Baldur Bjarnason

Web dev at the end of the world, from Hveragerði, Iceland

Getting outshone by great writers is a relief, not a failure (so many links)

I’m firmly of the Roy Peter Clark/Twyla Tharp “box” method of creativity.

The short version of the “method” is you start to collect “stuff” in a box – links, notes, ideas, observations – that fit a vague and poorly defined idea. Over-categorising at this stage is lethal. This sounds dumb but it serves as a form of discovering – over time – unconscious themes in your reading and thinking.

Roy Peter Clark, of “Writing Tools” fame, used boxes of index cards. Anne Lamott, who wrote “Bird by Bird”, used more organic stacks of index cards scattered around her home office. Twyla Tharp used an actual box of random objects. I tend to use folders of text files along with physical notebooks.

Most people who use this method have more than one “box” going at a time.

The common theme is that, at an unpredictable point, one of those boxes suddenly “makes sense” and you discover the thread – the core idea that runs through it.

All of my books came about this way – how I came to their core ideas:

  • Out of the Software Crisis: most software failures are down to management disregard of development as a holistic system that includes the people – both staff and customers.
  • The Intelligence Illusion: generative models are a crude facsimile of intelligence and because of that their functionality is severely compromised and pose too many systemic risks to be used safely by businesses or institutions.
  • Yellow: software development can be substantially improved by compassion and an understanding of a few relatively straightforward principles.

Even though I was absolutely certain that I was done with all of this book writing business – it’s probably the least effective way of paying the bills available to somebody with software development skills – I’m finding that my notes are congealing again and a crisp red thread is running through them, showing how the many ideas in the box are just multiple reflections of one.

I don’t know if it’ll actually cohere any further. A thread and an idea being there doesn’t mean pulling on it will lead to a book.

What it definitely leads to is more reading. Once you have a theme, more reading lights up the landscape and helps you see more clearly.

So, probably not coincidentally, this week’s collection of links has a good number of excellent reads, some of which I’ve tried to call out specifically below.

An in depth deep dive into the deep learning dilemma

P(Dumb) — Dorian Taylor

So many huge, conspicuous, world-changing events have to happen before any of those sci-fi situations are even close to plausible

Go read this

I think this is a fantasy. It’s concocted by people rich enough to already enjoy human servants, assuming—probably correctly—that there are people out there of lesser means who want the same kind of access.

No, really. Go read this.

On how the AI Bubble has revealed the software industry to be a garbage fire

Fire, Ready, Aim. The race to shoehorn half-baked AI…

Meanwhile, the customers whose loyalty businesses have worked so diligently for years to win have watched in dismay as the products and services they’ve come to rely on have been, practically overnight, colonized by an infestation of poorly considered, hastily implemented AI-fueled “features” that, like a break-out of acne, make almost any interaction a nightmarish exercise in humiliation and futility.

On how “AI” isn’t meant for customers or people in general

Creative Good: AI isn’t meant for us

How did we enter into an economy drenched in AI investments with (so far) little or no appreciable payoff for us, the citizens and students and customers? Because the AIs weren’t developed for us. They were never meant for us. They’re instead meant for the owners of the corporations: promising to cut costs, or employee count, or speed up operations, or otherwise juice the quarterly metrics so that “number go up” just a bit more – with no regard for how it affects the customer experience, the worker experience, the career prospects of creators and writers and musicians who have been raising the alarm about these technologies for years.

The politics of corporate loyalty

TBM 307: The “P” Word, Legitimacy, and Challenging Our Default Models and Frames

In countless conversations with friends in the industry, there is a pervading sense of decreasing company loyalty and increased questioning of the core tenets of working in tech.

A big contributor here is layoffs.

Frontend dev is a mess (Alex Russell’s “Reckoning” series)

Reckoning: Part 2 — Object Lesson - Infrequently Noted

Here I am obsessing about every K of JS while the industry standard for sites with potentially millions of users is megabytes of JS.

Reckoning: Part 3 — Caprock - Infrequently Noted

This is an industry-wide scandal. Promoters of these technologies have not levelled with their customers. Instead, they continue to flog each new iteration as “the future” despite the widespread failure of these models outside sophisticated organisations.

…

In practice, the complex interplay of bundlers, client-side routing mechanisms, GraphQL API endpoints, and the need to rebuild monitoring and logging infrastructure creates wholly unowned areas of endemic complexity. This complexity is experienced as a shock to the operational side of the house.

Reckoning: Part 4 — The Way Out - Infrequently Noted

Unacceptable performance is the consequence of a chain of failures to put the user first. Breaking the chain usually requires just one insistent advocate. Disasters like BenefitsCal are not inevitable.

…

Frontend’s culture has more to answer for than lost profits; we consistently fail users and the companies that pay us to serve them because we’ve let unscrupulous bastards sell snake oil without consequence.

…

But the crisis is incontrovertable in the data. If the web had grown at the same pace as mobile computing, mobile web browsing would be more than a 1/3 larger than it is today. Many things are holding the web back — Apple springs to mind — but pervasive JavaScript-based performance disasters are doing their fair share.

I respect Alex fighting the good fight here, and I hope his efforts lead to a culture change in web dev, but…

The simplest explanation for the status quo is that management in the software industry is largely incompetent and that executives largely don’t understand any strategy that isn’t “manipulate the stock market or investors”. In this version of history, the web platform had the misfortune of not being as mature as “native” stacks once the finance engineers and “I’m in tech to get rich” took over and industry brokenness set in, which I’d peg at as happening around 2008, which is why it feels disproportionally broken compared to other stacks.

I hope I’m wrong, though, because it’d be nice to have a healthier web dev industry.

You can also find me on Mastodon and Bluesky