When I first started running over a decade ago, I quickly built up to 2-3k. 5k+ was a little harder. I have flat feet, and it became a challenge to overcome the frequent foot pain I had. Then I invested in good shoes, and built up to the occasional 10k over a few years. Soon, I considered anything below 7k to be a “short run”.

I kept pushing into trail runs, 12-13k regular runs and finally I ran my first half marathon in 2018, about 3 years after I started running regularly. I ran 3 more after that, finishing in the 2:20-2:40 range. The first 3 HMs were extremely exhausting. The stretch after 18k was particularly brutal. Blisters on my feet, heart rate rising, my glutes complaining, I had to summon every bit of willpower to trot-walk-drag my way to the finish line.

An LLM has no intuition for when a contradiction means a human misunderstood something and when it means the spec is wrong and should change.

The Bug Bash

Last night I pointed Fable at my bugs list and asked it to review and fix everything. I mentioned some of them might already be fixed, so it had to review them before fixing.

It fixed four bugs, and I verified them. A good night’s work.

Getting value out of an AI is mostly context transfer: you hand it the constraints and context it can’t figure out on its own, add enough tokens and a few prayers, and value falls out. Or that’s the hope.

The problem is humans are not very good at sending context to AI. We’re impatient, so we send half-baked instructions and expect the model to fill in the gaps. Sometimes it does a good job; sometimes it totally misses the point.

LLMs are often the first “person” we talk to when we have a health concern.

This is especially true when the problem is not acute. If you have chest pain, you probably know you should not be debating symptoms with a chatbot. But if you are trying to optimize your sleep, reduce cardiovascular risk, understand your bloodwork, improve your diet, or decide whether a wearable metric matters, an LLM can feel like the perfect first stop.

I had an interesting conversation with my 4 year old daughter this morning.

Me: We usually wash your hair today, but you are going to swimming class after that anyways, so you will wash your hair anyways after that. Washing your hair now is pretty low RoI.

Her: What is RoI?

Me: Return on Investment.

Her: What is investment?

How do you explain investment to a 4 year old?

We have spoken a lot about delayed gratification. And how kids who exhibit delayed gratification are much more successful. She sort-of gets that concept.

I believe startups are best built by high agency individuals focussed on a joint goal.

Why? The nature of a startup typically involves:

1. A market opportunity - a problem faced by some set of people that is not being solved at scale.
2. A non-obvious solution - although the founders are typically in a place where the solution seems somewhat obvious to them.

That means often the path to take is not exactly clear and the primary thing that determines a startups success is the rate of experimentation.

One of Homo Sapiens’ most powerful skills is our ability to work together and strategize on a larger scale than any other animal. It is believed that this is what helped us edge out our other hominid competitors.

Why are we able to collaborate better than other animals?

It seems we have a unique ability to communicate abstract concepts using specialized language centers in our cerebral cortex. We are able to communicate concepts like possibilities, flow of time. For example “yesterday there was a lion near the river, and I saw it go near the mountain. It is possible you might encounter it near the hill today” vs “lion near river, dont go”. This way we can share more nuanced information and benefit from each other’s knowledge and coordinate better at greater distances and at a larger scale. 

Wow, it’s been 3 years since I wrote. I want to practice putting my thoughts into words again, so here we go.

Let’s talk about shared myths.

The reason we, homo sapiens have been able to scale our societies so well (> 7B globally and almost fully interconnected) is our ability to create and believe in shared myths. It is what differentiates us from our ape ancestors, and it is the reason we have become the dominant species on earth (by being able to coordinate at unprecedented scale). This is one of the core ideas of “Sapiens” by Yuval Noah Harari.

I usually take the train to office. Here is a breakdown of the state of my mind during the journey -

• 10 min walk to station, stand on the right platform - active

• Less than 5 min waiting - mostly on auto-pilot

• 5 min train ride, mostly standing - active, looking out for the right stop

• 5 min walk to switch tracks - active

• 5 min train ride, mostly standing - active, looking out for the right stop

Sometimes, you do better when you increase difficulty.

I occasionally play Flight Commander when taking a quick break. It’s a simple game - direct aircrafts to airstrips by drawing a path and make sure the planes don’t crash into each other. Each plan landed gives you a point.

I was terrible at it, till I noticed the “»” button at the bottom. It doubles the speed of the game, technically making it twice as hard to play. I started using it to get through the easy initial stages, and slowed the game down again once there were too many planes in the air. I was still terrible at it.