for our attention, the desperation sometimes of not making “enough” progress on things, etc. How to balance it all?

That’s when I explored what things looked like in my life and came up with this metaphor of cars and trains:

It’s virtually impossible to have many things require our attention all the time and also make tangible progress. The way to keep “all the balls in the air” is to not have to juggle them all the time but instead just have balls that float for long periods of time — like balloons.

Driving many cars at once is like juggling glass balls, while keeping many trains running on time is like juggling balloons: trains were made to run from origin to destination fast and with minimal supervision.

But they only take you to one place.

On diet and training

Here’s what I’ll eat tomorrow:

• I’ll wake up and take a protein supplement and a single Brazil nut.
• A few hours later, I’ll take an early lunch with kale or arugula, carrots, beets, broccoli or cauliflower, quinoa or rice, and beans or chickpeas.
• In the afternoon, I’ll have a protein bar snack, a nori sheet, and some oat milk.
• For dinner I’ll have a similar meal to the one I had at lunch, probably switching some of the options (cauliflower instead of broccoli, etc).
• Before I go to sleep, I’ll have a protein supplement.
• If it’s a BJJ training day or a heavy Strength Training day, I’ll have a carb supplement or a fruit if it’s available 1 hour before training.

I didn’t “plan” my diet for tomorrow. It’s not a plan in the sense that I have stopped and thought about “here’s what I’ll do tomorrow.” This is just how my diet works. It’s the default unless I interfere.

Once I got this train running, the effort to have a diet that aligns with my goals is actually quite minimal. All I have to do is not mess it up!

In fact, while someone might say “wow it must have taken a lot of discipline to be vegan!” and it might be true, it will take me virtually no self-discipline to eat a vegan diet tomorrow.

And I know this because I’ve experimented with not being vegan for virtually all of the past year, and it was really, really hard for me to run my diet as a car instead of a train again.

As an omnivore a few months ago, I didn’t know what I’d eat tomorrow: Chicken breast for more protein? Picanha for the taste even though it has more saturated fat? Pizza to close the day?

My diet as an omnivore was full of decisions: I was balancing my food cravings, my training needs, my health indicators, and my social events to try and find the “best” option in each situation. I struggled to clarify my goals and define how I ate.

Counter-intuitively, flexibility was leading me to worse long-term results based on the goals I had, even if it made for potentially better case-by-case decisions.

Importantly, my current diet presumably leads me to better results on the long-term goals I have, but it says nothing about whether I have the right long-term goals to begin with.

If picanha is tastier than chickpeas (and let’s all agree that it is), then my train isn’t optimizing for taste. If pizza is the right meal to unwind a busy day, then I’m not optimizing for unwinding, and so on.

As long as I stay on my train, I’m in exploit mode instead of explore mode. All those other options that may be better for me fall by the wayside..

.. so I have to be really certain I’m on the right track.

When to drive and when to check the map

Keeping your eyes on the road is the right way to make progress, but pulling over to check the map is the right way to know you’re making progress in the “right” direction.

Keep checking the map, and you barely move.

Keep driving, and you may just be speeding the wrong way.

While it’s therefore tempting to believe the optimal way to get where you want is to alternate between driving and map-checking, driving and map-checking, there’s actually an even faster way — albeit slightly more dangerous: pick your destination a priori, set the route, take a deep breath, floor it, and well.. hope for the best.

It sounds dumb but I’m serious. While it may seem like the fastest progress is made by constant reevaluation, from a theoretical standpoint, the fastest way to get somewhere is to pick the right route and never stop.

And if you’re that certain of where you’re going, why drive at all? Take a train: it can’t leave its tracks — and that’s exactly what makes it fast.

Now, when it comes to adjusting your path, there are two very different types of adjustments:

1. I have the right destination, but am I headed towards it? (Is the train on track?)
2. Do I even have the right destination? (Are the train tracks right?)

Number 1 is about having good systems, i.e. good trains. Number 2 is about having good goals, i.e. high confidence in where you want to go.

Goals AND Systems

James Clear creates what I think is a false dichotomy when he says “forget about goals, focus on systems instead” in his bestseller Atomic Habits:

The goal in any sport is to finish with the best score, but it would be ridiculous to spend the whole game staring at the scoreboard. The only way to actually win is to get better each day. In the words of three-time Super Bowl winner Bill Walsh, “The score takes care of itself.” The same is true for other areas of life. If you want better results, then forget about setting goals. Focus on your system instead.

What do I mean by this? Are goals completely useless? Of course not. Goals are good for setting a direction, but systems are best for making progress. A handful of problems arise when you spend too much time thinking about your goals and not enough time designing your systems.

Clear, James. Atomic Habits: An Easy & Proven Way to Build Good Habits & Break Bad Ones (p. 24). Kindle Edition.

Systems are not an alternative to goals — they are a means to achieving them. You obviously can’t achieve your goals if there are no goals to achieve, and if you build systems without goals you’ll just be building effective trains that quickly take you to the “wrong” destination — or rather that will take you somewhere you presumably don’t care about. There’s no way to tune them, because you have nothing to tune them for.

Another incredibly important reason for goals is that they act as pairing indicators, exerting backpressure on initiatives you have and preventing neglect. We have limited resources to invest and, once we invest them somewhere, we by definition limit the resources available to invest elsewhere.

If all you have is a goal to squat as much as possible, you’ll be able to do heavy squats, but you may have bad health indicators, be it from accumulating fat or from anabolic steroids. If you also have a health goal, they limit each other’s progress, but also prevent each other’s neglect.

The term comes from Andy Grove’s High Output Management: pair your indicators so that both effect and counter-effect are measured — track inventory levels and the incidence of shortages — and joint monitoring keeps things in what he calls the “optimum middle ground.”

The “optimum middle ground” is actually what trying to excel at a lot of initiatives looks like in practice. Rather than a ton of peaks, it’s more like a number of pairing indicators balancing each other out given your available resources.

You can’t create this backpressure and coordinate your resource allocation without goals. Goals are the means for building indicators and deciding how to tweak your system in the first place.

And then, you build systems that most strategically use your available resources to make progress towards your goals.

So.. car or train?

Well — it depends on where you’re going, and how sure you are about it.

Cars are for exploring. Trains are for exploiting. While I’m still figuring out where I want to go, I want a car: stop, take detours, taste things, change my mind cheaply. Once I’m confident in the destination, I lay down tracks and put the thing on a train — it runs fast with minimal supervision, and my attention is freed up for whatever still needs a steering wheel.

That’s what my year of not being vegan actually was: taking my diet off the train and putting it back on a car to re-explore the terrain. Exploration done, I’m laying the tracks again.

And that’s the real answer to “how to do it all”: you can’t drive many cars at once. But you can run many trains — and drive one car at a time.

My trains and my car

Cars are flexible and easy to steer, but they require close supervision and don’t make as much progress as trains do. Right now, AI study is the closest initiative to my “car.”

I study AI virtually every evening, but I’m never quite sure what I’ll do. The other day, I had voice agents doing phone calls based on a conversation I had at work. The field is fast moving, and there’s a lot of uncertainty in what I should and shouldn’t do.

Despite investing a lot of my time in AI, I’m still balancing a number of priorities:

• I write on my blog almost daily, in the mornings.
• I’m exercising every day in the evenings, a mix of Physical Therapy, Strength Training, and BJJ, aside from rest days and deloads.
• I read books on nights and weekends.
• I have a stable and healthy diet, and track my health indicators every other month.

It’s interesting to think about these activities because coming out of my July vacation, I actually spent time exploring and changing all of them, and that impacted my consistency but helped me re-tune for my new goals.

Some of that exploration ended in real re-routing: I went back to being a strict vegan, I tried writing blog posts on evenings and weekends but settled back on mornings, and I moved Physical Therapy to the evenings, alongside the rest of my exercise.

In other words, my vacation was a station stop: the trains pulled in, I inspected the tracks, asked myself whether the destinations were still right, and re-laid some rails. Now the trains are running again — and my hands are back on one steering wheel.

Some of these stops are now in the timetable itself. On that same vacation, I decided to deload from training every week that’s a multiple of 5 — week 25, week 30, week 35, and so on. Good trains don’t just run; they have scheduled maintenance windows.

In fact, yesterday I was feeling exhausted, and only this morning did I realize: it’s week 30. I should be deloading (lol).

I suspect that’s the actual rhythm of “doing it all”: consistency in long stretches, reevaluation in concentrated bursts. Not a little bit of doubt every day, but scheduled moments where looking up is the whole point.

Which brings me back to my mentee at the pool. His question — how to balance it all? — was never really about finding more time or more discipline. It was about looking at everything on his plate and asking: which of these can become trains? What are the defaults, routines, and systems that would let each one run without a driver? And which single thing deserves the steering wheel right now?

One car on the road. Everything else on rails.

That’s how you juggle balloons.