Since the ChatGPT API was made available, numerous apps have incorporated AI functionality. This trend is particularly noticeable in email clients like Spark and Canary, which proudly promote their built-in AI features. These features often include writing replies or composing entire emails based on a prompt, summarizing lengthy emails, and even summarizing entire email threads. While these integrations may seem like a great idea in theory, I believe they actually make communication less efficient rather than more efficient. You are welcome to try out such features for fun, but don’t expect them to revolutionize your life. Let me explain why.
The overwhelming amount of emails and communication, in general, is not solely a technology problem but also a social problem. People may argue that they receive too many emails, but what they mean is that they feel mentally incapable of responding to all the messages they receive. Attempting to solve a social problem with technology often leads to the creation of new social problems.
For instance, let’s say I want to invite you over for beers and instead of writing the email myself, I use ChatGPT to compose it for me. The resulting email is 220 words long, including a polite introduction, reasons to have beers together, and even fabricated details about the atmosphere and snacks. Most people would feel overwhelmed upon receiving such a lengthy email and may resort to using AI tools to summarize it. When I asked ChatGPT to summarize the email, it basically gave me back my original prompt in a single sentence: “Would you like to come over for beers?”
In this context, AI serves a similar function as alcohol, as humorously stated by the philosopher Homer Simpson, being both the cause of and solution to life’s problems. AI creates the problem of lengthy emails and then offers a solution by summarizing them. It’s like a snake eating its own tail, a technology created to solve the problems it generates.
In my opinion, it’s better to examine the cultural assumptions underlying these issues rather than relying on unnecessarily complex technological solutions. We should question the cultural forces that make us believe we can’t simply write a one-sentence email. If ignoring those forces leads to better communication, then why not do so?
Of course, cultural problems are more challenging to grasp than technological ones. While you can start sending one-sentence emails right away, some people may perceive it as impolite or strange. However, any individual or organization striving for efficiency should consider these factors. Unless, of course, you want an AI bot pretending to know the contents of your fridge, including a range of beers from local brews to classic favorites.
Furthermore, we don’t yet fully understand the contexts in which AI works best. My friend Kay-Kay and I had a running joke where we would use LinkedIn’s suggested replies for our conversations. This social network provides recommended response options for messages. It was always entertaining.
Courtesy of Justin Pot.