Think about this -
We used to go to war against entropy with our hands - handmade homes, pots, pans and weapons. Then we made machines with our hands - the sewing machine, the machine gun, and then we made machines with machines. Today we have robots building the factories that build the robots.

We used to go to war with a handful of soldiers, fight for half the day and take a break at night. Inventing the machine gun and the tank should have led to great uncertainty in the profession of soldier no? Like this thing could do in an afternoon what took a trained unit a week to do. Why would we need soldiers any more?

And yet, here we are! The machine gun is a product of industrialisation - and industrial processes never make things in handfuls. They need 'economies of scale' to function. And it is still amazing that the 5 orders of magnitude more efficient machine gun led to larger armies of much more intensely trained soldiers. Go figure.

The same thing is going to happen to software engineers. And it is a tide that will lift all boats, but some boats are going to be yanked into the stratosphere.

George Dyson, the scientist who theorised Dyson sphere's had once said how we will know if an AI is on planet Earth. His idea was that if an AI arrived on the scene, it would try and stay hidden for as long as possible but you would still be able to tell because it would

  • have a hunger for processor cycles
  • try and secure its own autonomous power supply and
  • surround itself with happy, well-nourished humans.

He said that his visit to google campus in the early 00s and he was convinced there was an AI there! And he might well have been right. The AI wants to surround itself with the best and the brightest. If you have the talent it needs, whether you be minding your own business in some village in India, it will suck you up to itself with irresistible force.

In other words, in the long term, software engineering jobs are going nowhere. and there has never been a better time to go deep! As we get deeper into the era of AI, it's starting to get clear what that looks like. My clients keep asking now for the same things - someone who can sustain speed, someone who has gone deep...and they are both the same thing.

It is a habit of software engineers that they pick up a problem and start coding first and then read the literature later. One quick blog post read and then they're like I'll figure out everything myself as I go. A wonderful, devil-may-care attitude. But when earlier we would only see our large program form through the collection of thousands of keystrokes, giving the mind plenty of time to adjust to the stressed and strains on our architecture, today programs arrive fully formed. And very few people know how to verify that they are designed and working correctly - TDD, benchmarking, static analysis.

Many developers treat system design as something to be studied to clear an interview but when AI coding, it's one of the few tools still left in your harness - the architecture and big picture of the thing you're building. Somehow you have to get good at this.

Then you will need to know the fundamentals really well. Programming with an LLM requires not just a theory of mind for the machine, but also a theory of the body of the machine - the network, the database, the queue.

And last but not least, it is now possible to polish a program till it shines! It used to be very tedious in the old days. But now it's possible to get exact tooltips, popovers, gradients, copy - whatever you need the LLM provides. And if you don't want to go deep technically, then you will have to prove that you use AI to build systems that are polished, industrial engineering. What fun! What a treasure!!

Y'all are going to have to go deep into the machine to be able to ride it. All the best!

And now on to the jobs - I've added quite a few exciting new roles this week and I'll go through some of the notable additions here -

Zamana.com might be in stealth mode but won't be for much longer. What I loved about this company is the crackedness of the founders (ex Duke U, ex Lossfunk, both of them ekdam kadak) and the ambition of the vision. Founders have put together an excellent team so far, so you'll be joining a place with very high talent density. They're looking to hire for two excellent roles - a Head of Engineering and an IC.

The HoE role especially is one of the most interesting mandates I have with me for the moment. Check it out and apply here if you're 8+ years of experience and cracked. Zamana has very generous compensation for ICs too - find out more and apply here.

The next exciting client I have onboarded is Runable - started last year as the SME automation agent for global businesses, Runable has secured excellent funding on the back of their crazy adoption. A small team of 4 people is now serving millions of customers. You get to get in on the ground floor of a rocket ship but please be cracked! Details and apply here

And now on to the full list of open positions.