Thread by Steph Smith

twitter.com
1 Recommender
1 Mention
Recommend
Post
Save
Complete
Collect
Thread
In 1994, Jeff Bezos famously spotted a stat that made him leave his high-paying PE job to start Amazon:

💡 The Internet was growing 2300% per year.

What are the generation-defining stats of today?

I'll post a few to kick things off...
💡 The cost of mapping a genome has fallen ~100,000% over the last 15 years.

If that doesn't speak to you, we're talking about $100m down to <$1k – faster than Moore's law.
💡 College tuition has outpaced how much people earn by almost 10x.

I'll just let this one speak for itself.

h/t www.reddit.com/r/dataisbeautiful/comments/r2s0aw/oc_college_tuition_has_outpaced_how_much_people/
💡 Cyber insurance loss ratios have increased for years, despite exponentially growing rates.

Loss ratio = (costs + claim payments)/premiums

In other words, cyber security is becoming an exponential issue.
💡 The percentage of men under 30 not having sex has tripled in the last ~decade.

I'm not sure what percentage of those men are now having sex in the "metaverse".
💡 The @nytimes estimates that there will be a shortage of 300k semiconductor workers by 2025.

Most emerging tech depends on semiconductors, whether we're talking EVs, AI, gaming...

And making chips is highly complex – taking up to 1500 individual steps and up to 6 mos.
💡 In 1942, 8 hours of sleep was the norm. Today, people get an average of 6.8.

New research and the $411B toll this costs the US each year has people finally "waking up" to its importance.

Think the meditation industry is big?

Try a TAM of literally every person on earth.
💡 The electronics system was 5% of car costs in 1970.

It's expected to be 50% by 2030.

Self-driving is not going to be the only advancement when it comes to "car computers"...

Exhibit A: The Ford F-150 has 150m lines of code!

Look out for car malware, app stores, etc
💡 Here's one that I can't believe it took folks a pandemic to realize:

As of 2020, Americans saved over 60m commute hours per day with remote work.

Per day!

Assuming 5-day weeks, that's ~16.3B (billion!) hours saved per year, equivalent to:
• 1.9m years
• 23k lifetimes
Lots of great comments! Will comb through them and share the best soon.

Gotta catch 🌅 first.

In the meantime, if you like this stuff...

1. Bookmark this 🧵. I'll add to it!
2. Give my podcast @sydlis a listen. You'll like this ep on tech ppl feared:

keeplearning.buzzsprout.com/987811/9788436-43-technology-that-people-feared
@sydlis I'm back with more on solar, lab-grown meat, crypto, eldercare... but first:

💡 The stat everyone knows, but not everyone FEELS (yet!)

19 of the hottest years on record were in the last 20.

A concrete example? The # of days 70⁰F+ in NYC:
• 2000: 111
• 2020: 141
• 2050: 184
💡 The OG lab-grown burger costed $330k in 2013.

Now, 1lb of lab-grown chicken costs $7.70, down from $18 mid-2021.

And 1 kg of lab-grown shrimp?
• 2019: $7398
• 2021: $739
• 2022: $37
• 2026-28: $4

Yet another trend that's challenging Moore's law.


💡 Global fertility has halved in ~50 years.

This might not sound meaningful until you realize that 10k baby boomers turn 65 every day in the US alone.

We can only hope that all these robots we're building will lend a "hand".

h/t to @heykahn @healthwealthgen for calling out
@heykahn @healthwealthgen 💡 A new solar plant is now 3x cheaper than a new coal plant.

Lower costs + increased capacity = a sunny, economically viable future

☀️☀️☀️☀️☀️

h/t to @azeem @ExponentialView @ramez @OurWorldInData @tsungxu for their excellent coverage on this topic
💡 Get this: the fastest growing 10-yo company isn't in software. It's a lithium-ion battery company: CATL.

Like semiconductors, batteries are a fundamental layer to many other technologies.

And yep, you guessed it: prices have decreased exponentially.


💡 Much of the coming decades will be defined by the growing global "middle class", ascending from 3.6B (2018) to 5.5B (2030).

Examples:

🇮🇳 India went from 70% to 93% electricity access from 2007 to 2017.
🌍 Africa reps 3% of global GDP, despite 17% of the pop'n. h/t @iaboyeji
Olympics coverage is down 55% (vs 2018)
Disney channel is down 88% (vs 2014)
Nickelodeon is down 75% (vs 2014)

You know what's not down..? Gaming.

💡 Esports viewership outpaces some major sports leagues, while prize money grew 4000x since 1998.


You didn't think I'd get through this entire thread w/o crypto, did you?

The maximalists must be sweating. 😅

💡 Crypto adoption closely mirrors internet adoption minus a few decades. This trajectory predicts 1B users by 2027.

Others peg as soon as 2023.

h/t @AzazelAyers
I'll end (for now) with a stat shared by @JustinQuda:

We pick up our phones 58x per day. 58!!!

Here's your reminder to:

|￣￣￣￣￣￣￣￣￣￣￣￣￣￣|
      Get some sunshine today
|＿＿＿＿＿＿＿＿＿＿＿＿＿＿|
                    \ (•◡•) /
                      \      /
And if you learned a thing or two...

1️⃣ Comment with a trend I missed! I'll keep adding.
2️⃣ Give @sydlis a listen. Ep 35 is our most popular.
3️⃣ Follow me @stephsmithio! Uncovering trends is literally my job, so there's more where that came from.

See you in the future! 🚀
@sydlis Wow. This has blown up.

In 24h, it's reached >5m people!

Perhaps the craziest stat is still the power of the Internet. Each min there are:

• 148k Slacks
• 575k Tweets
• 5.7m Google searches
• 694k hours YT streamed

PS: I wrote a book on this at doingcontentright.com 👀
@sydlis Many folks asking where to get more data!

Lots of great newsletters:
📊 @ExponentialView
📊 @OurWorldInData
📊 @NumlockAM
📊 @chartrdaily
📊 @TheHustle

Also:
🎧 @SYDLIS (w @calvin_rosser)
⬆️ @DataIsBeautiful
🛠 @f_l_o_u_r_i_s_h
🐦 @PlanMaestro

What's missing from this list?
Some stats feel "otherworldly".

This one takes that sentiment literally.

💡 The cost to launch a kg into low orbit has gone down – you guessed it – exponentially.

Down 10x in the decade, Starship projects another order of magnitude decrease.

h/t @ShepJoel @guyfosel @lizvstein
Perhaps the only thing harder than getting to space is grasping huge numbers.

Like the scale of AI models.

💡 The # of training parameters quickly went from 175m (GPT-3) in 2020 to 1.75T (Wu Dao 2.0)

That's 1,750,000,000,000.

AI compute power has also grown exponentially.
Mentions
See All