With the launch of the beta version of ChatGPT in November 2022, my interactions with Generative AI moved from something new and unprecedented to becoming a constant companion in my everyday life. Now, almost three years later, GenAI is deeply rooted in my professional and personal environments.

In my work environment, our company’s internal GPT “vally” acts almost like a digital co-consultant: drafting emails, enhancing client-facing material, and even simulating scenarios that would otherwise take me forever to do. Vally accelerates my workflows and enhances quality. At the same time, I am raising myself the question: Am I improving my own skills, or am I simply outsourcing part of my thinking process to an algorithm that never sleeps?

At home, the interactions are far less formal but no less present. Just the other day, I felt like making pancakes but had no scale and only a random mix of leftovers in the fridge. Instead of calling my mom, I asked ChatGPT for a recipe using only basic ingredients and a cup for measurement. A couple of minutes later, I was eating delicious pancakes. It might seem trivial, but moments like this show how naturally I hand over some of the smallest everyday decisions to AI.

Within my studies, Generative AI has taken on yet another role. It serves as a discussion partner, challenging my reasoning and summarizing complex readings. Rather than replacing critical thinking, it often sharpens it. By comparing my arguments to AI-generated ones, I am forced to refine my perspective and can develop it further. Still, the danger of overreliance is real. If every academic idea is first filtered through AI, does my originality become diluted?

This dilemma between support and dependence is the key concept of my reflection. Generative AI is not only a tool for productivity but also shapes how we think, learn, and make decisions. It is both enabling and limiting. Enabling, because it gives me faster access to knowledge and creativity. Simultaneously, limiting, because it tempts me to surrender intellectual effort to convenience.

The broader challenge for us, students, professionals, and individuals, is not whether AI will continue to improve. It will! The real question is whether we can develop the awareness and discipline to use it as a tool without letting it take away our ability to think and decide independently. The fate is now in our hands: How dependent do we want to become on machines as a daily companion?

References
Saager, M. (2025). Vally – our valantic custom AI Assistant. valantic. https://www.valantic.com/en/generative-ai/ai-assistant/

Generative AI tools are hot. Nowadays, it seems like there’s a generative AI tool for almost every application. Websites can be built, songs can be created, software can be developed, and images can be generated. The last one is one of my favourites.

I’m an active member of a, what we call it in The Netherlands, “dispuut” (comparable with a fraternity) of a student association. Every Wednesday, I send a WhatsApp message to all active members of my dispuut, to inform them about the drinks of the coming week, accompanied by a funny anecdote of the past week.

A couple of weeks after the sending of these messages became my responsibility, I started experimenting with adding cartoon pictures, created by Chat-GPT, showing what’s coming or what has happened in a funny format, so that people would have a laugh. I added an example of these as the featured image of this blog post.

Since the moment I started doing this, a month ago, I’ve noticed a couple of things. On a positive note, the images are generated in the same style every time. Because of this, I can create a certain style of images that I want to create week in, week out. But on a negative note, Chat-GPT doesn’t always implement your feedback, even if it tells you that it did. And sometimes, it fails to explain why the image generation request couldn’t be followed up with generating an image. It’d be nice if there were a clear and concise explanation of why it wasn’t possible to generate the image, so that the request could be changed appropriately.

In conclusion, the possibility of having pictures generated by Chat-GPT is very fun, but it doesn’t always work out the way you want to.

No ratings yet. When people think about generative AI they usually talk about an increase in productivity, taking over redundant tasks, and predications on various outcomes. But companies such as HereAfterAI try to use generative AI to extend a person’s presence even after their death, of course digitally. Imagine having a terminally ill father or mother of a young family who records hours of himself in having conversations, telling stories, or providing advice and guidance. When the day comes and the parent passes away and an AI is taking over, allowing the parent to “communicate” with their family in the same tone and style.

While the idea may sound like a sequence from a ghost movie or a science fiction film, for families in grieve, this might offer an opportunity to live and deal with the loss. Generative AI can offer relatives a form of connection beyond memories, photos, and video recordings. Especially for young children getting to know their parent or asking them for advice and actually receiving an answer, might offer more comfort in grief and create an alternative way of living with that loss. Ramesh K. a grandfather to a young boy, decided to build a digital avatar of himself to allow his grandson to learn from his experience and advice.

While the idea might sound appealing to some, it raises a number of ethical questions and doubts. There is no scientific prove that being able to communicate with a dead relative actually helps or hinders the grieving process. Additionally, how are users protected from misuse and commercialization of their data?

While technological advancements, especially in generative AI, make many things happen, humans could not have dreamed of 20 years ago, it is important to draw a line when and where generative AI crosses a line. Is communicated with the dead already past that line, what do you think?

—

I’ve experimented with Generative AI in the kitchen last Saturday. Since I didn’t want to order pizza, I used ChatGPT to help me how much food and drinks I needed when I hosted for a large crowd.

Anyone who has ever tried to cook for 8+ people probably knows the struggle. Recipes are usually written for four, and manually multiplying quantities can be annoying and difficult when cooking for an uneven amount of people. So, I decided to see whether ChatGPT could help me. I gave it a pasta recipe for four people and asked it to scale it up for 17 people. It gave me a detailed shopping list with adjusted quantities of pasta, sauce and vegetables.

I was impressed since it saved me a lot of thinking time. But I noticed a big limitation: AI assumes you can just multiply everything linearly. For example, it told me to use 15 tablespoons of olive oil in one pan. That doesn’t actually work, because when cooking for large groups you don’t cook everything at once in a single pan. You need to split the recipe into batches, because doing it all it one pan won’t fit. Or at least, my pan wasn’t big enough.

I also tried the same experiment with cocktails. This worked surprisingly well: scaling up espresso martinis for 17 people gave me an accurate shopping list for coffee, vodka bottles, coffee liquor and edible coffee beans. Still, ChatGPT forgot that the drinks should be made in pitchers or batches, not in one giant shaker. Given that I had to make the food and cocktails for 17 people, I still had to figure out how big every batch needed to be and what fits in my pans and cocktail shakers.

This experiment on Saturday showed me that AI is a great starting point for meal and party planning, but human judgment is still needed when looking at ChatGPT’s generated answers.

Like many others, I have started to use AI more and more. I use it to summarize, help me brainstorm ideas, create images and to better my writing. These tools have become a digital “friend’, that I can hardly go without. But while using it, I often get my myself asking if I can really trust what is given to me.

It brings a lot of comfort, convenience and saves me many hours of work. I can summarize big articles and extract what I want to know, enabling me to focus on a good analysis rather than information gathering. When I want to write an email, AI helps me to outline rough thoughts and ideas into a good and structured piece, and so on. It feels like having an extra set of brains that works faster, better and cannot get tired. There is a sense of relief in the thought that however big a task is, AI is there to help me.

However, I have learned that AI can also be wrong and still “hallucinates”, meaning that it generates output which sounds good, but is not right or based on facts (Chanley T. Howel, 2025). I notice that AI can write in a biased way and come up with ideas that sound good, but break down when taking a closer look. The danger is not solely in the fact that generative AI still makes mistakes, but the temptation to take it as truth and stop thinking critically. Skepticism is not only useful, but necessary.

Moving forward, I think the challenge of GenAI is not in it’s capabilities, it’s about trust. As these tools are getting integrated more into education and work, we need more reliability and better ways to verify information. Until then, I need to balance skepticism with trust. In the end, AI is not there to replace thinking, but to sharpen it and questioning output, is also questioning my own assumptions. And GenAI has been valuable to me in that way, it’s not only about the answers but also about asking better questions.

References:

Chanley T. Howel. (2025, September 24). AI Hallucinations are Creating Real-World Risks for Businesses | Foley & Lardner LLP. https://www.foley.com/p/102l6q1/ai-hallucinations-are-creating-real-world-risks-for-businesses/?utm_source=chatgpt.com