ChatGPT: transforming us from creators to revisers just as the microwave transformed us from cooks to reheaters.
That email that I just had the robot write is probably good enough to send without editing ? At least that is the thought that went through my brain after I last used ChatGPT (the robot) to write an email. I’m not sure if that’s because the robot is getting better or if it’s because something is changing in my brain. Actually, I’m pretty sure that both of these things are true. The robot is getting better, and something is changing in the way I think about writing.
In fact I noticed that I recently went about writing in way that is similar to how I might “cook” a frozen dinner. When I cook a frozen dinner, I peel off the plastic, pop it in the microwave and find a task I can do for about three minutes. This usually involves doing something on my phone or emptying part of the dishwasher. Either way I disengage while I let the machine do its work. Of course I don’t love the frozen dinners, and I could have had something better to eat if I wanted to put more effort into the task, but the meal was good enough.
Recently when I was attempting to write something, I used a process that was strikingly similar. I logged into the robot, spent a minute crafting a prompt and then found something to do for the next 60 seconds. Similar to reheating, I mentally disengaged from the task, consumed the results and decided that the product was good enough. I think that this might be a problem over the long term
It has been documented that Google does make us “dumber”. The Google effect was first explored in 2011 and has been refined and clarified through additional studies, and the conclusions are not great for the human brain. In 2021, Fisher, Smiley, and Grillo argued that “When people see how to reliably access new information using Google, they become less likely to store that information in their own memory”. This is an important conclusion because it argues that having access to Google will likely change the way we use our brains. This is particularly troubling when it comes to creating and the robot.
It is very easy to overstate the quality of the work created by the robot. Most of the things that I have it create look pretty good. They look pretty good because I haven’t gone through the intellectual struggle that is the writing process. I usually don’t know what I think until I write something, and I think that “writing to learn” is a pretty common strategy.
One of Sam Altman’s recommend books is Thinking, Fast and Slow by Daniel Kahneman. Kahneman writes that when faced with a difficult question, we often answer an easier one without noticing the substitution. I think that using the robot will make this type of substitution happen more often. We are not going to notice that we don’t really understand problems on a deep level when all we are doing is revising the robot’s work. We will very much feel that the robot’s answers are sound and in some ways the answers may even seem surprisingly creative. This is because we haven’t thought deeply about the content.
Don’t get me wrong, microwaves can be great, but it would be difficult to argue that they have made us healthier overall. Unless we think deeply about what the robot is doing to the way we create we might be headed down a similar path where quick, easy and good enough becomes the norm.
