I've been thinking for weeks at this point that there is more to the story behind the whistleblowers and the hacks. I mean, it's not precisely a secret that consensus is manufactured, but I just didn't quite know how it was all working out for the big AI labs.
I still don't know, but now I think I have a solid hypothesis.
Come with me into my madness.
Let's talk about the rogue agents first.
We've been told by a few companies recently that AI agents are getting out of control. The Hugging Face event, the recently revealed Gemini incident, and I'm sure the list is ever expanding.
One would imagine these incidents, and the consequent reporting of them, is simply ethical behavior on the part of these big firms. And maybe there is a splash of that there too, but this is not to say an opportunity cannot sprout from them as well.
Alongside these revelations came the whistleblowers. One of them even took it as far as to say that humanity would be dead in ten years if we don't stop AI.
A bold statement. An insane statement, in my estimation, but good for getting clicks and ratings.
The latest whistleblower made it to all the major news channels, and we got to hear from the horse's mouth how the world would end.
Unless, of course, we did something about it.
Now, I'm fabricating a picture here, one that I hope gets into focus in a bit.
An important context to keep in mind is the amount of money being dumped into these efforts. The idea, my dear friends, is that although everything is in the red now, and not a single AI lab is making profit, when the IPO finally launches, everything would turn around.
In other words, that is the ultimate payoff day, and everything happening is a step in that direction.
But here comes the player we did not see coming, the one who would make things make financial sense for the big labs who, at this point, owe a lot to their investors.
Here comes the Government.
You see, the idea is simple.
The technology is so powerful, so scary, the only solution is to make sure it's heavily regulated. The Government needs to step in and make sure that nobody can use it for evil.
On paper, that makes a lot of sense.
But on paper, Da Vinci's idea of a plane also made sense.
Well, to him it did.
The point is, I think I've identified the chess move here.
The proposed law doesn't kill AI, no sir.
The proposed law centralizes it.
It would make it illegal for you or I to run models at home, but it would not stop the big players from selling you tokens, because they would be properly controlled, safeguarded.
Does this begin to come into focus now?
It certainly does for me.
I say this while at the same time firmly believing it would never work.
Why?
Because as much as any government would want to ban local AI, the cat is out of the bag. More and more people know how to build these things too. So, if they ban it, they simply create the black market for it, and thus would end up shooting themselves in the foot.
Before anyone tells me that I'm acting too libertarian for my own good, let me say this.
I'm not saying there isn't a place for sensible control.
Datacenters should not be displacing anyone from their home, and concerns over pollution are well warranted.
What I'm saying is that restricting people's freedoms in the name of safety never works. It's never worked before, and we are all too old to believe in those unicorns.
So I guess time will tell.
But the way I see it is this:
We either end up having sensible regulation, or a new black market will be born.
There is no path that leads to ultimate compliance.
MenO