When your daily average user numbers are just a little too good to be true.

Chrome, Firefox, and yes even you Opera, Falkon, Midori and Luakit. We need to talk, and all readers should grab a seat and prepare for some reality checks when their favorite web browsers are in this list.

I've tried literally all of them, in search for a lightweight (read: not ridiculously bloated) web browser. None of them fit the bill.

Yes Midori, you get a couple of bonus points for being the most lightweight. Luakit however.. as much as I like vim in my terminal, I do not want it in a graphical application. Not to mention that just like all the others you just use webkit2gtk, and therefore are just as bloated as all the others. Lightweight my ass! But programmable with Lua, woo! Not like Selenium, Chrome headless, ... does that for any browser. And that's it for the unique features as far as I'm concerned. One is slow, single-threaded and lightweight-ish (Midori) and another has vim keybindings in an application that shouldn't (Luakit).

Pretty much all of them use webkit2gtk as their engine, and pretty much all of them launch a separate process for each tab. People say this is more secure, but I have serious doubts about that. You're still running all these processes as the same user, and they all have full access to the X server they run under (this is also a criticism against user separation on a single X session in general). The only thing it protects against is a website crashing the browser, where only that tab and its process would go down. Which.. you know.. should a webpage even be able to do that?

But what annoys me the most is the sheer amount of memory that all of these take. With all due respect all of you browsers, I am not quite prepared to give 8 fucking gigabytes - half the memory in this whole box! - just for a dozen or so tabs. I shouldn't have to move my web browser to another lesser used 16GB box, just to prevent this one from going into fucking swap from a dozen tabs. And before someone has a go at the add-ons, there's 4 installed and that's it. None of them are even close to this complete and utter memory clusterfuck. It's the process separation. Each process consumes half a GB of memory, and there's around a dozen of them in a usual browsing session. THAT is the real problem. And I want to get rid of it.

Browsers are at their pinnacle of fucked up in my opinion, literally to the point where I'm seriously considering elinks. Being a sysadmin, I already live my daily life in terminals anyway. As such I also do have resources. But because of that I also associate every process with its cost to run it, in terms of resources required. Web browsers are easily at the top of the list.

I want to put 8GB into perspective. You can store nearly 2 entire DVD movies in that memory. However media players used to play them (such as SMPlayer) obviously don't do that. They use 60-80MB on average to play the whole movie. They also require far less processing power than YouTube in a web browser does, even when you download that exact same video with youtube-dl (either streamed within the media player or externally). That is what an application should be.

Let's talk a bit about these "complicated" websites as well. I hate to break it to you framework web devs, but you're a dime a dozen. The competition is high between web devs for that exact reason. And websites are not complicated. The document itself is plain old HTML, yes even if your framework converts to it in the background. That's the skeleton of your document, where I would draw a parallel with documents in office suites that are more or less written in XML. CSS.. oh yes, markup. Embolden that shit, yes please! And JavaScript.. oh yes, that pile of shit that's been designed in half a day, and has a framework called fucking isEven (which does exactly what it says on the tin, modulo 2 be damned). Fancy some macros in your text editor? Yes, same shit, different pile.

Imagine your text editor being as bloated as a web browser. Imagine it being prone to crashing tabs like a web browser. Imagine it being so ridiculously slow to get anything done in your productivity suite. But it's just the usual with web browsers, isn't it? Maybe Gopher wasn't such a bad idea after all... Oh and give me another update where I have to restart the browser when I commit the heinous act of opening another tab, just because you had to update your fucking CA certs again. Yes please!

Data Disinformation: the Next Big Problem

Automatic code generation LLMs like ChatGPT are capable of producing SQL snippets. Regardless of quality, those are capable of retrieving data (from prepared datasets) based on user prompts.
That data may, however, be garbage. This will lead to garbage decisions by lowly literate stakeholders.
Like with network neutrality and pii/psi ownership, we must act now to avoid yet another calamity.

Imagine a scenario where a middle-manager level illiterate barks some prompts to the corporate AI and it writes and runs an SQL query in company databases.
The AI outputs some interactive charts that show that the average worker spends 92.4 minutes on lunch daily.
The middle manager gets furious and enacts an Orwellian policy of facial recognition punch clock in the office.
Two months and millions of dollars in contractors later, and the middle manager checks the same prompt again... and the average lunch time is now 107.2 minutes!
Finally the middle manager gets a literate person to check the data... and the piece of shit SQL behind the number is sourcing from the "off-site scheduled meetings" database.
Why? because the dataset that does have the data for lunch breaks is labeled "labour board compliance 3", and the LLM thought that the metadata for the wrong dataset better matched the user's prompt.

This, given the very real world scenario of mislabeled data and LLMs' inability to understand what they are saying or accessing, and the average manager's complete data illiteracy, we might have to wrangle some actions to prepare for this type of tomfoolery.

I don't think that access restriction will save our souls here, decision-flumberers usually have the authority to overrule RACI/ACL restrictions anyway.
Making "data analysis" an AI-GMO-Free zone is laughable, that is simply not how the tech market works. Auto tools are coming to make our jobs harder and less productive, tech people!
I thought about detecting new automation-enhanced data access and visualization, and enacting awareness policies. But it would be of poor help, after a shithead middle manager gets hooked on a surreal indicator value it is nigh impossible to yank them out of it.

Gotta get this snowball rolling, we must have some idea of future AI housetraining best practices if we are to avoid a complete social-media style meltdown of data-driven processes.
Someone cares to pitch in?