INTERVIEW: ‘Destination Undefined’ is ‘too real to be taken as fiction’
Photo: Changshuo Liu’s new show is called Destination Undefined. Photo courtesy of the artist / Provided by DARR Publicity with permission.
Changshuo Liu’s new play Destination Undefined, presented by Cellunova Productions, brings a genre to the theater stage that is often missing: science fiction. But Liu’s unique brand of sci-fi feels eerily prescient, as if the commentary is not about a future on the horizon, but on the troubling signs in the present day.
Destination Undefined, directed by Yibin Wang, continues through Sept. 7 at Theatre 154 in the West Village of New York City. In the show, the audience follows five AI researchers who are employed by the Human Research Institute. They make an amazing discovery: a mysterious Memory Cube that holds a living record of human consciousness from 2051, according to press notes. Liu’s show promises to be a “genre-bending sci-fi absurdist play about trust, identity and the cost of being human in a world that no longer knows what that means.”
Recently Hollywood Soapbox exchanged emails with Liu to learn more about this forward-thinking play. He has a foot in two professional worlds: theater and investments. His shows are particularly interested in focusing on working-class and immigrant communities, according to his official biography. Questions and answers have been slightly edited for style.
When did the idea for Destination Undefined first come to you?
When ChatGPT came out at the end of 2022, I was shocked. It is like a toy that comes to life. I was so compelled to write a show about this upcoming technology, but as I started writing, the world spins, AI advances, economics gets reshaped, geopolitical tension rises. And I start to see all these thematic topics interlink and mirror each other. These are big topics and too broad to be possibly discussed in a single play. On the other hand, these “big topics” are seriously and deeply impacting each one of us.
After many rewrites, the story eventually settled into a very specific time and place: an underground gold vault, during the announcement of the results of the AI voting rights bill. The characters’ backgrounds and motivations are so layered that their relationships naturally give rise to subtle, complex conversations around all these themes. Many readers and collaborators have contributed their insight along the way, and the play has grown into something far more intricate than I originally imagined — in a good way. It feels alive and unpredictable, just like the world we live in.
When you consider the future of AI, are you excited or scared?
Intellectually excited, but deep down, seriously worried. The emergence of multimodal AI agents is foreseeably the most profound technological breakthrough of this decade. Their reasoning ability across text, images, audio and video allows them to perform tasks we previously could only imagine a human doing.
Intellectually, it motivates many interesting research and engineering problems within AI — across pre-training, alignment and inference — as well as challenges beyond AI, such as how to leverage this technology in daily business across different industries. It enables a significant number of opportunities to disrupt the status quo with a very lean team, and we’re already seeing that happen.
But I’m worried that the long-term social impact of AI will be more profound than we imagine. What hits us first is exacerbated inequality, and we’ve already started to witness that within the tech industry. While AI-related developments receive fat investment checks, disproportionate mass layoffs in other parts of the industry have become the norm. The number of entry-level jobs has shrunk dramatically — as they are the most easily replaced by AI — and the younger generation is hit the hardest.
But what happens to the industries in the long run without young blood? As this creeps into other industries adopting AI, and production becomes more capital-intensive than labor-intensive, income inequality widens. It’s the beginning of a dystopia on the land of plenty.
The sudden emergent ability of AI felt like a singularity moment because we had enough original content on the internet — handcrafted, well thought-through and diligently vetted datasets crystallizing the entirety of human knowledge. But as AI-generated content floods social media and content-sharing platforms, it disincentivizes manual content creation. After all, why care so much about correctness when you can get the work done in one click? And this is dangerous because AI is literally biting the hand that feeds it. The question is: How can we still reward originality, as efficiency takes over the world?
Are you hoping that audience members connect the dots and see how this futuristic play comments on today’s society?
I think the dots are so close that it will be obvious to the audience. Specifically, there are two themes beyond AI that I hope the audience will notice.
The first is about immigrants. Most members of this production team are immigrants, and we all know how hard it is to stay in America. After graduation, you need OPT [Optional Practical Training], then a work visa, then a green card — maybe with a backlog and a decade-long wait for your priority date. It’s like all the stars have to align. On top of that, consistently — but especially earlier this year — the amount of hostility against immigrants has risen. It’s an ongoing uphill battle. And if you think about it, there’s a kind of loose duality between the immigrant’s conundrum in relation to American society, and AI’s in relation to humans. There’s fear, misunderstanding, competition for resources — and in both cases, the stakes are high.
The second is the dynamic between crypto, gold and the dollar. In the past decade, we’ve seen the rise of crypto — more regulated than it was meant to be, but still a major market. Then this year, gold has been all over the news, and the dollar has fluctuated quite a bit. The reason people choose one form of saving over another comes down to trust. For the dollar, it’s trust in the government. For crypto, it’s trust in the proof-of-stake mechanism. For gold, it’s the trust in physics — the purity and scarcity of the metal. And without trust, the systems we rely on every day are at risk. We’ve seen many moments this year where that trust is on the verge of collapse.
While this play is certainly not a Bloomberg report, it’s too real to be taken as fiction.
How would you describe the Human Research Institute in the play?
The Human Research Institute employs robots to study human logic and consciousness. They keep trying — only to realize it’s an impossibly hard problem. The institute is set up as a kind of mirror: The subjects (AI) we study are also studying us. More broadly, it reflects how difficult it is for two groups of living beings to truly understand each other — and how fear and greed ultimately get in the way.
In the play, the researchers attempt to fulfill their mission through deception and by learning from archived memories. But it’s only through moments of vulnerability — when they let their guards down — that real understanding begins to emerge, and that kind of openness has become increasingly scarce in today’s society.
Have you ever used AI to write or make a task “easier”?
Absolutely — I use it every day. It has significantly increased my productivity: writing messy, mundane code snippets; polishing the language in emails and articles; planning trips; even creating marketing videos. Sometimes it feels like a “cheap way out” when I don’t want to put in the extra effort.
And every time I use it, I can’t help but wonder how unsettling it is to anchor one’s self-identity to what one can do — especially when that playground is now being rapidly leveled by AI. I keep asking myself: What makes us human? As capitalism pressures everyone to boost productivity through AI, how do we preserve the personal color in the things we create?
Do you feel that we should be cautious about living in an “algorithm society”?
Unfortunately, I think we’re already living in one. Resource distribution channels are increasingly centralized and controlled by recommendation systems. The ads and content pushed to users are all trained on interactions, and these numbers directly influence what we eat, what we buy and what we do.
If Google search still required some effort — reading multiple articles and piecing together a solution — large language models now spoon-feed users with the “best” suggestion they have. It becomes dangerous when we start treating that as the only source of truth. In this sense, while we are already living in an algorithm society, LLMs [Large Language Models] have dramatically accelerated this slippery slope.
By John Soltes / Publisher / John@HollywoodSoapbox.com
Destination Undefined by Changshuo Liu continues through Sept. 7 at Theatre 154 in the West Village of New York City. Click here for more information and tickets.
