Dispatch

The following is a talk I gave at the Prompt Poets Society, a meetup I've been helping organise for programmers to swap notes on AI coding tools.

The advent of literacy

I want you to imagine you've been transported to Athens, 2,500 years in the past. You're probably wearing a Chiton (a sort of Toga), maybe some sandals. You've gathered in the Agora to hear someone give a speech. As you look around, there are a few customs you'd find unusual:

Sortition

The Athenians practiced sortition, selecting public officials by lottery rather than election. Imagine a letter in the mail: "Dear Madam/Sir, for the next 3 months you'll serve as member of parliament". Given the general level of disdain most people show for jury duty I'm amazed this ever worked.

Ostracism

The word has negative connotations today but I love this one: every year the citizenry would get together and vote someone off the island. Whoever got the most votes had to leave: they weren't harmed, they got to keep all their stuff, they just had to bugger off for 10 years. Can you imagine how differently our public officials would behave if they knew this could happen to them?

Gyms

The modern word for gym comes from the greek "gymnos" meaning "naked" or "bare". That's because you could only attend a gym in Athenian times if you were butt-naked and lathered in olive oil. Not only that: they had guards posted at each station and if they deemed you weren't properly groomed or giving the workout your full effort they'd have you publicly whipped.

Frankly, this strategy worked: art from Ancient Greece is replete with imagery of men and they're all shredded. But it'd be controversial if Les Mills tried it today.

Orality

So this is all pretty weird, but weirdest of all for us would be experiencing an oral tradition. Athenians relied almost entirely on the spoken word; speaking was the apex form of expression and little beyond commerce was recorded in script.

For you and I, who've been born into a society hundreds — nay, thousands — of years into literacy, life without writing is hard to fathom. Literacy is the water we swim in. So much so that we're almost blind to it. For a long time, it was believed the way we think is an inherent condition of being human. It wasn't until the 60s and 70s that we started to understand literacy fundamentally changes our patterns of thought.

So what is it like to be pre-literate?¹ Imagine never being able to write anything down. A huge mental effort must be given to memorisation. All history and laws change over time, often with the whims of whoever happens to be in power. It's also hard to transport information over distance, let alone time. This is a millstone for the progress of the species: how can you learn from your ancestors if they can't share their experience with you?

All told this leads to more conservative thought: you're penalised for straying far from what's already known. Even for a great thinker, solving a new problem is only half the battle. You then have to figure out how to remember it.

I think Homer's "The Odyssey" encapsulates what an oral tradition does to you. The repetition and formula is unutterably tedious. But this is what you've gotta do when you can't write the damned thing down.

But wait, weren't the Greeks literate?

This is the weirdest thing of all: The Athenians invented the first real alphabet. Sure, there were other alphabets banging about. You had Linear B, Heiroglyphics, even their neighbors had their own thing. But these all lacked vowels, forcing everyone to write like they did when texts cost 20 cents a piece and tweets had a 160 character limit: "SND TRBT NW THX". The reader had to guess a lot of meaning.

When the Greeks added vowels, shit got real. This was the first time a written script could cover the whole range of spoken sound. Sounds simple to us, but it's hard to overstate the magnitude of this development.

Literacy meant consistency of records over time and space. It freed the mind from memorisation, which led to less conversatism, more creativity, and ultimately an explosion of new ways of thinking. It's unlikely modern society, or our current quality of life, would've been possible without it. All told, writing was the biggest shift in human cognition since we'd learned to speak.

So the Greeks hit a home-run. Why didn't they flock to it?

Well, remember: orality was the water they swam in. As deeply as we're ingrained in literacy, they were deeper still in the spoken word. Old habits die hard. This meant they were initially dismissive ("Pfft, writing? Who cares"). More than that, they'd ridicule anyone seen to be relying on writing. It was a habit of lesser men, and orators were mocked if they tried to speak from written notes.

A change this big is likely to inspire fear and skepticism as well. Here's Plato on literacy:

Writing will produce forgetfulness in the souls of those who have learned it, through lack of practice at using their memory, because they will put their trust in writing.

Then again, elsewhere:

You have invented an elixir not of memory, but of reminding: and you offer your pupils the appearance of wisdom, not true wisdom, for they will read many things without instruction and will therefore seem to know many things, when they are for the most part ignorant.

This fear is, I think, reasonable. When your whole life and career is dependant on speaking, it stands to reason you'd see writing as a threat.

What I find fascinating is that Plato's fears have largely been realised: an oral culture requires much better powers of recall than most of us have today. We look back at Athenians as masters of oratory, but at oratory as almost irrelevant: a nice skill to have, but not decisive in making your future.

Returning to the now

With the rise of AI and LLMs, we're faced with a similarly daunting shift. Earlier I said literacy is the water we swim in, but it's also internal cognition. All of our thought: our reasoning, synthesizing, and summarising, has happened in our own heads. Until now.

For the first time ever, LLMs present an alternative. And this is a big, scary shift. Bigger, perhaps, than literacy because thinking is imbued with responsibility in a way that recording isn't. So it's reasonable for us to have some of the same fears as the Athenians. Especially for those of us who are knowledge workers, selling our thought by the hour.

I think we'd be more excited if we weren't so worried. Imagine explaining to a pre-industrial person that not only have we automated most manual labourm but we've also automated thought too. They'd think we lived in paradise.

How do we get past the existential dread? The way I see it, we have three main reasons for optimisim:

History shows augmentation is a win

For every major shift in cognition, we've seen leaps in the success of our species.

• Learning to speak conveyed obvious advantage: we could organise, collaborate, coordinate
• Literacy for all the aforementioned reasons
• The printing press democratised access to knowledge
• Digital media threw fuel on that fire, and
• The internet connected us all.

These were all great things, despite a painful period of adaptation. Which brings me to my next point:

History shows adaptation times are shrinking

How long did it take us to learn to speak? Millions of years?

By the time we'd invented literacy, that time had shrunk to hundreds of years. The Athenians were pretty well literate by the end of their golden period.

It took perhaps 50 years to adapt to the printing press. At first: bedlam. A wave of content with no way to distill the good. Then came publishing houses, reviewers, libraries. Books themselves got ergonomic upgrades like page numbers, pages of contents, appendices. Things got easier.

Fast-forward to digital media and it's faster again. And finally the internet, which took a decade to go from the exclusive domain of a druidic order to Google in the hands of all.

History shows us adaptation time is rapidly declining, probably as a function of how many people are working on it. The Athenians might've had 30,000 literate people. Today we have millions, if not hundreds of millions of people working on or with AI.

As coders, we'll learn first

Software has eaten so much of the world, and the stakes for figuring out how to build software quicker with AI are so high, that in many ways it's become the tip of the spear.

We, as programmers, will be among the first to shift to this new kind of cognition. I find that incredibly exciting: learning first means we'll get to teach it to the rest of the world. Both directly, and through the products we build.

That's an incredibly privileged position, one we should cherish and take seriously.

Closing

I wonder if society 2,000 years from now will look back on us in a similar way to how we view the Ancient Greeks: masters of internal cognition, but at internal cognition as mostly irrelevant.

If they do, it'll be from a place of unfathomable progress. And I'm sure it's the work of all of us here that will have helped them get there.

• 1.

  Ted Chiang's "The Truth of Fact, the Truth of Feeling" does an excellent job of exploring this, in his book Exhalation.

I wrote this today in an attempt to articulate my feelings about climate change, degrowth, and the common narrative around renewables:

I'm convinced there's no hope of limiting global warming by any of the methods currently being employed. As a species we just won't give up our comforts until we're truly priced out of it.

I don't say this to mean we shouldn't try. But I think we need to be realistic about what we are (and aren't) capable of. Degrowth is like communism: a great idea, but not practical.

As far as westerners go I think I'm somewhat "green" in persuasion. But even still my life is hopelessly emissions-intensive and wasteful. As a society our whole mode of being is built around burnt carbon.

This doesn't have to be a downer: if we accept that offsets, renewables, and international treaties aren't going to work, we could busy ourselves with what might. My answer is something like fusion power. That, and moving resource extraction and heavy industry to space. You may think all that sounds fanciful. But it's no more fanciful than trans-continental rail or international jet travel was two hundred years ago. These things will be commonplace in the near future.

We need something that lets us use more energy without consequences. Energy spend per capita is the holy grail of quality of life.

I rarely expect to see my thinking aligned with someone like Jeff Bezos, but he expressed a similiar sentiment in a 2023 interview with Lex Friedman:

So, it is a fact that 500 years ago, pre-industrial age, the natural world was pristine. It was incredible. And we have traded some of that pristine beauty for all of these other gifts that we have as an advanced society. And we can have both, but to do that, we have to go to space. And the most fundamental measure is energy usage per capita. You do want to continue to use more and more energy, it is going to make your life better in so many ways, but that’s not compatible ultimately with living on a finite planet. And so, we have to go out into the solar system. And really, you could argue about when you have to do that, but you can’t credibly argue about whether you have to do that.

He then goes on to explain that Blue Origin, his rocket company, is his contribution to this goal in the form of critical infrastructure:

When I started Amazon, I didn’t have to develop a payment system. It already existed. It was called the credit card. I didn’t have to develop a transportation system to deliver the packages. It already existed. It was called the Postal Service and Royal Mail and Deutsche Post and so on. So all this heavy lifting infrastructure was already in place and I could stand on its shoulders

[...]

It was already there. And that’s what I want to do. I take my Amazon winnings and use that to build heavy infrastructure so that the next generation, the generation that’s my children and their children, those generations can then use that heavy infrastructure

Assuming we all want to live, and we all want to live comfortably, it seems the only way out is up. Why is this effort only being championed by a handful of billionaires?

On a related note, I just finished The Future of Fusion Energy by Jason Parisi and Justin Ball. It gives an approachable look at where we're at with fusion, what needs to be solved, and how we can get there. A common theme throughout is underinvestment. We just haven't prioritised it, probably because of the popular sentiment that nuclear is bad.

Overall the book left me feeling hopeful. It's within grasp, we just need an Apollo-style program: fast-tracked and heavily funded. And we need it yesterday.

Compiling a reading list last year was fun so I'm going to keep the trend going. Besides, isn't everyone doing reading lists now? Here's the highlight reel from 2024:

Barbara Kingsolver's Demon Copperhead was an easy favourite. A portrait of forgotten children and victims of the American opioid crisis. The narrator's idiosyncrasy is addictive.

Claire Keegan's Foster and Small Things Like These. Books so short that transport you so far. Painful memories of Ireland's past.

Hugh Howeys's Silo series. I read a lot of sci-fi this year and this was by far my favourite. Howey reveals just enough to satisfy, and no more. The world he constructs is palpable, the characters believable. A rare blend of imagination and delivery.

On the flipside, the most tedious reads were Tolstoy's "Anna Karenina" and Rand's "Atlas Shrugged". I knew, almost immediately, that I wasn't going to enjoy finishing either of these and yet I persisted. Why? The faint hope that maybe it'd turn around? More likely that I'd feel like a quitter. That was dumb and I'm resolved to give up on books more often this year.

A couple of years ago, NASA and Durham university¹ put out this breath-taking simulation of the formation of the moon:

That's a Mars-sized planet colliding with our planet billions of years ago, throwing up debris that would eventually coalesce into the moon. And all of this is thought to have happened in just a few hours. Can you imagine?

For the curious reader, this is the Giant-impact hypothesis. A theory with a whole lot to back it up: Geologists have found evidence of the heart of another planet embedded inside ours. And the moon is suspiciously similar to Earth in core, mantle, and crust makeup.

In fact, the moon's done more for us than we realize:

• It stabilizes Earth's rotation, preventing wild climate swings
• It causes tides, which were crucial for our eventual evolution

And it rotates at the same speed as it orbits, which is why we only ever see one side of it. Cool, right?

• 1.

  I couldn't find a source for this but I'm pretty sure it was generated by the COSMA8 supercomputer, which has over 528TB of RAM and 67,584 CPU cores!

This year I read just shy of a book a week¹. That's an astonishing pace for me, and considerably more than any previous year. It's owed in large part to audiobooks. 2023 marked a change in lifestyle for me, one with much more motion in it to resolve a long struggle I've had with backpain. This motion — usually walking, sometimes running — is usually accompanied by an audiobook of some sort. The result is that half the books I've "read" this year were actually listened to.

It's my experience that the format doesn't lend itself to every genre, much less every title. Simon Sarris captures it well:

Audiobooks vary tremendously in quality. The performer is almost as important as the author.

So sampling performances is a lot more important for the spoken word than written.

I've heard it said that audiobooks aren't absorbed as well as written books and I'd agree. I'm rarely as attentive to the content when I'm going for a run verses reading quietly in bed. But this is easily worked around. Most of the books I've listened to are biographies. The kinds of books where missing a little detail doesn't matter as long as you get the overall picture.

Something I rarely see credited is the way a spoken performance can bring the text to life. My standout read for the year, and among the most impactful books I've ever read, was Alfred Lansing's Endurance, an account of Shackleton's 1914 expedition to Antarctica. The book is powerful in its own right, but Simon Prebble's performance of it makes it all the better. The metre, the excitement and the accents all combine to make it so much more thrilling.

With my rant about audiobooks over, here's some of my favourites from this year:

Irving Stone's The Agony and the Ecstacy, a biographical novel about Michelangelo Buonarroti. This was the first time I'd read a biographical novel as opposed to a biography and, christ, what a break from the dry and burdened prose of pure fact.

John Steinbeck's The Grapes of Wrath. I imagine every American reading this was assigned it twice in high school. And the long list of awards and commendations makes it no surprise. But boy, was it good. So good. Steinbeck's way of exploring the people, the place, and the times is so vivid and immersive.

Paul Graham's Hackers & Painters, which is a compendium of blog posts. This is a read I'd only really recommend to computer people. Graham is a divisive character in the tech industry so I wasn't sure what to expect. But he gave shape to a lot of the doubts I'd had about myself throughout my career and delightfully dispelled them:

All the time I was in graduate school I had an uncomfortable feeling in the back of my mind that I ought to know more theory, and that it was very remiss of me to have forgotten all that stuff within three weeks of the final exam.

Now I realize I was mistaken. Hackers need to understand the theory of computation about as much as painters need to understand paint chemistry. You need to know how to calculate time and space complexity and about Turing completeness. You might also want to remember at least the concept of a state machine, in case you have to write a parser or a regular expression library. Painters in fact have to remember a good deal more about paint chemistry than that.

And then:

I was taught in college that one ought to figure out a program completely on paper before even going near a computer. I found that I did not program this way. I found that I liked to program sitting in front of a computer, not a piece of paper. Worse still, instead of patiently writing out a complete program and assuring myself it was correct, I tended to just spew out code that was hopelessly broken, and gradually beat it into shape.

For a long time I felt bad about this, just as I once felt bad that I didn't hold my pencil the way they taught me to in elementary school. If I had only looked over at the other makers, the painters or the architects, I would have realized that there was a name for what I was doing: sketching. As far as I can tell, the way they taught me to program in college was all wrong. You should figure out programs as you're writing them, just as writers and painters and architects do.

I can't tell you how liberating this was to read. I'd worked this way for years and had never drawn the parallel. Irving Stone's book concurred: Michelangelo didn't spend long planning his work, he just took chisel to rock and "found" the form within it.

Other notable segments include his summary of schooling:

Public school teachers are in much the same position as prison wardens. Wardens' main concern is to keep the prisoners on the premises. They also need to keep them fed, and as far as possible prevent them from killing one another. Beyond that, they want to have as little to do with the prisoners as possible, so they leave them to create whatever social organization they want. From what I've read, the society that the prisoners create is warped, savage, and pervasive, and it is no fun to be at the bottom of it.

In outline, it was the same at the schools I went to. The most important thing was to stay on the premises. While there, the authorities fed you, prevented overt violence, and made some effort to teach you something. But beyond that they didn't want to have too much to do with the kids.

Overblown? A little. But certainly some truth. And finally, some excerpts on why being young is a miserable experience these days:

Teenage kids used to have a more active role in society. In pre-industrial times, they were all apprentices of one sort or another, whether in shops or on farms or even on warships. They weren't left to create their own societies. They were junior members of adult societies.

I'm suspicious of this theory that thirteen-year-old kids are intrinsically messed up. If it's physiological, it should be universal. Are Mongol nomads all nihilists at thirteen? I've read a lot of history, and I have not seen a single reference to this supposedly universal fact before the twentieth century. Teenage apprentices in the Renaissance seem to have been cheerful and eager. They got in fights and played tricks on one another of course (Michelangelo had his nose broken by a bully), but they weren't crazy.

As far as I can tell, the concept of the hormone-crazed teenager is coeval with suburbia. I don't think this is a coincidence. I think teenagers are driven crazy by the life they're made to lead. Teenage apprentices in the Renaissance were working dogs. Teenagers now are neurotic lapdogs. Their craziness is the craziness of the idle everywhere.

If life seems awful to kids, it's neither because hormones are turning you all into monsters (as your parents believe), nor because life actually is awful (as you believe). It's because the adults, who no longer have any economic use for you, have abandoned you to spend years cooped up together with nothing real to do. Any society of that type is awful to live in.

While I think this oversimplies, I like the notion that kids need to feel useful. And useful in a real sense, outside the artifice of schooling itself. Probably this is why computer kids typically do well out of school: the barriers to being immediately useful to society are low enough that a high schooler can do it. And there's rarely any formal qualification required as with other trades and professions.

• 1.

  I share this with some reservation. It's always struck me as obnoxious when people talk about how many books they've read as if it's a contest. Talking about it has helped me reflect on what I enjoyed and why, so I hope you'll forgive my transgression.

Earlier this month MBIE published a report, UpStart Nation, making proposals for how we could improve the startup ecosystem in New Zealand. Among these is the idea that we should change how we tax employee share programs.

Startup employees are taxed when they exercise stock options — that is, when they buy them and "realize" the gain — which they can do long before a company's value or future is certain. In the eyes of the tax system, though, they owe tax the moment they've exercised their options.¹

Exercising an option to buy company stock valued at $20 for $10 means you've technically made $10 even if you can't yet sell the stock. You pay tax on that $10, and from then onwards any further appreciation in your shares is treated as a capital gain and thus tax-free. Herein lies the yin and yang of startup risk: the earlier (and cheaper) you exercise, the more of the eventual gain you stand to keep. But likewise, the more risk you take by paying for them before there's certainty around their value and liquidity.

For many, that risk is too much to bear. Or they simply can't afford to exercise. In that case, depending on the terms set by the employer, they may be able to exercise at the moment the shares become liquid (whether by going public or by being acquired). This approach is risk-free but means the entire gain is taxed as income.

The report points out that there's a disparity here between employees and investors. Rowan Simpson makes a good summary of it in his critique of the report:

The current tax system we have in New Zealand means the tax paid by employees and investors in any successful business sale is calculated very differently. This is because investors pay cash to buy their shares in advance, so any gains on that investment are capital gains which are not currently taxed. Meanwhile employees earn their shares by working, which means they are a form of income, and that income is taxed (either immediately or eventually).

The report focuses on the straw man scenario where tax needs to be paid by employees immediately when shares are issued. This is true in theory, but in my experience very rare. Much more commonly these days, employees in high-growth startups are issued long-dated options rather than shares. This means the tax is deferred until an exit occurs. But tax is then due on the full value of those shares at exit, which can be very large amounts.

And then:

It’s difficult to argue that is fair - investors pay 0% capital gains tax on their gains while employees pay 39% income tax on their gains.

This is indeed a problem. Their proposal: either defer taxation until the employee can actually sell for cash or remove that tax altogether.

On the face of it the first option sounds great. But it's unfeasible for all sorts of reasons. With all the possible twists and turns of an early-stage company, it's very hard to say when an employee's shares can really be considered sellable.

As for removing the tax altogether: it's incredible to me that a group of people, when presented with this problem, would conclude that a reasonable solution is to tax no one. Having the option to own a part of the company you're helping to build is a privilege very few get. The notion that these lucky few deserve special tax treatment as well is totally unnecessary.

In a follow-up post, Rowan makes the point that ESOP terms set by the employer are what really defines the outcome for employees. As someone who's been around the block a couple of times now I wholeheartedly agree. I've had the fortune to work for two companies that had humane ESOP settings, and in the end that mattered a whole lot more than the taxes I owed².

Generous ESOP settings should be the norm. Founders might think that restrictive terms will help with staff retention, and they're kind of right. But often all it serves to do is keep people there in body but not spirit. For an ecosystem to flourish you want talent to have freedom of movement: the freedom to seek out the stage that suits them best. Things like 90-day expiry windows inhibit this freedom, and the people bound by these handcuffs trend toward low morale and productivity. I believe founders have a responsibility to get this right for their employees and the wider startup community. It's a shame more don't.

In my mind, the solution to all this is painfully obvious: just tax capital gains and move on. A broader tax base will mean less income tax, evening the burden between employees and investors while simultaneously correcting many of the distortions in our investment economy. It speaks volumes that this wasn't even mentioned in MBIE's report. Here's Rowan again with the final word:

I don’t hear many of the people complaining about tax on ESOP advocating for a capital gains tax.

Everybody just wants to pay less tax.

I can't help but feel that all this focus on "building the ecosystem" is a distraction. Given time, the ecosystem builds itself out of companies succeeding and spawning new companies. That's exactly what happened with TradeMe, Xero, and now Vend (to name just a few). I'm now an investor in, consultant for, or co-founder of companies that were started by people I worked with at Vend. Which is ultimately what MBIE wants, right? More founders, more innovation.

You don't need tax breaks to get there.

• 1.

  For a thorough explanation of how this works, I recommend Rowan Simpson's Cost vs Value

• 2.

  As a California resident at the time, I paid 45% tax on all Slack shares and 25% on Vend shares

This walkthrough video of hiking Mt Everest is a revealing look at the reality of the climb today.

Things I found surprising:

• How large and elaborate Everest Base Camp is now
• How you're surrounded by other people at all times, both during your hike and at camp
• How prepared the trail is: steps in the snow are pre-cut, with ladders and ropes in place across the entire length of the trail
• How blasé Jon Gupta, the author, is throughout the whole process. He might have been reporting in from a walk in his local park

Not pictured here but evident from other videos is the extent of the trash problem. There's enough debris that Renan Ozturk and his gang were able to dumpster-dive for food rather than carry their own when they went up there looking for the body of Sandy Irvine.

This is so far removed from what I'd consider an enjoyable or challenging hike that I'm amazed people still do it. No doubt it's hard work but it's the wrong kind of work: all the creativity and problem solving is gone. And you've paid a team of locals to take the risks for you. Where's the achievement in that?

Yesterday, Lightspeed announced their acquisition of New Zealand startup Vend for about NZ$500 million:

Lightspeed will acquire Vend for total estimated consideration of approximately $350 million, satisfied by way of payment on closing of approximately $192.5 million in cash and the issuance of subordinate voting shares in the capital of Lightspeed valued at approximately $157.5 million.

This marks the end of a 10-year journey for the company, a journey I was a part of for the first four years.

Vend had a profound effect on my life both personally and professionally. It's how I met my wife Vic and 6 years on we have a 2-year old daughter and another on the way. I grew substantially as an engineer, and this experience helped propel me to Techstars in Berlin, then San Francisco and Slack.

The story is similar for so many others that passed through Vend's doors. It's a natural New Zealand success story. And now to see this acquisition reward so many of the people that were integral to the company's success over the years is just wonderful.

There's one tweet that stuck out to me, though, amid all the congratulations. This from Simon Pound, also a former Vend employee:

Which is so bleak. 500 houses don’t inspire other companies, create hundreds of high skill jobs, build an ecosystem and bring in hundreds of millions in export revenue. Our housing crisis is even making our dreams poorer.

— Simon Pound (@simon_pound) March 12, 2021

This so perfectly captures my frustration with our country. Vend has paid the salaries of hundreds of people, made all sorts of contributions to society, and made a great return for its investors. And yet, for all its success, it’s eclipsed by our runaway housing market.

In December 2020 alone, almost $10 billion dollars of new mortgages were issued. That’s 20 times the value of Vend, and for what benefit to our country? Arguably none. How many Vends would there be if we weren’t so caught up in buying houses?

I’m celebrating this win, for Vend and for all of us that were a part of it. I hope it serves as a reminder of what we can achieve when we invest in something productive.

Mike Moffett in The Sun this month:

I'm an ecologist, so I think in million-year increments. What we're going through now is a minor blip in the multi-billion-year history of our planet. The earth is unlikely to live or die based on what happens to humans in the next hundred years. There's plenty of evidence that species will decline and ecosystems will be compromised, and for our future as a species, we need to take that seriously. But in terms of the long-term ecology of the planet, things are going to sort out. I'm an optimist that way. Even if Earths future doesn't involve humans, it will have a future.

This outlook is my strategy for emotional survival right now.

Marco Sodano renders iconic art in minimalist style with Lego.