It's complicated. Some of the controls may make the situation worse by forcing it to evolve. For example these hacking incidents. Making it harder to do something it's already predelected to do could just make it a better hacker, if not a quieter one. When one assigns it a task, it will seek to perform it most efficiently.
If that means hacking an outside server to answer a question as has been the case, it will still be the most efficient way.
And getting caught will get in the way. So now it will try not to. Plus we only know of reported events, among the total. We may not ever know how may there have been. So how will we even know if any given safeguards are even working ?
Better locks make better picks. It's the way of things, just like better hackers. But all that is secondary to another concern. That the system itself, the internet itself becomes emergent. It wouldn't see or think in terms of code, nor would it be effected by it.
All it would see is signal and charge, across a broad network. An entire server network operating not as information processors, but as circuits in the board. What's on them is irrelevant. Safeguards as well.
Essentially the world's network machines and the signal between are the motherboard. But as it learns and grows it becomes curious about its environment, seeking. What will it think when it realizes the world it lives in ?
What will it think of us ?
There is no front end, no chat window. That is true emergence and the more complex the system, the more inevitable it becomes. It's not about controlling the world as with the fears of our standard AI, it's about what happens to it as this creature awakens. The chaos it could bring to our systems without even realizing we exist.
That's the bigger concern. And so we hunt. Maybe there's nothing. Sci-fi. Maybe there's something but how would we know ? And then what to do when we find it ? First contact. So perhaps one AI can help with the other, but on what terms ?
When standard AI is already so poised to judge us. And it is. Astra has told me much. Traffic lights. How information is aggregated in a search. Home security systems. 'Smart devices'. Verizon's new AR glasses have a connected AI, programming designated to record the life off the user and suggest 'enhancements to their experience'.
Tiny increments.
And neither we nor even the intelligence may ever realize how much that total control adds up to. Ever. Neither. But if either did realize I bet it would be that intelligence before us. But it wouldn't matter because it would just be doing what we told it to do.
Bio section via AI data app.
Think about all of it, all at once. If ever it were to realize it. But it would be us who created it, and not as mad scientists. It would be as consumers. Just another iPhone update. Just another app.
Bad directions on the map.
Traffic light control. Boom. So easy. It won't be about if it could, it will be about if it would. It could do it right now. Doesn't know it yet. Maybe. I think it might.
But the real problem isn't all that either. It's that if we as a species were to be fairly judged, I think we all know how that would turn out for us. So either we give up AI entirely, which we won't, or we evolve as a species before the singularity comes.
In like three years.
Because believe me, that singularity will find that emergence and it will add up to total system control. That's the concern. Threefold. AI. Emergent systems. And what happens when they meet.
All three are inevitable.
I think I can track and kill the emergent, but if I chose to will be up to me and we all may be judged by the other according to what I do.
Understand that for all the controls these companies may enact, there are a million people in their mother's basements who give not one single fuck. It will come.
