Information Density and why you should always show your working

A fun thing about being alive now is how much knowledge and information is available in the form of books, blog posts, audio books, social media posts and YouTube videos to name just a few. But now more than ever there's no way we can consume all of it and so on which content we choose to focus our time and attention has the potential to drive wildly different outcomes.

In particular I believe an organisation that is driven heavily by short-form content loses it's ability both to reason about novel situations and discuss decisions constructively, leading to a lower likelihood of success for that organisation.

The crude heuristics that have consistently proved true with respect to non-fiction for me have been:

Valuable from highest to least information dense:

Books, with the value in popular business books being disproportionately in the first 1/3rd
Podcasts of over 1 hour with ~2 hours being the sweet spot (On 1.5x that's ~1.3 hours listening time)
High quality summaries of books (with a reading time of 15-45 minutes)
Blog posts with a reading time of > 5 minutes
Current affairs magazines published weekly or monthly (with an average article reading time of > 5 minutes per article)

Close to zero information density e.g. rarely valuable if my objective is learning:

Any social media post (especially if it takes less than 5 minutes to read)
All daily news (newspapers, websites, TV, daily newsletters etc)
Most podcasts of less than 30 minutes

Reading times are my own so YYMV

These are crude heuristics. I've read some terrible books, come across some podcasts that pack an incredible amount into twenty minutes and read some truly insightful twitter threads.

But on average, the vast majority of useful information I've learned in the last 15 years was from books, longer podcasts and longer blog posts. From the perspective purely of learning, I don't think daily news or social media has ever - for me - created anything I would count as genuine knowledge.

I think the common factor which drives the above categorisation may be the ability to convey "why" rather than "what".

The shorter the content, the less space is available for explaining why the author reached that conclusion rather than simply the conclusion they reached.

Genuine knowledge allows us to take a novel situation and reason about it to reach a decision on how to act. To do this, at a minimum, we need to be able to compare the specifics of our situation to other situations we have learned about to see to what extent we believe they are similar.

If we only learn the conclusion, we cannot effectively do that.

This is why maths exams require you to show your working and attribute more marks to the working than the answer; most of the value lies in how you got to the answer, not in the answer itself.

There are very few universal laws. Conservation of energy seems close and the speed of light also looks like a tough nut to crack.

But outside of that, things are broadly complicated. And so saying "Always do X" is default unhelpful. You have to get to "In situation Y, you may want to do X if you objectives are A,B,C, this seems to be the case because...".

Short form content doesn't lend itself to achieving this. Imagine two social media posts:

A trait of high performing organisations is that they always pay back their technical debt immediately

and

Technical debt is an important lever but like all debt it must be managed carefully. There will be times when the right decision is to incur more technical debt and times where it should be paid back. One of the key drivers of this decision will be the extent to which technical debt is creating drag on engineering velocity. There will be some technical debt which it never makes sense to pay back and some which has to be paid back almost immediately.

Neither are long form and neither does a particular good job of giving the reader a clear framework for evaluating technical debt and when to pay it back.

The second post starts to hint at some of the factors that might make up a framework but is far less compelling to read.

If I was going to screenshot a Linkedin post and share it, honestly I'd be tempted to choose the first one (despite not agreeing with it!). It feels sort of compelling to read despite being almost entirely devoid of useful information and probably wrong.

As humans we like things that state simple conclusions without ambiguity because they give us a brief sensation that a complicated world is simple. That a single rule will absolve us of having to spend a lot of time understanding and making non-obvious decisions about something hard.

The lure of making that 49%/51% decision that's hanging over us a 99%/1% decision is powerful.

The dark side of this when building a startup technology business is that the more "rules" are taken from low information density sources - e.g. the sources that don't provide a framework for the why - the lower your organisations ability to reason about about novel situations from first principles becomes.

This means increasingly a business can become made up of people who feel very strongly "always do X" and "never do Y" but cannot explain why or when these - often contradictory rules - may or may not apply.

On top of this they may "believe" multiple things which in reality are applicable to different situations but because they lack the why they simply "believe" - without realising it - things which directly contradict each other, usually in proportion to how recently they were exposed to the rule or conclusion in question.

One of the easiest ways to take a conversation into defensive territory is to ask sometime "why" or generally challenge them on a conclusion that they've stated publicly but sub consciously know they cannot explain the reasoning behind.

So the more of a company's operating principles end up being derived from short-form content, the less that organisation will be able to reason about novel situations it encounters and the more it's ability to discuss decisions objectively and constructively without regressing to defensiveness will decay.

This doesn't mean organisations should discourage or become hostile to short form content. Just that they should encourage people to think of it as a jumping off point to learn more. They should also encourage respectful but firm ways of challenging anything that is presented as a rule but doesn't show the working.

To quote Richard Feynman:

"Hell, if I could explain it to the average person, it wouldn't have been worth the Nobel prize."