I'm still pro-human, as it were. Algorithms, even the newest ones, still need hardware to run on. If I remember correctly, every time a reigning world champion had been playing against a machine (for example Kasparov versus Deep Blue), the machine had been either a general-purpose top-of-the-notch supercomputer or a custom built device with massively parallel processing technology crafted specifically for playing chess and nothing else. And even in these cases the outcome was far from predetermined. I understand that today's regular computers are a lot faster and have a lot more memory than they were and had some twenty years ago. But I still think that they are not quite there yet to crush humiliatingly a grandmaster level chess player.

Besides, what is an estimated elo of a chess algorithm? Any chess match is time restricted, and would the algo show the same elo on some decrepit hardware given the time limitations for making a move? In my view, an estimated elo of such an algorithm is more hype than reality.