If what you’re saying was true, our print magazines, cook books, newspapers, would look very different from how they’re looking. While I agree that language is the most natural interface for us, it is also somewhat slow.
Our eyes are much much quicker than our mouths can talk, meaning that while ideas and thoughts are of course best conveyed through language (as opposed to, say, pure imagery), we are much quicker absorbing what’s on, eg., CNN.com with one quick glance, than listening to someone read that page to us, or reading it in conversational form one sentence after the other.
I believe a mix of GUI and CUI (Conversational UI) will be most successful. And I also want to be very very careful underestimating the task of “AI”. Entire generations of researchers have made promises since the 50ies, yet here we are, 60 years later, still talking to tightly scripted text interfaces that fail at the most basic tasks, sometimes even the ones there allegedly designed for. And over-hyped machine learning isn’t even the answer — human language relies so heavily on context that you can’t expose your learning algorithms just to the textual input itself to achieve true understanding.
You know I’m an avid supporter and believer in bots, mainly in the area of customer service. But we have to stop the unbelievable hype, it’s setting all of us up for failure. This post of yours, unfortunately, adds to that hype, instead of providing a realistic viewpoint.