What do you mean by "the way it's able to access data"? AFAIK ChatGPT is just a fine tuned version of GPT-3.5 optimized for chat[1], it isn't any different from GPT-2 in the way it accesses data.
If the model is smaller, I still do that, sure, but I can only ask "basic" questions about the English language and so on. But if it has more neurons, it's capable to "index" a lot of questions about e.g. proper Python usage.