The internet and the wider digital revolution have delivered us a more complex, emergent world as people have interacted through social media and internet networks. Now we have a materially different experience with ChatGPT and what will emerge with generative AI is impossible to predict.

Over the past few weeks, I’ve received a non-trivial number of communications starting with a preamble along the lines of: “I would like to say that ChatGPT was not involved in the making of this. If you find what I’ve written soulless and derivative, I’m afraid that’s down to me.”

There’s a lot of binary thinking going on—either/or. Either ChatGPT offers no original thought therefore not useful or it’s highly productive and therefore very useful. (Plot bust: both are true). Aside from ChatGPT disclaimers being the 2020s version of “sent from my iPhone” why should you care about ChatGPT?

We’re going to give you three reasons ChatGPT (and generative AI more broadly) matters a lot.

‍

1. ChatGPT shows us that questions matter just as much as answers.

Much of the quality of answer generation depends on how you ask the question. Asking good questions is a highly creative and intuitive skill and ChatGPT’s best use is when you consciously work on your questioning skills. As well as being clear you’re using ChatGPT, be clear that building your questioning skills (and those of others) is what will make ChatGPT more valuable to you.

‍

2. ChatGPT changes how we perceive AI mistakes.

ChatGPT’s role is to help us better access knowledge from our community. Interacting with it is more intuitive and natural than most AI interactions. Human-machine interaction remains awkward, unnatural, anthropomorphic, and error-prone in many cases. This keeps us alert to AI errors. But ChatGPT makes human-machine interaction significantly more natural, which is a double-edged sword. Some errors are easier to spot—the ones when context is off or there’s an obvious absurdity. But other errors are much, much harder to detect. For example, when ChatGPT includes a value judgment that is overly narrow or skewed by the internet corpora. Automation bias further accentuates this. If there’s one golden rule for using ChatGPT it is this: assume it’s going to be wrong somewhere and that you are the one who is on the hook for the error.

‍

3. ChatGPT’s knowledge is emergent.

ChatGPT doesn’t give the same answer every time, even to the same question. Questions generate answers which generate questions which generate answers…Beyond that, as your questions become more advanced, ChatGPT reveals more advanced answers. What happens is emergence—humans and machines respond to the level of each other’s knowledge. ChatGPT is ultimately a cultural technology that produces cultural artifacts. We have to evaluate it in this light. It will make access to humanity’s knowledge easier and faster for new generations and this will change the way we pass on knowledge. This is a profound shift.

It is this final point that we think is the most important. The internet and the wider digital revolution have delivered us a more complex, emergent world as people have interacted through social media and internet networks. Now we have a materially different experience with ChatGPT and what will emerge with generative AI is impossible to predict. The only way to know is to immerse oneself in the system.

‍

‍Don’t miss out on more insights about ChatGPT and our advice about the intersection of human and artificial intelligence! Sign up for our newsletter and be the first to know when new blog posts are published. Sign up here.