Hey there!

Fall weather has arrived in Berlin, and our latest newsletter issue has arrived in your inbox. We hope you like it.

At Ponder, we do quite a bit of writing – for example, for concepts and content in client projects. Or when working on our newsletter.

So we understand how much work goes into good communication. That is why we tend to be sceptical when new tools promise to make this work easier – for instance, through artificial intelligence.

But we are still curious ... and today, we would like to share what we have learned in our recent dealings with this matter.

Let’s go!

But first: SOLIDARISPREE #2 🙇

Yes, this is already the third time we have mentioned our summer event, but it will be the last for a while, promise!

Today, we simply want to thank everyone who came, supported our fundraising efforts – and made last Friday such an enjoyable evening.

We had a blast!

PS: We’re still well-stocked with wine and other drinks … 😉

Pondering: Using AI for copywriting is intriguing, but … 🤔

Artificial intelligence has spread into almost every industry and business in the past few years. We can now find it in everything from manufacturing to healthcare. AI has even made its way into copywriting, which is why we’ll shed some light on this topic today.

Publicly available AI tools can automate some tasks, assist humans with coming up with ideas, and analyse text quality. This seems tempting, but we should still be careful when using AI for writing copy. Here are just three reasons why:

1 – AI is still in its infancy and prone to error.

The first thing we must discuss regarding artificial intelligence and its use in copywriting is that it’s still in its infancy. This means that AI will often produce errors that humans would never make. This can be due to various factors, including the complexity of the AI program and the amount of data used to create the algorithm.
In addition, AI can be biassed, which can significantly impact results. While humans can question their biases and learn from them, AI cannot. As a result, AI will have the same inherent bias no matter how often it is used.

2 – AI can use data gathered from people unaware of being monitored.

AI can be risky in general (and for copywriting) because it often relies on data gathered from people without their consent. In many cases, AI will acquire data from users who willingly provide information. However, in other instances, companies will collect data from users without their knowledge, making it ethically questionable.
Tinder is one example of a company that has faced controversy for using AI that collects data from users without them knowing. They use AI to analyse the user’s photos to determine their attractiveness. It’s generally problematic because Tinder gathers data without users’ knowledge. But also, users may not be presenting themselves in their photos the same way as they would in real life, resulting in low data accuracy.

3 – The danger of algorithms with inherent biases

Algorithms are modelled after the data they are given. If the data has inherent biases, the algorithm will replicate these issues. This means that if the data was gathered from people without their consent – or from a biassed source – the algorithm will stay biassed. A data set might have been collected over a long time, resulting in an algorithm being biassed for a long time as well.
This can be highly problematic for a variety of reasons. It can result in the loss of both customers and employees due to discrimination. Bias and discrimination can also damage your reputation, making it more challenging to draw in new customers in the future.

Conclusion

Artificial intelligence is a fascinating development and has already made many positive contributions to the world. However, we must proceed with caution when using AI for copywriting purposes. This is because AI is still in its infancy …

[ RECORD SCRATCH ]

Hold on.

Did you notice anything in the previous section?
Because you just witnessed a curious little experiment.

Unlike the rest of this newsletter, it was created not by a human but by an AI writing tool called Writesonic. All we did was enter an oddly specific description of the topic into a form field and work ourselves through the tool’s recommendations – from intro to outline to the actual article.

After using Writesonic to generate the initial draft, we ran it through another well-known AI service, Grammarly, accepting most word suggestions and grammar checks and hitting the „rewrite for clarity“ button a few times. This second step helped bring the text closer to what you’d usually read here.

Oh, and of course, the artwork for this section was also AI-generated (using DALL·E).

So, what should we make of this?

Honestly, we don’t really know just yet. However, putting aside the mostly superficial talking points and the slightly repetitive nature of the AI-generated content, the possibilities of today’s tools are pretty impressive.

This little experiment, driven by nerdiness and curiosity, has definitely sparked our interest – and we will delve deeper into the matter in the future.

But at this point, we are interested in one question above all:

What’s your take on the subject – what kind of thoughts, hopes and concerns are going through your mind?

Weekend send-off

As a regular reader, you know what’s next – our inspirational weekend send-off.

This newsletter was already pretty long and hopefully gave you plenty to ponder over the weekend. But to keep up the content-heaviness of this edition, we have another interesting (yet slightly more actionable) resource for you:

RESOURCE: Language, Please As most of us will be doing our own copywriting in the foreseeable future, we think you will find “Language, Please” quite helpful. They describe themselves as “a living resource for all journalists and storytellers seeking to thoughtfully cover evolving social, cultural and identity related issues.” The platform includes style guidance, an inclusivity reader directory, and editorial tools. At Ponder, we’re all for thoughtful language. So, naturally, we’d recommend you check out the website at your earliest convenience.

Thank you so much for reading!

We hope you’ve enjoyed this newsletter. If not, please let us know what bothered you before you unsubscribe. We’re always happy to hear your feedback and use it to make the newsletter more relevant for everyone. Here’s a simple form to submit your thoughts.

That’s it. Read you again in a month — or drop us a message whenever you feel like it!

Hej då,
Your friends at Ponder