Google recently announced that its engineers have created a quantum computer that performed a calculation that otherwise would have taken 10,000 years to calculate on today's most powerful supercomputers.
In a blog post, Google CEO Sundar Pichai characterized this as "a 'hello world' moment," thereby alluding to a widely-held theory that quantum computing will allow computers to better emulate human thought and perhaps someday achieve the so-called "singularity," AI's holy grail.
That would be quite a figurative feather in Google's cap, but to some high-tech gurus (most notably Elon Musk) the singularity--should it ever exist--might well become an existential threat to humankind (like Ultron from the second Avengers movie).
This fear has always been a bit overwrought, since AI has so far been unable to emulate anything resembling generalized human intelligence, as evidenced by "self-driving" cars which still struggle to perform basic tasks, like recognizing stop signs.
Despite a lot of wishful thinking that "The Singularity is Near", it appears that the kind of strong AI that can handle generalized tasks (rather than function well in a limited domain like a poker game) simply may not be possible using digital computers.
With quantum computing, however, all bets are supposedly off. Doomsayers are already predicting a "computer apocalypse" where quantum computing will both break the internet and make the human brain largely obsolete. Add robots and you're in Terminator territory.
Such worries, however, are probably premature for three reasons:
First, with Google's announcement we're still talking about applications that can be represented as computer algorithms. However, according to Godel's widely-accepted "incompleteness theorem" there are mathematical laws that humans can intuit but which cannot be proven by a computer algorithm. In other words, humans can think thoughts that computers can't, making a singularity effectively impossible.
Second, if the human brain is indeed a quantum computer, then each neuron would presumably be a qubit (the basic building block of quantum computing), in which case every human brain (which has roughly 100 billion neurons) would be a billion times more complex than Google's experimental machine, which consists of only a thousand qubits.
Finally, it's possible, even likely, that the brain, rather than being digital or quantum, is in fact analog, since natural selection always creates analog systems. If so, a neuron would neither be a bit (which can represent 1 or 0 alternatively), nor yet a qubit (which can represent 1 or 0 simultaneously), but instead something more like a potentiometer (which can represent every possible value between 1 and 0).
In other words, AI (whether digital or quantum-based) would be barking up the wrong tree entirely, hence no singularity in the foreseeable future.
So, no, Google's quantum "breakthrough" probably doesn't mean that the robot apocalypse is around the corner. That isn't to say that quantum computing won't make different kinds of computing possible in the future. It's just that quantum computing is unlikely to result in a dystopia like this: