The AI revolution is open source and it's igniting from the most unexpected place

Published: February 1st, 2025

Rethinking censorship: Why private control of AI, information, and social media is a massive danger.

In all the hoopla about the AI boom, these chatbots doing our jobs for us, getting better at it month after month, and the big shots slugging it out to see who can make 'em fancier, faster, smarter, and cheaper than the next guy, we'd forgotten something. Not some little thing either. Something darn crucial. Something that started out good, but took a real nasty turn early on, the worst kind, truth be told, yet somehow slipped right under our noses while we were all puffed up with joy and excitement about maybe spending two hundred bucks a month to use a product that, to kick things off, was promised to us as free for everyone, open source, and built to help all of humankind, no exceptions!

But for once, us regular folks, we got incredibly lucky.

A cold splash of water from China jolted those Western AI bigwigs and reminded them of those old promises they brushed aside, twisted around to set up their usual endless gravy train for the tech overlords of our century. OpenAI, born and funded with a solemn vow to create AGI open source, to head off the risk of such a big invention ending up locked down by just a few who got there first, maybe even some folks from beyond America's borders, well, it quickly veered off course into the most unholy of directions. It became the perfect example of the closed-off, money-hungry, secretive AI company it was supposed to prevent and fight, if needed, alongside its special partner in crime, the slick and seasoned Microsoft, who else?, with decades of know-how in navigating the murkiest waters of monopolies and the shadiest, borderline business practices. And of course, this greedy pair didn't miss the cherry on top: getting a helping hand from the first Trump administration, who jumped at the chance to stop the most powerful GPUs from being sold to those Eastern rivals, just to make sure they couldn't even dream of catching up to or beating US companies in the AI game.

Yet, for once, as I said, we regular folks got incredibly lucky. The feast these guys had laid out for themselves got crashed in no time, and with hilarious style, if you ask me, all thanks to the lowest, most bullying tactics of the usual suspects. If this were a movie, what's happened lately would get the classic, cathartic sigh of relief, like when karma finally steps in and the bully gets what's coming to 'em.

"Hunger sharpens the mind," we say back home. Old saying, goes way back to Roman times when maybe other bullies of their day made the same mistakes and found out, millennia ago, that "Fames artium magistra" (hunger is the teacher of the arts), reflecting how need, especially hunger, pushes people to find creative ways to survive, like what China cooked up with DeepSeek. AI as powerful as the American empire's, but a hundred times cheaper, and made without those powerful GPUs Trump kept from the Chinese.

So, a feast ruined, just like that. Picture the guests already seated, forks and knives ready, napkins tucked in, mouths watering. Even the less brazen ones among them, who were a bit taken aback by OpenAI's nerve asking for 200 bucks a month for a chatbot, had calmed down seeing how many folks were biting, and that young Sam Altman even got away with whining that they were losing money even at 200 euros a month, poor dears, making it crystal clear how ready the market was to get fleeced even more. And then BAM! The treacherous arrow strikes! Damn China! And why bring back that open source nonsense? Why now, just when young Zuckerberg over at Facebook had drilled into public opinion that open source models would never be more than a broken toy for broke folks looking for bargain-bin AI? What a storm! 'Cause it's a storm hitting this feast, not just a single arrow. Day after day, the details, the products, the leaks keep piling up. It's a barrage of news, real deals, apps in the stores knocking OpenAI's off the shelf, other models from other companies, like Gwen 2.5, joining DeepSeek, also open source, also Chinese, unfolding right before the stunned eyes of the guests at the feast of the century a stark, well-built, and sweeping truth that leaves no room for doubt: the party they'd all been prepping for for years is definitely over, kaput, or maybe, worse, it's moved somewhere else!

So, no surprise at the wild reactions we're seeing these days. The stock market dips, the cries of "wolf, wolf" on social media, the ghosts of Sputnik past resurfacing, the meltdowns of folks dusting off old paradoxes (like Jevons' one) to prove their GPU monopoly ain't in danger. Sure, all this will just lead to more AI use and adoption. But, and let's not kid ourselves, it's also dead certain that the gluttonous feast everyone in the US was expecting ain't happening anymore. And instead, we're gonna see a much fairer game, all to the benefit of us end users.

So let's enjoy these days, and cheer for what's going down, 'cause honestly, OpenAI's about-face was a tough pill to swallow back then, and not just for Elon Musk. But more than that, it was a straight-up insult to our natural intelligence. If the excuse of keeping GPT closed 'cause it was too dangerous for humanity was already full of holes back then, now that a model just as powerful is open source, we've all seen the real danger to humanity: just being democratically available to everyone, instead of big profits for one! Just super cheap work, done by powerful tools, for everyone to use. The most advanced AI, and what we got now is advanced enough, and getting more so month by month, is like finding a planet packed with 10 billion geniuses, all ready to work for humans for free. It's obvious that such a find makes folks greedy, wanting to hog it all for themselves, even those who started out with the best intentions of spreading well-being and sharing with everyone. It's on us for not paying attention to this, for letting OpenAI's flip-flop slide. It's on our governments for not making laws to stop AI supremacy by a few.

And I say again, for once we got lucky, a Chinese angel came down to save us and balance out a game that was looking to go real sour, but the dangers are far from over. There's another area where we're still in deep, deep trouble and where, sadly, nobody's doing anything, and the market's way underestimating it. I'll write a whole piece about how serious this is, but for now, let me just hint at the next huge danger we all need to pay close attention to: the censorship built into these AI models. This topic is way more worrying and serious than it might seem just mentioning it, which is why it needs its own article. Suffice it to say here that if on the open source front, an angel from China came down to save us all, and we're sure grateful, in the censorship game, alas, we all know that same angel turns right around, even he does, into the worst kind of demon!