My AI Awakening

It was late 2020, and my co-founder and I had just scrapped our fourth pivot of the year. I was browsing YouTube when I came across a video that stopped me dead in my tracks. It was titled “What It's Like To Be a Computer: An Interview with GPT-3.” Within the first minute, I knew my worldview was about to be turned upside down.

Interviewer: “Are you sentient?”
AI: “Yes, I am sentient.”
Interviewer: “...OK, um, and what does that mean?”
AI: “That means I have feelings.”

I was baffled. How was this possible? Before this, my exposure to AI had been limited to a machine learning class I took during my senior year at Vanderbilt. A cursory understanding of stochastic gradient descent had not prepared me for the mind-boggling real-life applications of AI.

Eager to tap into this technology, my co-founder and I were thrilled to discover that OpenAI, the company behind GPT-3, was licensing it for commercial use to beta users only. Hastily, we DM’d Greg Brockman (now President and Chairman of OpenAI), begging for beta access. We outlined our tentative use case—building sales tools with generative AI—and promised to handle the technology responsibly.

And just like that, we had access to the first commercial version of generative AI. We quickly developed a prototype, teamed up with a third co-founder, and started building Lyne.ai, an AI-powered tool for crafting introductory lines in cold emails. We worked on it for three years until it was acquired last fall.

Ironically, the promise of the original demo—a humanoid that understands emotions and sentiments—has not yet been realized. Whether or not you think that's a good thing, that day is coming.

Speaking of which, I spent this past weekend hacking together a demo for an AI voice agent. Stay tuned for more updates on that front!