I thought of a different title, “The new luddism is okay”. Apparently, neo-luddism is an established term, and I did not fall for the Ted Kaczynski meme yet.

Likewise, I would like to avoid the trendy rant on topics of “digital minimalism” and “information detox”, the subjects online celebrities are focused on lately. I quit smoking at least two times, so far, successfully; let’s not quoting Mark Twain here. Nevertheless, over past months my RSS feed inspired me to do the brain dump below, loosely triggered by a nice Andrey Listopadov’s post about the “vibe coding” (Jesus, vibe codding! It does sound funny).

The services I don’t use

In 2020 Donald Trump lost election, did Donald Trump things and consequently was banned from Twitter. I took it as a global censorship move (it is), and ceased most of my social network accounts. Can’t say I have been using them much, anyway.

Then, there was Telegram. For the post-Soviet internet, Telegram is the only network that matters. Of note and by the end of 2024, it still revolved around its messaging functions, and thankfully so - users are not flooded with explicit content suggestions and dubious monetization schemes yet.

The war started, some characters I followed succumbed to a genuine necrophilia, and the amount of clearly unhealthy content grew too much. But the final straw came from a less expected angle, with the French government efficiently achieving in a week what Vladimir Putin failed to do. I don’t miss Telegram, though.

So, the last mainstream social network of value is LinkedIn.

LinkedIn is bad

Indeed, LinkedIn is too full of folks thrilling to announce something. But most people are way too busy to broadcast their professional life online. LinkedIn certainly has its uses, to exemplify few: a big and universal address book, a digital collection of business cards. Although I fail to understand the intricacies of networking there, it comes in handy during the social events.

Within my interests, there are groups with really good content. There are scientists actively writing - not basic things, but stances on pretty specific problems I find interesting. And then, there is the rest - the murky noise, generated via the not so creative use of LLMs, and Microsoft’s algorithms that litter the feed. Distasteful and boring AI-generated pictures to supplement the very insightful messages are included. Seeing this makes one believe the dead internet theory, in a slightly modified form though. In this particular Brave New World, the human beings are the agents for AI, not the other way around.

In essence, the email spam problem from 2000s is back, but severely inflated. Still, I prefer the astronauts lost in space to the 10 empowering advice from a prominent communications leader.

Consumer technology

And here enters the argument I hear often, that a person won’t fit into the modern world unless they follows its developments. The argument is as mundane as it is fair, albeit the definitions of said developments are not so clear.

As an aftermath of affordable computing and online upbringing, my generation is supposed to have high technological literacy. This statement is flawed. The main, and oftentimes the only computing device a person uses is a smartphone. Modern smartphones are the pinnacle of the consumer-oriented tech. Designed and tailored for ease of use, with the associated services inevitably navigating the user towards the internet-enabled lifestyle. Such technology is trivialized to fit smoothly into the everyday life, and naturally the best experience should not require onboarding or training. Unsurprisingly, the technical competence is abstracted away.

Thus, why should mastering the omnibus and already standard practices give a competitive edge? My grandmother, a stereotypical 87 years old бабушка, uses her smartphone to find recipes on YouTube.

Scientific equipment

And here is a chance to elaborate on things I possess some experience in. The design of scientific instrumentation and research equipment is stirred in a wrong direction.

For a start, there is lack of quality documentation. I spent a good share of time driving relatively dated equipment, FPLC systems from the begining of 2000s, biosensors of the same era - what a wealth of knowledge the instrument handbooks were. Reading them is not a requirement to run the machines, but helps to build expertise and simplifies troubleshooting and interpretation of the readouts. Not a case anymore - at best, bits of documentation are scattered online. Usually, a direct contact with the dedicated helpline is expected.

To corroborate with an anecdotal evidence, the equipment with advanced autosamplers or embedded robotics used to be operated through scripting languages. The learning curve was an overhead then, but totally worth the effort - paving options for advanced experimental designs, allowing one to understand the origin of the results. One iteration, and block workflows were introduced instead. Fast forward through another iteration, wizards are in charge of the workflow configuration. Said wizards have a useful documentation too, kindly indicating at which steps the user presses the button “Next”.

There is a notorious, successful instrument manufacturer that started in early 2010s and aggressively markets its tools as being extremely user-friendly. All possible whistles are included, nice interfaces with smooth animations, the company scored numerous design awards. At a first glance, this is an outcome-based business model, and the company offers not the instruments itself, but a straightforward path towards the experimental data. However, black boxing complex matters and templating physics and chemistry into empirical guidelines stimulates overconfidence in the performed experiments and in the acquired results, quantities which are normally scrutinized in RnD and science. Yet, backed with a procurement experience, I see how the strategy works and sells well.

And then, the new trend is to offer external monitoring of the performance of the instruments, to wire them directly to the manufacturer’s or vendor’s infrastructure. There are certified laboratories, there are compliance requirements. But do SaaS and models akin constitute a good solution? I am not sure if it worked for the end user benefit in other areas, e.g. with consumer electronics.

Is it “transofrmative”?

Business models are established, platforms are built and are evolving. Do they have a strong impact on my life? As I discovered, definitely not on a personal level; of course, not without some degree of inconvenience, to exemplify with an outline by Wouter Groeneveld here.

The occupational impact is real. It does not hurt to bounce thoughts against ChatGPT (when I don’t have a colleague in proximity to bounce against), AlphaFold is great (and no, it did not solve the problem of protein folding, neither it makes structural biology obsolete or accelerates drug discovery on its own), LLMs help me with the rudimentary scripting tasks I face occasionally. The so-called disruptive technologies, when they are disruptive indeed, tend to blend into work coherently. Therefore, I remain skeptical of the accompanying claims that the evangelists of the “new digital era” have. The same era that started not yesterday, by the way, but probably quarter century ago.