I'm an experienced software engineer and ceramicist and I enjoyed this essay. I have stuff to say.
There are some wonderful industrial ceramic designs, e.g. Royal Copenhagen. But most of it is cold. Despite that people still become attached to their factory-produced porcelain mugs, because we want our daily objects to be our life-companions. We demand certain qualities that lead to intimacy. Studio pottery is a direct line to these deeper relationships, so when people own studio pottery for the first time they rarely want to go back. It's a red pill moment.
Studio pottery is not a luxury. You can buy a mug from one of the finest potters in the world for £50, and you can buy a beautiful piece for £25-£30. These might bring years of pleasure, and that registers with our customers as fantastic value. So there are plenty of studio potters all over the world who make a living with their craft.
However the product of craft-produced code and AI-produced code are, for a customer, mostly the same. So my fear is that writing code by hand will become little more than a challenging and pleasurable distraction, like a big-brained version of solving sudoku, whereas making pottery will always have a value that outstrips it's factory-produced counterparts.
But I think there is a parallel between clay and code here. There are night-school potters who just love making and getting away from their fucking screens. I have love for every one of these people. And there are those who take it much further, who read the books and design the kilns, who wood-fire in shifts over many days, who study glaze chemistry, who create objects no one has imagined before. And in software there are the line-of-business enterprise coders, and often they're handle turners who would really rather be doing something else but there are those who take it much further, who read the books and language specs, who work on foundational open source tools and study compiler design, who create idioms and paradigms no one has yet conceived of.
All that's very interesting. But for me the commercial side is prosaic and a bit dull. The pleasures of both are the creativity, which is itself a way of re-enchanting my materialist and bewildering late-capitalist way of life.
In the late 90s and early 2000s there was a bunch of academic research into collaborative multi-agent systems. This included things like communication protocols, capability discovery, platforms, and some AI. The classic and over-used example was travel booking -- a hotel booking agent, a flight booking agent, a train booking agent, etc all collaborating to align time, cost, location. The cooperative agents could add themselves and their capabilities to the agent community and the potential of the system as a whole would increase, and there would perhaps be cool emergent behaviours that no one had thought of.
This appears, to me, like an LLM-agent descendent of these earlier multi-agent systems.
I lost track of the research after I left academia -- perhaps someone here can fill in the (considerable) blanks from my overview?
As someone who got into ongoing multi-agent systems (MAS) research relatively recently (~4 years, mostly in distributed optimization), I see two major strands of it. Both of which are certainly still in search of the magical "emergence":
There is the formal view of MAS that is a direct extension of older works with cooperative and competitive agents. This tries to model and then prove emergent properties rigorously. I also count "classic" distributed optimization methods with convergence and correctness properties in this area. Maybe the best known application of this are coordination algorithms for robot/drone swarms.
Then, as a sibling comment points out, there is the influx of machine learning into the field. A large part of this so far was multi-agent reinforcement learning (MARL). I see this mostly applied to any "too hard" or "too slow" optimization problem and in some cases they seem to give impressive results.
Techniques from both areas are frequently mixed and matched for specific applications. Things like agents running a classic optimization but with some ML-based classifications and local knowledge base.
What I see actually being used in the wild at the moment are relatively limited agents, applied to a single optimization task and with frequent human supervision.
More recently, LLMs have certainly taken over the MAS term and the corresponding SEO. What this means for the future of the field, I have no idea. It will certainly influence where research funding is allocated.
Personally, I find it hard to believe LLMs would solve the classic engineering problems (speed, reliability, correctness) that seem to hold back MAS in more "real world" environments. I assume this will instead push research focus into different applications with higher tolerance for weird outputs. But maybe I just lack imagination.
Maybe this article can help you. It mentions the multi-agent research boom back in the 1990s. Later, reinforcement learning was incorporated, and by 2017, industrial-scale applications of multi-agent reinforcement learning were even achieved. Neural networks were eventually integrated too. But when LLMs arrived, they upended the entire paradigm. The article also breaks down the architecture of modern asynchronous multi-agent systems, using Microsoft's Magentic One as a key example.
https://medium.com/@openagents/the-end-of-a-15-year-marl-era...
Can I read a physical book when I'm running? Can I read a physical book when I'm doing house work? Or throwing pots in my studio? Or knitting? Or cooking? Or driving?
Can I stay focused if I'm just sitting and listening instead of reading? Absolutely not.
There's no 'vs' here. They occupy different spaces. All hail text, whatever form it takes.
> How do you listen to audio books when you're engaged in other activities?
The same way you can probably carry a conversation while walking. As long as the other activity doesn't engage the language processing part of the brain there's no conflict. I can't listen while reading, not even small package labels in a grocery store, but I can easily listen while washing dishes or exercising or eating a meal because different sections of the brain are responsible for them and can be active simultaneously.
The activities are mindless. Cutting grass, shoveling snow, cleaning the house, gardening, painting, and going on a run. Sure, I'll get distracted by something and need to rewind it back in 20 seconds, but that happens in books, and I'll need to re-read a page.
I think it entirely depends on what audiobook you're listening to. If I'm listening to something that's lower complexity or aimed at younger audiences (e.g. The Martian by Andy Weir or The Hitchhiker's Guide to the Galaxy by Douglas Adams), I can miss bits and pieces of the narration without losing too much of the story. As long as you don't lose track of several uninterrupted minutes at a time, you'll probably get most of the relevant context. The same applies if I'm in the mood to listen to a story that's already familiar to me, like Lord of the Rings.
It definitely becomes more difficult to multitask with harder reads; which is where I prefer to have a book in hand. I'll have to rewind several times per chapter to catch everything. Though it's still doable for some, I'm sure.
That's a great question! It comes up for me a lot.
I can walk through my local woods and listen very carefully but I don't think I could read a recipe book and pay attention to an audiobook!
So for me and I guess for most people it depends on the intensity of the activity and on the nature of the book. I have gradually found where the balance is for me and I've learned to hit the pause button or rewind 30 seconds.
I also pick different types of books to read and to listen to. I _love_ having an engrossing story as an audiobook. And popular science is a great listen, e.g. I loved the audiobook of Stolen Focus. But with physical books I tend to choose more involved or complex novels (e.g. Deep Wheel Orcadia), poetry, and for non-fiction books that are more specific to my interests (psychotherapy, ceramics).
It depends on the book and the activity. Not just audiobooks, but I can read a nice fiction book at a busy coffee shop no problem, but I need a controlled quiet environment for a dense, technical read. With audiobooks, I have the mental bandwidth for a fiction book or nontechnical story based non-fiction book. If I’m going for a walk or a light jog, I can listen to a book no problem. If I’m going for a run, I can’t concentrate enough. Vacuuming and other mindless chores are fine, but cooking is too much of a mental load. It isn’t a binary thing, it is just depends on how much mental bandwidth you have vs the book you are consuming.
These type of audio books tends to not be very dense, key points are often re-iterate, the pacing is slow. The type of activities they do are often not very engaging also, e.g.: if they're cooking, it is just a common dish that they already know from memory, not some new recipe they are trying to learn, etc...
I was en route to St Helena and I had several days of a raging fever on Ascension, and my memories of the place on either side of my illness are suitably strange. I remember walking through a landscape of sharp, anthracite grey volcanic rock and throwing a banana peel into the sea, to watch the fish churn around it like piranha. I remember going past a rock covered in paint -- everyone who was determined to never come back added a new splash of colour. I think it was right next to 'the worst golf course in the world'. I remember leaving the barren low-lands and climbing the mountain switchbacks, into rainforest-like verdancy. A very odd place.
Normal firebricks wouldn't withstand mid or high fire temperatures. They'd crack or melt or both. An iron grate wouldn't fare well either. You could probably make modifications to the fireplace that could make it possible, but that that point you'd probably decide to take the whole thing away from your wooden floor/carpets.
The raw clay wouldn't be vitrified, so it would be porous and would seep or leak. A glaze is a glass, more or less, and to get the silica to melt you need a flux. There are different fluxes for different temperature firings, but suddenly things are getting a little more precise. Without substantially levelling up the kiln tech and design you'll be at best low-fired and probably using lead as a flux. E.g. terracotta and earthernware.
Early peoples would have used wax or fats to seal pots like these, to make them functional. People do that with modern pit-fired pots too, or use other sealants.
I designed and built my own high-fire kiln, but it uses industrially made light-weight insulating and refractory brick, and gas burners, and I use Orton cones to know when I've hit the right amount of heat-work, and a pyrometer to take temperature readings. But some brave souls make their own bricks, and look at the colour inside the kiln to know when they're at temperature.
I used to use an electric kiln to do firings for my partner. It was fun working on the process. Our house had a lowish supply voltage so I ran 10mm2 cable back to the consumer unit, and experimented with the impedance of the elements, as we did a lot of quite high firings. I used an electronic kiln controller mostly but occasionally verified it against an Orton cone. Gas would have been more economical but a bit more scary, for me anyway.
It's scary for me too! And way less economical in the UK than electric. My small electric kiln costs £5 to fire. My big (perhaps 4 times the size) gas kiln costs £70-£80.
I'd much prefer firing with wood. But I'm too suburban and firing with wood takes much more effort in prep and during firing, but it's a pleasant and exciting experience! Which gas firing is not.
You can eat as many fermented foods and do as much tai chi as you like, and you can get therapy and manifest and go on retreats. But nothing improves life as much as a 4-day week.
Another reason for using tmux: never worry about changing your muscle memory for splits and nav ever again.
I've been using tmux for five or six years. In that time I've used Linux and MacOS and several different terminals. But how I navigate around a terminal has stayed the same thanks to tmux.
Another reason: put complex tmuxinator configs in source control and start up a bunch of stuff for your project with a single command.
Your last point is why I use it. I think many don't know you can configure it to open apps, run commands, and open splits on start. Super helpful when projects get big!
There are some wonderful industrial ceramic designs, e.g. Royal Copenhagen. But most of it is cold. Despite that people still become attached to their factory-produced porcelain mugs, because we want our daily objects to be our life-companions. We demand certain qualities that lead to intimacy. Studio pottery is a direct line to these deeper relationships, so when people own studio pottery for the first time they rarely want to go back. It's a red pill moment.
Studio pottery is not a luxury. You can buy a mug from one of the finest potters in the world for £50, and you can buy a beautiful piece for £25-£30. These might bring years of pleasure, and that registers with our customers as fantastic value. So there are plenty of studio potters all over the world who make a living with their craft.
However the product of craft-produced code and AI-produced code are, for a customer, mostly the same. So my fear is that writing code by hand will become little more than a challenging and pleasurable distraction, like a big-brained version of solving sudoku, whereas making pottery will always have a value that outstrips it's factory-produced counterparts.
But I think there is a parallel between clay and code here. There are night-school potters who just love making and getting away from their fucking screens. I have love for every one of these people. And there are those who take it much further, who read the books and design the kilns, who wood-fire in shifts over many days, who study glaze chemistry, who create objects no one has imagined before. And in software there are the line-of-business enterprise coders, and often they're handle turners who would really rather be doing something else but there are those who take it much further, who read the books and language specs, who work on foundational open source tools and study compiler design, who create idioms and paradigms no one has yet conceived of.
All that's very interesting. But for me the commercial side is prosaic and a bit dull. The pleasures of both are the creativity, which is itself a way of re-enchanting my materialist and bewildering late-capitalist way of life.