Back To Atoms

It's November 2022. ChatGPT just launched.

You regularly scroll tech twitter so you had heard of AI and maybe even GPT 2.

You've even been to some big data and machine learning conferences over the years to see what was up and maybe see if you could use some of it in your business.

You never really found anything interesting to do with it yet.

Not that it couldn't do cool things it just wasn't inspiring you to specific actions on your current path.

But damn did GPT catch you off guard.

AI has been coming your whole life but it's always "20 years away".

It's not the scifi stuff yet exactly but something is different now.

--

This is essentially the story of a background task I've been running in my mind the last few years.

I remember reading books about the singularity and other wild visions of the future of computing when I was young.

It helped cement me as a techno optimist with a wide eyed belief that it was all coming faster than anyone around me realized.

I still feel this way but I've changed my mind a bit through hard won learnings from fighting with reality and trying to build new things.

Main learning.

It's hard!

So now I'm just as bullish I just expect it to take a bit longer than I hoped.

ChatGPT caught my timeline pessimism off guard.

For me it has always felt easy to "see around the corner" or kind of easily come up with say 50 versions of the future I expect to happen knowing what is now possible with a certain technology.

The way I experience this is when I learn about something new I automatically come up with a slightly modified fuzzy shape of the future. This generates something like 10 - 20 specific ideas and if I find one very interesting personally and have time I start digging and see if I can build something useful.

The one's I don't work on have a good habit of being built by someone else.

ChatGPT and LLM's ( Large language models ) are the first technology in my life that sufficiently "magical" that when I extrapolate I can't easily "see around the corner".

I end up in weird places or nowhere at all.

I expect this "magical-ness" property is the same feeling driving every AI doomer on the planet saying we should regulate AI and stop "AI Fast Takeoff" or the idea that at some future point AI will start improving itself and become "infinitely smart" over night and at that point it will just decide what to do with us and we won't really have a say.

I'm honestly not worried about a "fast takeoff".

Humans are good at fixing problems mostly because we are good at making them.

If it happens we will fix it.

I do however worry about what to spend my time on.

How do I build things people care about so I can have fun, make money, and have an awesome life?

I'm definitely a believer that technology accelerates and we are in an interesting point on the exponential curve kicked off by the industrial revolution.

-

The new stuff comes in waves.

We're all meme surfing monkeys in some ways.

We go to the beach when we are ready to do something new.

We try to catch a wave.

Some of us do.

-

The waves your probably familiar with are likely the personal computer, the internet, mobile, software as a service, and crypto.

The waves keep coming faster and faster.

Catching a wave is not the only way to build things that matter or have a good life but it can drastically change the experience you have from feeling like your pushing a boulder up a hill forever with no reward to one where you are the master of a ship pushed by the winds of time creating something amazing that people actually use and care about.

Timing has always been important.

Waves are what bring in new curious minds, money, excitement, permission and a Lollapalooza effect of every other important thing that matters to create a new version of the world today.

So is AI the next big wave?

It's undoubtably going to create many wonderful things and is a worthwhile wave to try and catch.

But perhaps its also the last wave before the tide goes out.

In hindsight its a bit sad that some of the brightest minds of a generation have gone to optimizing ad clicks, e-commerce conversions and B2B SaaS.

I know from time to time when I stare out the window and realize the days of my life are spent looking at a screen I feel this might not be how I should be spending my years.

Since almost all of the waves of the last 50 years have been digital it has pulled all the air out of the room to do anything else.

If you have been smart, ambitious, greedy, insane or otherwise motivated towards doing something big you have probably ended up working on software.

It's where people have been winning your entire life.

But maybe thats over now.

-

When I do get glimpses of the future trying to "see around the corner" of AI one thing keeps screaming at me.

We may no longer need to build vertical specific software in a few years.

AI's magic is partly related to it's "general purpose-ness".

Computers until now have been "deterministic" or very very single minded inflexible machines that only bent to our will towards being "user friendly" after 50 years of trying.

Everything you use today is essentially a really well organized spreadsheet with nicer colors and lots of safeguards to make sure no one can screw it all up.

What happens if you give something akin to "intelligence in a bottle" a clock, a calendar, and a notebook with the sole purpose of helping you accomplish things and it works great?

Well that might just eat 90% of all software.

I'm not saying the software industry will disappear.

Human purchasing patterns are way slower than technological change.

I'm just saying unless your working at one of the few companies who figures out how to provide that system. A general purpose AI for life and business. Working on software might no longer be a very interesting thing to do.

Software may be a solved problem.

-

So whats next?

Where does all the creative energy and ambition of a society go when software just gets boring or out of fashion?

The only place it can.

The real world.

Maybe we look back at 50 year of software production as a necessary sort of dark age between the microchip and the future. Every generation sacrifices something for future generations.

It feels like the ball is starting to roll too.

We are regularly building and launching rockets into space.

We are fighting climate change with science and industry.

You can even get VC money to build military defense hardware.

-

So whats around the corner for AI and humanity?

Maybe we get lucky and go back to atoms.