“No man is an island, entire of itself”. John Donne, Meditation XVII

I want to know what’s going on in the world. I’m a human; I’m interested in what other humans are up to; I value them, care about their triumphs and mourn their deaths.

But:

• There’s far too much going on for me to keep track of all of it
• I think that some parts of what’s going are likely far more important than others
• I don’t think that regular news providers are picking the important bits to report on

I would really like there to be a scope sensitive news provider which was making a good faith attempt to report on the things which most matter in the world (see digression one for more thoughts on this). But as far as I know, this doesn’t exist (see digression two for some cool things which do exist).

In the absence of such a provider, I’ve spent a small amount of time trying to find out some basic context on what happens in the world on the average day. I think of this as a bit like a cheat sheet: some information to have in the back of my mind when reading whatever regular news stories are coming at me, to ground me in something that feels a bit closer to what’s actually going on.

Methodological notes

• I picked the scale of a day because it felt easiest to imagine. I think that weeks, months and years would all be interesting scales to have something like this for. (For decades and centuries, I mostly think you should just read most important century, and I strongly recommend doing so.)
• The numbers I’ve used are rough, but I think they are ballpark right, and a big improvement on no numbers.
  • Please read all of the numbers as having a ‘~’ in front of them.
  • I used the most recent years I could find, but this varied (2014-2022).
  • I mostly pulled numbers from Our World In Data, and sometimes other places.
  • I’ve only put links into the text directly where the number in the text is an annual figure that you can find directly in the original source. For everything else, I think looking at the spreadsheet of my workings makes more sense than going straight to the original source.
  • I’ve included a bunch of charts to give a sense of trends over time. The charts are all in years, so the numbers they are tracking on the y axis are different to the daily numbers I give in the text.
• I picked some things that happen on an average day that I care about and think are important. I expect there are big gaps here, particularly where it comes to technology.^[1] I’d love for people to suggest things which should be added to this post.

Headline version

Things that happen on an average day:

• Almost 170,000 people die.
• 380,000 babies are born.
• Humans kill 800,000 cows, 1.6 million sheep, 4 million pigs, 200 million chickens, and 300 million farmed fish.
• $263 billion is spent.
• 95.4 million tonnes of CO2 are emitted.
• Warning: these numbers are ~nonsense
  • 898,271 people fall in love.
  • 628,350,575 people cry.

Things which remain the case on an average day:

• 8 billion humans are alive.
• A quadrillion fish are alive (a thousand trillions), 550 billion wild mammals, and 100 billion wild birds.
• There are over 9,000 stockpiled nuclear weapons.
• There are around 56 active state-based conflicts.

Spreadsheet version here.

Things that happen on an average day

Almost 170,000 people die.

Since 1950, this number has been up and down, at around 140,000 deaths per day. Tomorrow, the number of deaths will be more than today, by around 6 deaths - and this number will keep growing every day for the next 80 years or so (nearly 30,000 days). On the average day in 2100, around twice as many people will die as did today.

15,000 of these people are under 5 years old.

5,000 of them die from diseases which are preventable by vaccines.

380,000 babies are born.

This has been the case every day since around 2010 - so nearly 4,500 days - and looks set to remain the case every day till around 2050 - so another 10,000 days.

Probably around 15,000 of these babies will die in the next 5 years.

Looking just at life expectancy at birth estimates, these babies will live to the year 2093, on average.^[2]

Humans kill 800,000 cows, 1.6 million sheep, 4 million pigs, 200 million chickens, and 300 million farmed fish.

$263 billion is spent.

That’s double what was spent each day in 2005, just over 15 years ago.^[3]

$26 billion is spent on health, $10 billion on education, $7 billion on R&D, $6 billion on the military, $1 billion on AI^[4].

[Flagging that this leaves over $200 billion spent on other stuff which I haven’t found numbers for.]

95.4 million tonnes of CO2 are emitted.

Warning: these numbers are ~nonsense

They are nonsense in various ways: they each pull on one random paper I found in a popular article, each of the papers is a small sample size study, neither falling in love nor crying are clearly defined activities…

See here for nonsense workings, but it’s probably not worth it.

892,657 people fall in love.

628,350,575 people cry.

Things which remain the case on an average day

8 billion humans are alive.

3.3 billion of these people go to work.

Around 990 million children go to school.

277 million of these people have depression.

A quadrillion fish are alive (a thousand trillions), 550 billion wild mammals, and 100 billion wild birds.

Annually in farms, there are around 26 billion poultry, 1.5 billion cows, 1.2 billion sheep, 1 billion goats, and 1 billion pigs. I don’t think it makes sense to say that on a given day there are, say, 26 billion poultry alive though, given death rates in farms. You’d need to do more stats to get an estimate of the number of poultry birds alive right now.

There are over 9,000 stockpiled nuclear weapons.

And it's been roughly this way for the past 10 years - so the last 3,650 days.

There are around 56 active state-based conflicts.

As of 2020. Each day, around 135 people are killed in these conflicts.

The number of conflicts has been going up since ~1950; but deaths in state-based conflicts have declined a lot.

Thanks to my modern studies teacher who helped me set up a school newspaper, Planet Press, when I was 14; to Nicole Ross and Max Dalton for helpful conversations around this post; to Sam for inspiration on one of the nonsense numbers; and to Our World in Data for the amazing work they produce.

Digression 1: my desiderata for a news provider

Here is my list of desiderata for a news provider, in roughly the order of how much I want the thing:

• Scope sensitivity: a serious, good faith attempt to tell the stories that matter most to the most sentient life.
• Reasoning transparency: explicit about what is known, how, and how confidently. Evidence-based, links to sources etc.
• In context: giving relevant and ideally quantitative context for stories, including:
  • Forecasts and odds on the things in question (e.g. Metaculus is at x% that this thing will actually happen by Y date)
  • Similar stuff (e.g. Congress has passed $z billion for A. The total amount spent on A-type things each year is P, and so-and-so project that we need z+c if we want blahblah to happen)
  • Historical trends (e.g. this is double what it was 5 years ago)
• Tone: something about the tone not being persuasive/newsy even if that means it’s too nerdy/dry for a general audience. I just want someone serious to tell me honestly, in good faith, in a call-a-spade-a-spade way, what’s going on that they think might be important.

Digression 2: cool news-related providers right now

Worldometer shows live counts of various different things - births and deaths, expenditure, books published, water used, energy produced - some annually, some daily. 

Because the counts are live, it’s a much better way than this post of getting a sense of something like what happens each second or minute, and I recommend checking it out. I wrote the post anyway because I wanted to pull out slightly different numbers to worldometer. 

NB I used different data sources to them, so our numbers are presumably a bit different. (My only complaint about worldometer: they give sources but not workings, as far as I could find.)

Improve the News pulls news stories from existing providers, but gives users a series of sliders where they can express their preferences, and see the stories which reflect those preferences. The main sliders are left/right for political stance, and critical/pro for establishment stance. You can also choose to split either of those sliders, and then see the headlines split half and half across your screen, left and right or critical and pro.

In my dream world, there would be a website like this, but the sliders would be on complicated empirical and ethical questions (things like ‘what’s your discount rate’, ‘how many chicken lives is a human life worth’ etc), and the news stories being drawn upon would all be reasoning transparent and high quality.
 

1. ^{^}

    Seems like if you were only to have one number, maybe the best one to have would be something like ‘total FLOPS in the world today’. I’m not sure this is coherent or how hard it would be to estimate.

2. ^{^}

   I’m not sure how much life expectancy at birth makes sense. Like, does this factor in existential risk, and the probability of huge breakthroughs in anti-ageing? And if someone were to try to factor those things in, would they average out?

3. ^{^}

   Can someone tell me things about when I should expect the next doubling, i.e. in what year should I expect daily global spending to exceed $526 billion? Feels complicated and important; I’m ignorant about what sensible projections are and how much uncertainty there is.

4. ^{^}

   I got the R&D and AI figures from different sources; I imagine there’s some (maybe a lot of) double counting.