Why has Japan lost its edge in new technology supremacy? – Software Craftsmanship Weekly vol. 119

Today we’ll devote most of the edition to the Land of the Cherry Blossom, as I’ve gotten my hands on some insanely interesting texts about it in recent weeks. However, we’ll also talk about the cost of cloud computing (here we go again) and lawsuits in the world of generative AI.

Article cover

1. Why has Japan lost its primacy of new technologies?

When I grew up in the 1990s, Japan appeared to be the technology capital of the world. And while this belief probably persists in the broader consciousness, the reality is somewhat different. Yes, Japan is extremely modern…. for the 1980s. But at some point, the Cherry Blossom Country fell behind and never again caught up with our heavily “digital” modernity, remaining in a world of robots serving hamburgers and toilets which simultaneously play music and wash the users.

When I’m old and rich, I’ll install one at home.

What happened? The answer is probably very complicated and multi-component, so it is difficult to explain it in a coherent way with one publication. Nevertheless, we will try. That’s why I have for you today as many as three sources (two texts, one video) that lean into the subject and try to analyze why (while remaining an economic behemoth – after all, it’s still the third country in terms of GDP) from being the world’s High-Tech powerhouse Japan surrendered the fields to competitors from Silicon Valley, Korea or China.

And in my individual ranking, the country’s most innovative export A.D. 2023 remains Chainsaw Mansee, and then readtrue gold.

We’ll tackle software first – the first text to share is The forgotten mistake that killed Japan’s software industry, which inspired today’s edition, being published last week. Celebrating the bicentennial episode of the Disrupting Japan podcast, its host, Tim Romero, published a very accessible analysis that presents the reasons why Software Engineering never became such an important part of Japan’s economy and industry, and what repercussions it has had to date. The text dives into historical conditions and presents changes that Japanese industry went through throughout the 19th and 20th centuries, which made Japan, despite being a very early player in the PC market (or maybe because of it), somewhat sleepy about the personal computer revolution, focusing on the corporate sector. If you like analysis of the software market in a somewhat broader economic and social context – I highly recommend it.

Okay, but even though the Japanese never became a behemoth when it came to app development, their hardware remained world-class for years. So why is it that at one point it was Apple and then the Chinese and Koreans who became (outside the game console market) the innovators when it came to consumer electronics, especially miniaturized ones? The text Why Japan didn’t create the iPod again cites lower PC penetration than in other developed countries as the reason. In addition to the corporate-market reasons identified in Disrupting Japan, the second of today’s texts also points to… katakana – the Japanese alphabet. It was this (or, more precisely, the difficulty of creating a convenient way to input text, especially in the early years of computers) that caused the Japanese to focus on developing external media that allowed them to skip computers, instead of managing their gadgets with PCs. Such things are slow to change – there’s a reason why in the most prominent spot in Tokyo (Shibuya) stands a… a multi-story CD rental store.

Seriously, I was on my honeymoon in Japan and it was definitely the biggest shock to me.

The third publication, meanwhile, is a video Why Japan’s internet is weirdly designed, about the very specific style that Japanese websites adopt. The Answer in Progress channel decided to look at the myths that accompany the web design of Japanese websites. Are sites created in the Land of the Cherry Blossom really so different from the rest of the world? It turns out that there is a great deal of truth in the legend, and once again the reasons for this are most interesting. Indeed, the author of the video points to Japan’s very rapid entry into mobile, thickly ahead of the rest of the world, which meant that the country had to develop its own practices before “Mobile-First” became a thing around the world.

As you can see, the answer to the question “Why didn’t Japan become a software powerhouse?” is not obvious. I hope this insert has convinced you that it is worth deepening your knowledge of the nuances of the software market in more exotic (from a Western perspective) countries.

BTW: I am writing this text on Saturday morning, watching the Ski Jumping competition from Sapporo with one eye.


Discover more IT content selected for you
In Vived, you will find articles handpicked by devs. Download the app and read the good stuff!

phone newsletter image

2. Why is 37signals abandoning the cloud? As always: 💰

37signals is a very interesting company to watch. The creators of Basecamp and Hey very often make controversial decisions, against the established patterns and their approach to software development has as many supporters as opponents.

  • They’re the ones who relied on a small team and balanced growth, with no money from funds (which, in an era of layoffs caused by the growth of companies with VC money, will probably allow them to survive the current recession better than their competitors),
  • They’re the ones who were very quick to point out the advantages of remote work (and the book they published, Remote, became a kind of industry bible even at the beginning of the pandemic)
  • or banned political discussions at work (which caused a very wide resonance in the industry).

And now they have taken aim at cloud computing.

The decision was shared back in October by David Heinemeier Hansson, the company’s CTO and creator of the famous Ruby on Rails framework. The reason is simple – money. But why are we returning to this topic at the beginning of the year? The reason is simple – in addition to the initial declarations, the company showed calculations. These points that the company has spent last year more than four million dollars on the infrastructure needs of both of its main products (Basecamp and HEY). Now 37signals will try to approach the topic of its infrastructure differently. They have decided to try an intermediate solution – instead of building the whole thing from scratch, it will try to use the services of a server leasing company. However, the entire “platform layer” will be handled by the 37Signals itself.

We will be keeping a close eye on this experiment, which is already generating widespread discussion. In addition, the publication itself is a fairly detailed breakdown of the costs of the various services, so if you are curious about how these types of accounts look in global companies, I think this is one of the most transparent sources available to the public.


Discover more IT content selected for you
In Vived, you will find articles handpicked by devs. Download the app and read the good stuff!

phone newsletter image

3. Lawsuits raining down in the world of generative AI

And finally – a brief insertion about AI. This is because we have historically written about preparation for lawsuits concerning generative ML models. Last week brought the first two major cases of this type.

The first was launched by a group of artists – Sarah Andersen, Kelly McKernan, and Karl Ortiz – against StabilityAI, MidJourney, and DeviantArt (which also entered the generative AI market). The case will be led by Matthew Butterick, a well-known activist who in late 2022 began preparations for a lawsuit against Google Copilot.

The second lawsuit, also against Stability AI, was launched by Getty. It accuses the company of illegally scraping its resources. It will be interesting to see whether the case will be decided differently than it was in a similar lawsuit against scrapping by LinkedIn, which lost its case. Getty, however, is not accusing StabilityAI of violating its terms of service, but rather of violating the copyright of the images it inhabits, so the whole thing is subject to different lawmakers.

In both cases, we are talking about very early stages, but they are worth watching – their results are likely to shape our legislation on ML models for many years to come.