--- /dev/null
+<!-- TITLE: AI is not bad, tech in general is -->
+<!-- DATE: 2025-07-11 -->
+<!-- SUMMARY: AI cannot be considered bad, the problem is just overhype, marketing and this stupid money pursuit as always. -->
+<!-- TAGS: rant,not-code -->
+
+<p>
+ Looking at my first two posts this may feel like I'm avoiding at all costs to talk about actual tech and I'm going
+ for some kind of "capitalism bad" agenda or something. This is not my objective and this will <s>possibly</s> be my last
+ rant towards this. See, for quite some time I had some stuff stuck at the back of my throat I just wanted to write it
+ down somewhere.
+</p>
+<p>
+ I grew up messing around with computers, video games and eletronics in general. I wrote my first line
+ of code when I was around 13 (more than half my life) and never stopped since. One of my main sources of joy is
+ writing and talking about code, watching videos about it, learning, studying, seeing about areas within tech I could
+ never fully understand due to limited time (job+personal life takes a big chunk of it). But for years now I'm just
+ tired. Full tired. Not because I dislike coding now, not at all. I still enjoy writing and doing my own projects. My
+ problem is with everything else.
+</p>
+<p>
+ If you're not part of the tech community or don't work on tech, it might be quite shocking when I say that tech is
+ one of the most toxic and stupid areas out there -- not because of tech itself but because of how it presents itself
+ towards the market. See, there's this whole notion of "we need to be innovative" or something similar (every company
+ has it's own "motto") that actually can translate to "we need to make way more money". At first you may think "Oh,
+ that's normal. It's a job and a company, the main objective is indeed money" and you're right, but the biggest
+ difference is the shit ton of money available in tech and this makes ALL the difference.
+</p>
+<p>
+ As of 2025, out of the <a href="https://www.forbes.com/billionaires">10 richest people in the world, 7 is from tech</a>.
+ The meaning of this is that tech has a SHIT TON OF MONEY available, with a bunch of rich guys trying to get more money
+ out of it. This makes tech an insufferable extremely competitive field, meaning that people is always striving to find
+ one single scenario they can grow on. It's really easy to earn money in tech! It's even easier to go bankrupt. Most
+ startups out there create a product doing something -- always claiming it's some kind of innovation, of course -- and
+ start having customers until they can capture VC from some backer. Most startups NEVER get to actually make money from
+ their products, everything comes from VC. That means the company and the product itself becomes a shit show, the
+ mission is not "let's make it better, we need to build something good!", but "let's make it better, or else
+ shareholders won't invest on us again!" -- and this impacts DEEPLY on how software is developed. For years now most
+ commercial software out there was not written with you, the user, in mind, but the shareholders. Reason?
+ <a href="https://jsdaj.com/blog/you-most-likely-dont-need-aws">You don't pay our AWS account</a>, they do.
+</p>
+<p>
+ So the cycle has been always the same: one comes up with multiple ideas (not everyone makes it first time),
+ implements each one of them in the most shitty and rushed way possible (we need to be quick, or else someone can
+ implement it first!), tries to push it down everyone's throat (if we have a good sales and CS team, that means
+ engineers can work in more features instead of fixing stuff!), tries to capture VC ("we're a new startup focused on
+ innovation" I'm sure you heard that before), get's VC money, uses all VC money (our software is really shitty <s>
+ because we rushed it</s> for some reason no one knows, maybe we need more engineers? yeah, totally, we have VC),
+ shareholder gets angry and threatens to stop pumping money (don't worry, we'll have everything you need here), pushes
+ harder towards some imaginary deadline + starts cutting costs (layoff, suddenly everyone cares about infra, etc), and
+ either it works for some more years or it goes bankrupt sometime after.
+</p>
+<p>
+ But, from times to times, there's some mystical and "innovative" new shining piece of technology that, whatever shit
+ you do with it, some shareholder will give you money. That means the first steps I described above can be virtually
+ skipped (if this technology is innovation by itself, I don't need to be innovative! That's genius!), which results in
+ a bunch of people that pretend they have any idea of what they're doing but in fact they're just playing the same
+ game as everyone else and are hoping it works for as long as possible. The current "entrepreneurship silver bullet"
+ is (you guessed it) AI.
+</p>
+<p>
+ AI is not, by any means, bad tech. It's a whole field within Computer Science with a bunch of amazing solutions
+ involved -- from hard problems optimization (such as Lossless video scaling, rendering frame prediction, analysis of
+ genome data, etc), to computer vision, voice recognition, video game analysis, video-game AIs and more. For many
+ years now we've been using AI daily in some way or another, waaay before LLMs were a thing. The reason why people
+ didn't refer to them as AI that much is probably due to the <a href="https://en.wikipedia.org/wiki/AI_effect">AI effect</a>, and
+ the reason why LLMs are still being referred as AI is because of marketing. That's the only reason, marketing. LLMs
+ are very cool indeed: they're basically a text prediction machine. If your data can be parsed as text, they will most
+ likely be useful in some way. And not only that, but thanks to LLMs 1. AI fields are being more recognized then never
+ before and 2. a lot of <b><i>real</i></b> innovations were accomplished on the way to achieve what we have today. That's
+ marvelous... but the problem with LLMs are the overhype and the instrumentation of this overhype as a market maneuver.
+</p>
+<p>
+ Companies such as Anthropic and OpenAI lives from LLMs. And they need every penny possible because LLMs are <i>really</i>
+ pricey. There's just NO WAY they'll let the hype die. That's why every time the market is cooling off SUDDENLY there's
+ a new "OpenAI Product X" or "ChatGPT X.X -- now surpasses everyone's benchmarks" or even "Sam Altman says 'we
+ achieved AGI, but we're afraid to release it'"... really? This is bullshit. If you take any real researcher that
+ doesn't work for any of those companies the predictions and conclusions are always astronomically different. People
+ tend to forget that Sam Altman's <b>is NOT</b> a researcher. His an entrepreneur. While yes, he went to Stanford for CS, he
+ never graduated and never worked with tech directly. His first job was "being CEO of Loopt". Now you tell me, you
+ really think this guys has any idea what the hell is actually going on with researches? You really think his predictions are
+ accurate? That's the same as asking <a href="https://en.wikipedia.org/wiki/Chris_Kempczinski">Chris Kempczinski</a>
+ to make a Big Mac, that won't happen. Different jobs. They don't need to know what the shit they're saying, they just
+ need to sell. And this is not a Sam Altman-only thing, or AI LLM-only thing -- this is how ALL the industry works:
+ 1. have some (usually shitty) idea, 2. get tons of money, 3. hire a bunch of people, 4. sell it as a technical
+ panacea for as long as you can.
+</p>
+<p>
+ AI is not bad tech. The problem is that tech is bad. I'm not talking about tech itself, but the market around it.
+ It's just some stupid billionaire's race. People that aims only for money and only that. One may argue that it's always
+ been like that -- and that's true. The difference is that now with all the tools available this can (and will) be
+ really harmful for the field (+everyone's mental health in the industry is shit). I'll probably never stop coding or
+ messing up with old junk circuits around the house. I'll never stop thinking about software architecture, and how
+ software is cool and can help lots of people. I'll never stop trying to help people to learn software, or help people
+ through software. But I can't imagine working on the industry for more than 5 years now. It just sucks.
+</p>