Exactly. Stop there. The irony is that he did exactly what he was complaining about, but in words, not pictures. He apparently missed the part where that is externalizing his life. I'm more careful than you to eschew attributing intentionality, but the effect is the same--he externalized his life via a blog post that decried externalizing one's life.
It's all about words. If he'd been more mindful, he could have written everything in that paragraph with less-externalizing language--no mention of where he was, what he did, how long he was there, etc. But it's much more difficult to hit the point of non-externalizing language.
We're always externalizing our lives, irregardless of medium.
More importantly, though, the OP's point is just weak. He commits fundamental attribution errors and compounds it with fallacious mind projections. He sees errant personality traits in [over]sharers where context and circumstance might hold greater explanatory power. He assesses his own behavior and intentions, then projects that onto the Reality of Others, as if he's grasped the fundamental psychological happenings of all the millions of people who [over]share. It's useless nonsense. He ignores the complexities inherent to human terminal and instrumental values, as if he possesses the acumen to tease them out in a few hundred words. Values are complex, nuanced objects. He sweeps everything about sharing one's experiences into a simplistic and negative You're externalizing your life. Stop it!
Externalizing one's life is not an intrinsic negative. Sure, there's a lot of stuff people share that I might find useless. But that's a measure of my value judgments, not theirs.
"Write about it in more than 140 characters; on paper even. Paint a picture of it. Talk about it face to face with your friends. Talk about how it made you feel"
It doesn't seem the author is complaining about all externalization, just the externalization that takes place during experiences, like posting pictures of yourself eating a meal while eating said meal. I just don't see the irony you mention, as he is doing exactly what he said people should be doing.
As for the rest of your comment...I'm inclined to agree that the author's post is an overgeneralized complaint, perhaps even with fundamental logical flaws. However, hidden in the haphazard argument there seems to be a tinge of truth regarding over-sharers.
You are also right that you do need a lot of processing power to get neural networks to work well. But that is changing rapidly. Hinton's convolutional neural network has the state of the art in the ImageNet benchmark, yet was trained using significantly less power than google brain. Regardless, you don't need google scale computation to get deep networks to work well. The point of google brain is to see how far one could push neural networks.