top of page

Down the AI Rabbit Hole

Several months ago, I went down the AI/GPT/LLM (Large Language Models) rabbit hole. My interests are three-fold:

  1. to develop personal fluency in the technology,

  2. to understand the threats/opportunities the technology represents, and

  3. to begin formulating a set of tactics/strategy for our company.

So far, here are my high-level conclusions:

  1. AI/GPT/LLM is not the end of humanity. This is some of the most-disruptive technology man has invented, and like much disruptive tech before (think Guttenberg press, digitized telecom, digital audio tape, the Internet, VoIP, the cloud, streaming media, etc.) there are and will continue to be calls for regulations because of the disruptions. Oftentimes, those calling for regulation because of the “threat” are those most disrupted. The fear of calamity drives towards regulations intended to protect the hegemony of those calling for the regulations.

  2. The new “priesthood”. As we saw with the introduction of the WWW/HTML, a new, highly-compensated profession will emerge, that of Prompt Engineer. The run will last 5-7 years and will greatly reward those who have: 1) command of the written/spoken language, 2) and understanding of LLM architectures.

  3. Websites are dead. (Or at least endangered species). Think about it: why do we have websites? Websites exist to structure content and/or data in specific ways to be discovered and consumed by very specific software applications, namely browsers.

  4. The new discipline: optimization. The website-browser relationship gave rise to the search engine; search engines gave rise to search engine optimization. LLM/GPT applications will give rise to LLM – rather large data model (LDM) optimization: how do you make data discoverable by LLM/LDM applications; how do you make data un-discoverable by LLM/LDM applications? There is a new construct. Remember how hard it was for companies to optimize their websites for mobile? This is a much bigger and deeper optimization.

The implications are clear: AI threatens the ~$700B annual, digital advertising market and the valuations of several trillion-dollar companies. That’s why the “regulatory” alarm.

The real questions we should all be asking are:

  1. How is my business threatened by AI, and

  2. What new opportunities for my business are propelled by AI?

bottom of page