Large Language Models and Time
Putting it to the test…
One day, people will be looking back at today and researching the beginning of the AI Age. They will ask themselves questions like, “Who chronicled this revolution and renaissance into an age of broad-based enlightenment?”
In short, people will be asking if people were aware of the sheer size and scale of what was just about to happen to the world on the release of Large Language Models (LLMs).
Central to that investigation will be perceptions of our perceptions as to this moment in time. How did we test LLMs? How and when did we realise their full potential? As AI approached human level intelligence, what enquiry did we make into the many facets of human intelligence and human capacity to reason.
Part of that enquiry will be our enquiry into whether we enquired of the best Large Language Models their, if any, conception of time.
So, tonight, I did just that. A simple test with ChatGPT 4o testing its ability to place events in a point in time and have a concept of the time between those events.
Medium is about writing, so in the spirit of Medium and to capture this point in time for future researchers, my conversation with ChatGPT 4o, below:
