CHATGPT is not your friend

Posted on LinkedIn on the 28th April, 2025

Remember when we drove cars without navigation systems? In those olden days of yore, we knew approximately how to get somewhere, maybe looked at a map or a signpost or two, and generally figured out how to reach our destination. If not (and, as the bad joke used to go, if you were a woman) we asked the actual real-life stranger who was conveniently standing on the street corner for help. Once navigation systems arrived, it suddenly felt like we had collectively thrown our orientation skills out of the car window. We couldn’t get anywhere without them.

At the moment I can’t seem to get through a single day without at least one person saying “So I asked my friend ChatGPT, and he (have you noticed, it’s always he?) said…”

I hate to break this to you.  But ChatGPT is not your friend.

I have no problem anthropomorphising objects. Our first navigation system was called “Gertrude”. (I do, as a side note, have a problem with objectifying and dehumanising humans, which also seems to happen every day.)

I mean a friend would not act like this.

1. A friend does not destroy your home

Chatbots are not just a funny piece of software floating in the clouds. They need massive data-centres, which:

• Need staggering amounts of raw materials and rare earth elements to build
• Produce electronic waste, including hazardous substances
• Need water for cooling the electrical components. According to UNEP: “Globally, AI-related infrastructure may soon consume six times more water than Denmark, a country of 6 million, according to one estimate. That is a problem when a quarter of humanity already lacks access to clean water and sanitation.”
• Need masses of energy which in most places still comes from the burning of fossil fuels, producing planet-warming greenhouse gases. A (text-based) search through ChatGPT uses 10 times the electricity of a classic engine search.

Not only does AI consume huge amounts of energy and water, the big corporations Amazon, Microsoft and Google are expanding their data centres in water scarce areas (for example in the Aragon region in Spain).

That’s a lot of resources to show people a plastic doll version of yourself on social media.

2. A friend does not isolate you from your other friends

A few people in an online room. Finding connection remotely is hard enough, but we have found our ways over the last years. Suddenly there is one more guest that didn’t get the invite. A new unsettling presence that creates distance.

People are even getting AI therapy. Mental health is still so taboo, that to many this is a golden opportunity to keep their shame and their pain hidden, as they never have to interact with a real person. But artificial intelligence cannot give us authentic empathy.

Research shows that “Heavy users of ChatGPT tend to be lonelier, more emotionally dependent on the AI tool and have fewer offline social relationships.”

The source of our problems is disconnection, and the solution these Chatbots are offering is less connection. Talk about throwing oil on the fire.

3. A friend doesn’t damage your skills (or: “ChatGPT, you’re shaking my confidence daily”)

ChatGPT can swoop in and do what we humans had been struggling with for hours. But I worry more about us repeatedly pushing the button that turns our confidence to think critically and solve problems to mush.

It’s the navigation system saga all over again, which this article describes as «quintessentially infantilising, leaving us unable to get around without it, or cope with its shortcomings.» In a similar way, reliance on Chatbots, according to this research, “could limit users’ exploration of alternative problem-solving strategies”, limiting our ability for independent thinking as well as our adaptability to complex problems and promote “passive acceptance of information.”

We can make similar assumptions about creativity, inspiration and learning. These things take practice. By not sitting with your problem for more than the time it takes you to write your “friend” a prompt, you are telling present and future strokes of genius that they are not welcome to the party.

As Julia Cameron says: “It is impossible to get better and look good at the same time” and “we must trust our process, look beyond “results.” “

4. A friend doesn’t enable your exploitation

An old favourite myth when technology increases productivity is that we will soon all sit around sipping mojitos and reading books all day. As we become more productive, we will automatically be better compensated and have more free time. Right?

How has that worked out for us so far? This handy little calculation, the productivity pay gap, shows us that in the US (other countries follow suit) for example, since the 1970:

• Productivity grew by 86,5%
• Compensation only grew by 31,7% (i.e. productivity grows 3 times faster than compensation)
• Average working hours reduced by only 7%.

Hmm I wonder where the benefits of productivity went? On a completely unrelated topic, did you know the income of the 1% richest people in the US grew by 50% in the same period? (World in Data)

Make no mistake that if profits of higher productivity have until now gone to bosses and shareholders, they won’t change their MO now when it is all working out so well for them.

5. A friend doesn’t pretend to be your saviour (to a crisis they help create)

What enrages me the most is the belief that “the solution to the climate crisis (or insert crisis you care about) will come from AI.”

This is how the story goes: If we have not yet solved the meta-crisis, it is because nobody has thought of a solution yet. Some cool superman entrepreneur is just about to fly in with his best buddy.AI and save the world at the very last minute (making millions in the process). Phew.

If we are not changing, it is not because we have not yet thought of a better way. It is because those who currently benefit from our current extractive, imperialist, capitalist system, don’t fancy doing what is desperately needed, what is morally right. Instead they are doing everything in their power to stop change from happening.

So what, stop using ChatGPT completely? Un-friend the bot?

After listening to people who know more about these things than me, I come to the following conclusions:

• Be aware of its impact: many of us (at least in the « bubble ») think twice about flying, we should also think twice about prompting.
• Avoid it when you can. If you don’t absolutely need it, use a classic search engine instead (here prefer Ecosia to Google)
• If you do really need it

…only use it for text – not images or video (which need even more energy)

…and use more ethical platforms like earthlyinsight.ai

• Opt-out of the systems that work automatically work in the background (like Google in Austria – read this post by Katy Shields for more)
• Talk about how you navigate these decisions with others.

The bottom line (irrespective of the question!) is prioritise connection and empathy with real people. Whether they are in the room or not, whether they look like you or not. Ask for directions. Greet your neighbour. Support other humans.

Most importantly: Remember your brain has worked just fine until now and that your creativity needs practice. You have something unique to contribute.

We will find our way together.

What are your thoughts on this? How do you make decisions about using Chatbots or not? How can we raise people’s awareness about this?