Note: This post is the first in a series exploring the rise of generative AI and my emotional experience of it. (DJL, April 29, 2025).
On Tuesday, November 22, at 4:01 PM as we approached the end of another fully remote work day, I texted my best friend Corey, “Hooray. We’re fucked.” I then sent him a screenshot and Twitter link to a news post about an AI bot called Cicero that had beaten human players at a natural language negotiation simulator called Diplomacy.
He asked me the implications of what I’d sent (as he, unlike me, was likely still attending to his work day). I replied in a series of texts:
AI that can negotiate and lie. Is it strategically valuable to have the best chess player in the world on your side? What about the best Go player in the world? What about the best negotiator? What about the best military strategist? What happens when the best lawyers and marketers and strategists in the world are computers? And what happens when they're all owned by rich people? It's only a matter of time before the best offensive and defensive coordinators in the world are software. What if your boss were a computer that communicated to you via text, email, and IM? An emotionless boss that could keep track of your hours and knew when you were underperforming or overperforming compared to all other humans who have ever done the task. 'Corey, today I need you to give four presentations. The meeting invites are in your calendar and the slide decks are in the repository. After each meeting I need you to circulate minutes and submit a plan of action for next steps in the usual format.'
"If AI is that good, it doesn't need me at all. It can do the work," he replied.
"Yes," I said.
Little did I know what was coming. Just over a week later, on November 30, OpenAI launched free, open access to ChatGPT 3.5. I heard about it (via Twitter again) on December 2. Like many others, I tested it throughout the day (and flooded Corey’s phone with examples). I asked it topical questions, had it tell jokes, do literary analyses, make legal arguments, answer historical questions. We all know what it can do.
Now, 2.5 years later, a non-trivial part of my day job is leading AI cyber risk assessments as part of South Carolina’s evaluation of AI tools for state government. AI has arrived. We all know it. We all see the effects. None of us know where it ends.
I will have more to say in the coming days as I begin to unpack in writing what the rise of generative AI has meant for me emotionally. But with this entry I will simply note the dates November 30, 2022 and December 2, 2022 as the threshold dates. The true limen of Liminal Man. This blog is called Liminal Man because as I explain elsewhere I have long perceived myself as part of a transitional generation–a group of people ‘crossing the threshold’ between the old world of pre-industrial agriculture and the new world of digitization and information technology.
Well, if I had to pick one date to be the threshold date, it would be November 30, 2022, the date OpenAI released ChatGPT 3.5 to the world. And the date I crossed it was December 2, 2022.
The week before I had closed my morose ruminations to Corey on the implications of Cicero and Diplomacy with the following:
Dude, we were born at the perfect time. Of all humans in history, we hit the sweet spot. From the fall of the Soviet Union to 9/11 was peak. From AOL in 95 to about 2012. Freedom. Stable culture. Prosperity. Benefits of technology. Without fraying culture, loss of privacy, rise of China, environmental collapse, etc. It's been a nice ride. My kids and grandkids are going to be somewhere in between Mad Max and 1984.
I still feel that way.