Artificial intelligence is a hot topic. There are several AI-powered tools on the market, whether in a commercially-viable or prototype form. Perhaps most significant is ChatGPT by OpenAI, which was made public at the tail end of last year. What’s ChatGPT? Why don’t we ask it:
To describe itself as a “language model” may be correct but it’s a little modest. ChatGPT’s abilities are vast. It can answer questions, generate code, write articles, translate documents, tell you a joke, or engage with you in some good old-fashioned chit chat.
This of course is an impressive piece of technology and a fun tool to experiment with. But its scope and usefulness extend far beyond curiosity. It has practical, real world applications. Why spend time programming a website component when an AI can do it for you in a fraction of the time? Why struggle with writer’s block when a copywriting deadline is looming? Why be presented with adverts when searching for information online? Prompt ChatGPT and it’ll take on the task and output some code, content or answers within minutes if not seconds.
Amazing, right?
Let’s take a step back
The validity of the phrase artificial intelligence is often contested. As a species, we are yet to develop something that is actually aware, or truly understands what it is doing. Perhaps a more accurate description is fancy algorithms. Patterns, machine learning… and sometimes very wonky output.
On that topic – ChatGPT makes mistakes. That’s not a criticism. Just like a human, ChatGPT learns from mistakes… but it doesn’t feel embarrassed about it. When things go wrong, you can tell it so. It’ll then take steps to rectify the error. There’s some give and take when working with ChatGPT.
What artificial intelligence could mean for email
The combination of email and AI (or fancy algorithms) is nothing new. One-to-one product recommendations have been around for years, picking relevant items based on previous shopping behaviour. Spam filters automatically guard inboxes against emails of the shadiest kind. Customer journeys and automated emails are made possible through complex workflows with little post-development need for human intervention.
What is new however is the concept of emails that are predominately or even completely designed and coded by computers. Imagine high-quality, on-brand design and copy that is generated in seconds. Perhaps the same AI tool could then select the audience, send the email, read the report and optimise the next send. Is there a point at which human input becomes zero?
What artificial intelligence currently means for email
Don’t worry – we’re not at the human irrelevance stage yet. In my tests with ChatGPT, I’ve seen it output some erroneous facts and broken code. From what I’ve seen it’s neither ready to fly solo nor likely to achieve such independence any time soon.
That doesn’t mean it’s not a revolutionary and practical technology in its current state. I already use it almost daily for writing inspiration and to help with coding questions that would otherwise mean trawling through forums for answers.
ChatGPT and other pioneering AI technologies have already changed the way we work, and they will only continue to evolve. It’s incredible to think that what was only recently in the realm of science fiction is now becoming a reality. The future really is here.