BS Things Big Companies Do to Control Us
And How OpenAI Is Trying to Do It With Artificial Intelligence
Keeping an eye on AI has become my obsession in the last year and a half.
For starters, I see it as a public safety issue. Anything that can reshape the job market overnight and aid in the commission of criminal behavior needs all eyes on it that we can get.
A world in which 60 percent of jobs are impacted, and people are laid off because their marketable skills have become obsolete can lead to desperation.
Desperate people do desperate things.
Throw in uber-realistic voice-cloning, image generation, and video creation capabilities. Enter:
Underage pornography AI images depleting law enforcement resources, thus preventing the investigation of real-life crimes;
Predators creating nude deepfakes to embarrass and humiliate;
Disgruntled workers targeting employers with voice clones of their bosses saying racist things
It’s not hard to see Total Anarchy becoming one of our possible futures.
I don’t pretend to have all the answers. AI literacy is our best bet.
But, I know what the answer isn’t. It isn’t handing this technology over to one or two key power players.
Unfortunately, a new patent filed by OpenAI shows this is exactly what they’d like us to do.
The patent is for a process that generates docstrings—natural language descriptions of computer code. At first glance, this patent might seem harmless, even helpful. But dig a little deeper, and you’ll see why it’s anything but.
This isn’t about OpenAI protecting its work. This is about a big company laying claim to processes and techniques that are foundational to AI development as a whole.
By doing so, OpenAI isn’t driving innovation—it’s locking it up and throwing away the key. Here’s why you should be concerned, along with what OpenAI will likely tell you to defend itself.
A Wolf in Sheep’s Clothing
The patent claims to protect a method for generating docstrings using machine learning. Sounds niche, right?
The reality is that the steps outlined in the patent—training models, generating outputs, and refining results—are standard techniques in the AI playbook.
These are the kinds of things every AI developer does to solve problems.
But now, OpenAI is attempting to fence these processes off, making it harder for others to experiment and innovate without fear of stepping on legal landmines.
Why This Patent Should Worry You
1. It’s Broad and Vague
Patents are supposed to protect specific inventions, but this one seems to describe a general recipe for solving machine-learning problems. Train a model, generate outputs, evaluate them, and refine them.
Sound familiar?
That’s because it’s how AI has been developed for years. The problem is that OpenAI’s patent wraps a common process in legal red tape, making it harder for others to do what they’ve always done.
What OpenAI Will Tell You:
“This isn’t about the general process—it’s about our specific implementation for docstring generation. The patent only covers this exact method, so others can still innovate freely.”
Why That’s BS:
The claims may sound specific, but they’re broad enough to give OpenAI room to claim infringement on almost any similar process. And when smaller companies can’t afford a drawn-out legal battle, even the risk of a lawsuit can push them out of the market.
2. It Stifles Innovation
By claiming ownership over techniques that are broadly used in AI, this patent creates a chilling effect. Smaller companies and independent developers, who don’t have the resources to fight legal battles, might avoid working in this space altogether.
Instead of fostering competition and new ideas, this kind of patent discourages them. It’s not about protecting progress—it’s about protecting market dominance.
What OpenAI Will Tell You:
“Patents like this ensure we can keep investing in cutting-edge research. Without protections, competitors could steal our innovations without putting in the work.”
Why That’s BS:
The research that led to this patent is built on decades of shared, open-source knowledge. OpenAI’s work wouldn’t exist without that collective foundation.
To now claim ownership over such broadly applicable methods feels like taking advantage of the system rather than pushing it forward.
3. It’s a Corporate Power Grab
Let’s not kid ourselves—this isn’t about protecting a groundbreaking invention. This is about OpenAI and its corporate partners, like Microsoft, using patents as weapons to corner the AI market.
They don’t just want to compete; they want to control who gets to play the game. And this patent is just one more tool in their arsenal.
What OpenAI Will Tell You:
“Patents are a way to level the playing field and protect innovators from being overrun by tech giants with more resources.”
Why That’s BS:
OpenAI is one of those giants. Backed by Microsoft’s billions and entrenched in the AI market, OpenAI doesn’t need to "level the field."
Instead, it’s building walls to protect its territory, locking out competitors and smaller players who can’t afford to fight back.
4. It Hijacks Shared Knowledge
The AI field thrives on collaboration. Techniques like transformers, natural language processing, and code generation have been developed through decades of shared research and public investment.
OpenAI’s patent takes that collective progress and slaps a “Do Not Enter” sign on it. By privatizing ideas built on the backs of many, they’re claiming ownership of something that should belong to everyone.
What OpenAI Will Tell You:
“We’re not patenting shared knowledge; we’re patenting a specific application of machine learning that we developed internally.”
Why That’s BS:
If the techniques used to generate docstrings are so unique, why do they look like standard ML workflows? The patent stretches far beyond any “specific application” and puts basic, community-developed practices under OpenAI’s control.
5. It’s Not Even Enforceable
Here’s the kicker: the patent is so broad that it’s practically unenforceable.
A minor tweak in methodology—changing the architecture or using a different evaluation criterion—could sidestep the claims entirely.
So why file it?
Because the mere existence of the patent can scare off competition. It’s not about enforcement—it’s about intimidation.
What OpenAI Will Tell You:
“We only intend to enforce this patent in cases of clear and deliberate infringement. This isn’t about scaring anyone; it’s about protecting our intellectual property.”
Why That’s BS:
History shows that patents like this are rarely used with such noble intentions. Even if OpenAI doesn’t actively sue, the chilling effect alone is enough to stifle competition. The system itself is designed to favor the patent holder, and OpenAI is exploiting that.
Why This Matters
Patents like this don’t just harm competitors; they harm progress.
The best ideas in AI don’t come from walled gardens—they come from open fields where everyone is free to experiment and build on each other’s work.
OpenAI, for all its talk about “beneficial AI for humanity,” is doing the opposite by fencing off foundational techniques and discouraging collaboration.
This isn’t just an OpenAI problem. It’s a Big Tech problem. Companies that claim to champion innovation are the same ones filing overly broad patents, hoarding talent, and creating barriers for everyone else.
And, too often, we let them do it because they package it as “protecting our investment.”
The Bottom Line
OpenAI’s docstring patent might seem like a small, technical issue. But it’s a symptom of a larger disease: the corporatization of AI research.
When big companies prioritize market dominance over shared progress, we all lose. Innovation slows, ideas are stifled, and the tools that could benefit humanity end up benefiting only the shareholders.
It’s time to call BS.
AI doesn’t need more barriers—it needs more bridges. And the sooner we stop letting big companies control the narrative, the better off we’ll all be.