It's hilarious how gleefully we're trying to bring about our own demise.
We're building better AI's at an incredible pace. And at some point the AI will take over building better versions of itself.
Once that happens, it could easily develop millions of iterations of this Lego robot until it's effectively a real T-1000.
Then the only 2 possible outcomes are that it sees humanity as a threat so it eliminates us...
OR
It's so powerful that it doesn't even consider us a threat, which probably means that it runs the world and we are slaves to it.