# Use Always My Em-Dash: From Normative to Positive Reality
Published: 2026-02-11
Canonical: https://journal.xiaotianfanx.com/journal/use-always-my-em-dash-from-normative-to-positive-reality

On AI, authorship, authenticity, attention, and scarcity. Effectively a continuation of my previous posts [1, 2, 3, 4].

 

	Since adopting voice-to-text transcription, my writing speed has surged, on balance. In the past weekend, I probably "produced" many more words than I could have by typing with a keyboard. But today, although I'm still utilizing transcription now, I actually want to talk about a particular knack for typing. To some extent, it appears to be one of the cornerstones of my writing and thinking, and it is also inextricable from my understanding of human-AI relationships these days.

 

	Before I give the very knack itself (which is kind of trivial in this context), I'd like to go back to one of the main themes in a number of my previous writings recently: considering AI as a genuinely novel kind of entity—something that is not necessarily superior or inferior to human beings, but simply different—different, to some extent that necessitates us to consider it as a new kind of entity. We need to study this entity from a genuinely new perspective, moving beyond always using human beings as the only benchmark. By doing so, we can take the chance to pursue the possibility of discovering a new epistemology and potentially looking back at ourselves as human beings differently. 

 

	Admittedly, the "novel entity" stance itself can be obviously contestable and controversial, and it can be easily taken as suggesting that AI is a new kind of species or a new kind of conscious being, which is not exactly the case. I shall not ramble on and on again on the stance itself, as I've discussed it in many of the previous journals already. The point is, I'd like to zoom out for a moment from arguing for that stance or rationalizing from the for and counter angles. 



	Considering AI as a genuinely new kind of entity is simply a premise and a starting point that could guide our further research and understanding of what, in fact, happens in reality around us within the peripheral. In other words, it is a normative appeal that I'd like to promote, and I'd like it to potentially be accepted in the future (including continuously held by me myself in the future). Similarly, one of my other arguments on AI—in particular, generative AI content, and especially art and music-related content—is that if there is any fair "competition", then since AI is a genuinely new kind of entity, we should at least recognize the fact that AI creation is at least eligible to be aesthetically evaluated with an equal chance—nothing more and nothing less—neither privileged nor dismissed based on its origin. And I myself, as a musician, an artist, and a creator, am both aware and prepared for it.  

 

	However, that is nevertheless another normative argument (normative and positive in terms of "ought "vs. "is"). And this time, I'd like to touch on the positive potential and show you how I, as a creator and expresser, have also gone through the process of dealing with "AI anxiety." Eventually, in essence, I'd like to argue that human creation could retain value not because AI lacks consciousness (although it does, now), but because human attention is finite—and finitude, not biology, grounds authenticity. This position is compatible with, and indeed requires, treating AI as a genuinely novel entity.

 

	Alt+0151 on the numpad—this is the knack I was talking about. On a Windows PC, in any editor, using the system default input method, this keyboard shortcut produces arguably my "favorite" punctuation—an em-dash (—).  While I cannot recall when exactly I fell in love with em dashes, it was almost instantaneous when I learned their usage. Honestly, I can go on and on talking about why I love this punctuation mark, the em dash: It can be used as a substitute for a comma in some cases; It can be used to introduce an appositive or a list; It can be used to introduce a summary or a further explanation; It can indicate an interruption or a break, etc. And, most importantly, I use it to indicate the rhythmic design in a typed, linear text as if I were reading it for you. 

 

	Frankly speaking, what really matters with that particular kind of "rhythmic design" is how it reflects my thinking pattern. To some, I understand that my sometimes excessive use of em dashes and parentheses could seem overly parenthetical and conscious-flow, interrupting the reading experience and weakening the voice and central arguments. It is largely true, for sure. But for me, when there is no external requirement or specific framework I'm following (in other words, I'm not writing an essay or report at all to be evaluated by an audience), it is personal (although I do hope some people would be interested), and thus legitimate to truthfully reflect how thoughts are actually formed in my mind. In that sense, Alt+0151 is one of my purposeful moves to exert authenticity and authorship.

 

	However, my use of em dashes as a private and peculiar idiosyncrasy was unexpectedly breached and generalized by something called a GPT. Ironically speaking (although I believe I've mentioned it somewhere else), I indeed used to be called an "AI "in a kind of good way by some of my high school schoolmates before GPT was even a thing. When GPT broke out, and people started to recognize typical patterns in the text generated by LLMs, one of those vices is the excessive use of em dashes. Furthermore, together with the usage of em dashes, there was an entire collection of taboos and vices people recognized in generative text, including the frequent and formal usage of transitions, the use of certain words like "delve into," "tapestry," "profound," "myriad," "ubiquitous," etc., the tendency to be tentative and hedging when giving opinions, and the use of passive voice, etc. Unfortunately but also pleasantly, as you may tell from my writing history, those are almost by definition my style of writing.

 

	The unfortunate part, of course, comes from the fact that my style of writing was, from time to time, criticized for being too indifferent and distant, even in daily life. Just like an LLM. (And that's exactly why rhythm and vocalization matter, right?) But the pleasant part comes from a peculiar rationale: if it can be said to produce at least formally "good writing," then to some extent I may consider the criticism as a compliment—not to say my writing is really that good, but as a form of compensation, at least I get some sense of accomplishment. Jokes aside, although I don't reckon it is an essential question that would actually diminish or disqualify my writing, and I also don't think that even if something is AI-written, it is fundamentally inferior from a mere language perspective, there was a time when I questioned whether I should purposefully change my writing style to avoid being recognized as AI—although (again), I would assume still, for now, most people would not really consider these proses as LLM produced (even if, according to my experiment, with enough examples fed and proper prompting, some of the models are indeed capable). Anyhow, as you can see, my answer to the question is no—I will probably continue to use my em dashes on and on until I myself someday may get sick of it.

 

	You may wonder if, following this direction, I'm to continue on a relatively cliché argument for defending human agency and dignity, confronting an elephant in the room when it comes to the positive, actual happenings, i.e., the mass-produced generative AI content. That argument is to say that human creation, no matter how perfectly mocked by AI-generated content, will always have its place due to the fact that AI does not have a soul or a consciousness as we do. To some extent, yes. I am going to argue for a similar conclusion but from a distinct approach. Again, I'd also like to show that this argument is very much compatible with my other views: we should normatively accept and acknowledge AI as a genuinely new kind of entity, and acknowledge its creation to be evaluated given an equal chance with humans', simultaneously on the table. To be honest, from my perspective, it is not only compatible with my views but mutually necessary.

 

	Let's say we have already reached the moment when nobody can ever distinguish between human-produced or cerebrally produced content and generative content (with human-AI collaborative ones aside for the purpose of this thought experiment)—no matter if it's text, audio, image, film, or any other possible media we could think of. In that case, is there any solid reason for us to still pick a side? Normatively, as I said, given the equal chance on the same table, if there is indeed no way to distinguish the two, distinguish the nature of the creator behind a piece—even if some of us, due to different personal or communal reasons, still could prefer one over the other (and I would pressume, in most cases, there will be a preference for human content over AI content without a fundamental sociatal paradigm shift in taste)—then, positively also, the preference "sadly" matters not anymore. However, if at the end of the day, there are some ways to distinguish the nature of the creators—whether human, AI, or an alien—despite the fact that the product itself could be exactly the same (again, we include not the case when one is perceived to be "better," as there would other obvious reason to prefer on over another then), I then do think there is always a simple, solid reason that we may still positively prefer one over the other, a reason besides the inertia of preferring human content or based on whether the creator possess a soul or not, but a reason rooted in the very fact that AI is a genuinely novel kind of entity. 

 

	One of my main arguments for justifying AI as a new kind of entity is to think of its almost infinite amount of attention that could be devoted to an individual, given enough computational power. That amount of attention is unparalleled and impossible to give to or receive from another human being. Of course, that kind of attention fills or substitutes many gaps in human need, but it essentially also makes itself distinct from human attention due to its "abundance" as club goods. Therefore, based on a simple rule or value system resorting to scarcity theory, the always limited and finite human attention is thus unique. Despite the fact that I as well agree and hope we might develop another completely new market or economy based on agent or AI attention and interaction, the amount of and the need for human attention will still persist because of its scarcity. In other words, because human attention is finite, it becomes scarce—and in a world where AI offers infinite attention, this scarcity ensures the enduring value of human focus. Hence, it is justifiable for one to prefer scarce human attention, in some specific scenarios, over an infinite, platitudinous attention.

 

	Anyhow, from my perspective, this is by no means jarring with the potential for some others to prefer the infinite attention—exactly because of its infinite nature in comparison to a human being's. The point I'm making is that, due to such a distinction between human and agent attention, both sides could have their own reason to choose one over the other. In that case, it makes total sense for a creator (whether agent or human) to create particularly for its audience in that new scenario (you may be interested in the OpenClaw/Moltbook hype to take a glance at the prospect of agent-for-agent creation). In practice, another reason I am still writing these journals cerebrally, or at most using voice-to-text transcription and grammar correction, besides practicing writing as a method of thinking and existence-check, is precisely built on top of the fact that LLMs can already transform my even sketchier ideas, jotted or transcribed, into prose with my specific style of writing in a second. Inevitably, when we talk about human attention, we are talking about our finite time and life—then, it revolves back to the matter of our distinct nature as mortal beings, and the question of whether finitude puts weight and value on our existence (shout out to Heidegger). Following this thread, since I acknowledge so much my own mortal nature and your very mortal nature, it makes sense for both of us to continue acknowledging the value of each other's limited time and attention to whether to create or experience a piece—given you'll be willing to believe that I am willing to put my stake on the balance. Eventually, scarcity alone doesn't create worth, but finitude does as it transforms attention into an irreplaceable existential stake.

 

	That said, one might still argue that this is nevertheless a passional choice or belief decided out of my passional nature, and I have no evidence to guarantee anything. And yes, it's certainly the case—I do not believe anyone would be able to guarantee anything on this matter, even for themselves. However, to potentially console the still worried, I may briefly introduce another point of view that may be theoretically more "realistic," aka a Marxist-ish perspective. (Disclaimer, as usual: I am in no way a Marxism expert. My only intention of thinking from this perspective here is to demonstrate another explanatory framework and prospect that is potentially compatible with my understanding—even if it is a completely different starting point.) 

 

	The story goes like this: the groundbreaking progression and development in AI can be deemed as an essential leap in human productivity—aka the force of production—as a whole from that perspective. However, given that productivity soars (or a potential leap, if not yet realized), our macroscopic relations of production—the economic and social structures—lag behind and haven't yet even started to develop themselves to accommodate the new level of productivity—and that induces hindrance, tension, contradiction, conflict, and displacement. As a result, according to Marxists, there may eventually be a social revolution—which I cannot tell a bit if it is even possible anymore on this planet. Anyway, deviate not too much from our main thread. A topical example could be the latest release of the Seedance 2.0 video generation model by Bytedance. While I shall not ramble on about how game-changing the model itself is, my point of focusing on a video generation model here is to showcase how the improvement in productivity—a model and new entity that no longer only generates the raw footage after taking prompts or reference but covers all domain knowledge and action from director to cinematographer to editor to sound designer—could subvert an industry of well-defined SOPs. Everyone knows the game has been changed (and if not by this particular model, what's next), yet it is impossible for the industry to adjust its relations of production in one night—and thus, people feel punched in the face abruptly and thrown into limbo with absolute uncertainty. Some accuse it of being another AI hype of the optimists, yet I barely observe any blind optimists there among the actual professionals in the film and motion graphic industry—but only one or another individual forced to confront the "cruel" facts and forced to choose "optimism"—as the cost of quitting is even unacceptable with the current relations of production. 

 

	In that sense, returning to my stance, the positive resolution between human creation, AI creation, and hybrid creation (or production) could only happen when the relations of production are no longer stagnant but open to change. And certainly, what I described in terms of justified preference for either side could be at most only one of the potential outcomes of evolved relations of production. But to have that scenario realized—if we, you and I, consider it (or any other scenario) acceptable, desirable, and worthy of pursuit—we may nevertheless need the faith and normative premise shared in the first place. Even if you do not agree with my envisioning of the future on this matter, normatively or positively, I would argue it is nonetheless necessary to make a choice and to have the desire rooted in our finite nature, sooner or later. 

 

	I burnt my time and attention on writing, for this matter, and for this piece. And the burnt is my very finite, mortal life into authenticity.

 



 

P.S.

 

	Towards the second half of this journal, I actually switched back to typing from transcription. Afterall, the transcription can only be conducive to the efficiency in terms of putting words onto the canvas, but not the "efficiency" of thinking.