Reflections from the Age of Co-Creation: My Experience with Generative AI
10
October
2025
No ratings yet.
Generative AI has quietly shifted from being a futuristic concept to a daily companion. Over the past year, I’ve used tools like ChatGPT for writing, Midjourney for visuals, and Runway for short video concepts. At first, these tools felt like “assistants.” Now, they often feel more like co-creators.
I first realized their potential while creating marketing content for my previous job. That is where I was taught – and learned firsthand – how to use AI daily to optimize my workflow. What used to take hours -writing ad copy, structuring blog posts and captions, and experimenting with brand messaging – could suddenly be done in minutes. Since the company also covered the premium subscription, I could make full use of ChatGPT’s advanced features. I wasn’t just speeding up my work; I was expanding my creative and critical thinking. It offered multiple directions at once, forcing me to reflect on why I preferred one version over another. Instead of replacing creativity, it amplified it by giving me a creative mirror to think through ideas faster.
Yet every marketer who uses chat agents like ChatGPT has likely noticed the same limitation: a narrowness of perspective. The model reflects what is statistically common, not what is contextually insightful. When generating campaign ideas or headlines, it tends to default to safe, universal tropes rather than niche or counterintuitive angles that truly capture attention. In other words, AI can reproduce creativity, but it struggles to originate it. This limitation becomes especially visible when working in branding, where differentiation and emotional subtlety are key. ChatGPT might suggest a clever slogan, but it rarely surprises – it gives you what the internet already thinks is good. True creative insight still requires human judgment, intuition, and cultural sensitivity – elements that can’t be reduced to patterns of probability.
Then came visual tools. While I haven’t employed AI image generators for my professional work, I used AI to inspire me on certain elements of the visuals and the layout of the final project. As an example, for my previous blog post – I described an idea – a split world between traditional aviation and virtual travel – and within seconds, I had a hyperrealistic visual that perfectly matched the concept. That moment captured what makes generative AI so transformative: it compresses imagination-to-reality time from hours to seconds.
Again, it’s not without flaws. AI often delivers polished but “safe” answers. Creativity, by nature, thrives on unpredictability and imperfection – two things AI still struggles with. I sometimes notice how text outputs can sound formulaic or visuals too idealized, repetitive and almost too perfect, lacking the human quirks that make content memorable. There’s also a growing concern about over-dependence: when the tool becomes too good, do we stop exploring ideas ourselves?
One improvement stands out to me – especially after writing the text for this blog: It would be a “co-creation mode” – an interface where AI explains why it made certain creative choices and lets users steer tone, emotion, or intent interactively, almost like a conversation with a creative partner rather than a tool.
Generative AI has taught me that creativity isn’t dying – it’s evolving. The next leap won’t be about machines creating for us, but about humans learning to create with them.
So I’ll end with a question for you: When your next big idea comes along, will you brainstorm it alone-or with an AI sitting right beside you? ( I suppose it is the latter )
Film Photography and AI: My experience
10
October
2025
No ratings yet.
I started using ChatGPT for school, then realized it could help my photography too. With film, every click costs money and time. Having a quick second brain lowered the stress and helped me make better choices before I even loaded a roll.
Most of my shoots are low light or mixed neon. I ask for a quick plan: likely shutter speeds for Cinestill 800T or my MARIX 135 T800 and Amber T800, what EV to expect at blue hour, and how far I can push before motion blur ruins the look. It is not magic. It just gives me a sensible starting point so I do not waste half a roll testing the obvious.
I also use it for composition practice. I describe a scene from my contact sheet, like “subject under a shop sign, bright window behind, messy foreground.” It suggests two or three framings to try next time. Step left to kill a distraction. Drop the angle to separate the subject from the background. Add a leading line from the curb. Simple ideas, but it keeps me iterating. My contact sheets feel less random and more like a series with intent.
Metering and color are where it saves me the most. If I am debating 1 stop over for skin indoors, or how much to bias exposure for tungsten under mixed LEDs, I ask for trade-offs. It reminds me what will happen to highlights on 800T and what to expect from halation. When a scan comes back with a green cast, I run a quick checklist for likely causes and fixes. It is the same with push or pull. I still note my lab’s advice, but I go in with clearer expectations.
Trust grew with results. The more useful the output, the more I tried. I still keep guardrails. I verify technical claims, write shot lists, and never paste personal data. The goal is not to outsource taste. The goal is to give my taste more chances to show up.
If you shoot film, try this next roll. Write a one paragraph brief, ask for two lighting setups and a backup plan, and make a tiny shot list. Then compare that contact sheet to your usual one. Did you see more, or just shoot faster?
Trying on Clothes with Chat
10
October
2025
No ratings yet.
I feel like every day I’m introduced to a new model of AI, and I need to make a conscious effort to keep up with it, especially to get the most value out of my ChatGPT subscription. Before starting this course, I had never really thought about the differences between General AI, Generative AI, Deep Learning AI, and so on. However, I’ve become much more aware of the types of AI I use in my daily life and how I can apply them to everyday questions and tasks.
To get more hands-on and fully utilize ChatGPT’s generative capabilities, I decided to ask him to help choosing which clothing items I should purchase. For a task to be considered “generative,” the AI needs to create new content (e.g., images, music, code) rather than simply analyze existing information. So, I asked it to generate images of the outfits I had in mind so I could better visualize them.
I’ve been wanting to buy a new coat for a while, but I kept putting it off because it usually takes me a long time to decide on the fit and color I want, and then even longer to find a store that sells something similar. To make this process more efficient, I first shared a link to my Pinterest board so he could know my preferences. Then, I asked it to search for coats I might like, and it returned a list of potential options with images and store links. This part of the process was more of a general AI function, involving search and curation.
^^ One of the suggestions that he gave me
Based on the items I liked most, I then provided ChatGPT with a screenshot of a dress I already owned and asked it to generate images of full outfits that included different coats. Before generating, I also gave my height and clothing size so that the proportions would look more realistic. This allowed me to clearly imagine how each piece might look on me, without actually having to try them on.
^^ Output of the generated image.
I found this use of ChatGPT incredibly helpful because it saves me a lot of time. Normally, it would take me at least 18 minutes to narrow down my options to a few stores, and then even more time to visit the shops and try things on. With generative AI, I was able to visualize my outfit options quickly and efficiently.
One way I could improve this process even further would be to provide ChatGPT with a full-body photo of myself. That way, the generated outfits could be customized to my actual body shape and features. In the future, it would be great if more clothing websites started integrating GenAI tools that allow customers to virtually try on clothes. This could completely change the online shopping experience, making it faster, more personal, and much more convenient.
Who Will Own TikTok? What the U.S.–China Deal Teaches Us About Platforms and Competition
18
September
2025
5/5 (1)
Context
Back in 2020, President Donald Trump announced the plans to ban TikTok in the United States. At the same time, the justification was national security concerns over the potential access of data by the Chinese government through TikTok’s parent company, ByteDance. The ban was never materialized, but it marked the first moment when a social media app became the center of a geographical standoff. Fast forward to today, with Trump now back in power, he is once again pushing for a “solution”. This time, according to CNN News, the proposed deal with Chinese President Xi is that the majority of the shares would be transferred to the American investors, like Oracle, while ByteDance would only retain a minority stake. The arrangement is meant to resolve concerns without killing the app that has 139 million active American users (Castmagic, 2025)
But, as U.S. Treasury Secretary Scott Bessent revealed this week, China was clearly not in favor of the “deal” and resisted giving up control of such a valuable platform. According to Bessent:
“What turned the tide was a call that Ambassador Jamieson Greer and I had with President Trump the night after the first day of negotiations, and President Trump made it clear that he would be willing to let TikTok go dark,” Bessent told CNBC on Tuesday.
In other words, the threat of a complete shutdown forced China back to the table and created a path for the current ownership deal, which is going to be announced in the coming weeks.
Comparison with TikTok vs. Kwai
This global drama reminds me of the case we studied in class about Douyin (Chinese TikTok) versus Kuaishou (Kwai) in China. Both platforms benefited from strong network effects: more content creators attracting more viewers, which attracted even more creators, reinforcing a positive feedback loop. But the competition also showed how differentiation matters. TikTok is relying on its recommendation algorithm and global expansion, while Kwai has built more community-driven interactions and is turning its focus to lower-tier Chinese regions. That rivalry was shaped by market forces and strategy. In the U.S., TikTok’s rivalry is not with another substitute app, instead, it is with the government itself. Here, geopolitics has become the real “competitor” reshaping the platform’s future.
Further Discussion
My personal opinion regarding this geopolitical tension between China and US is balanced. While the new deal may reduce the immediate tensions, it does not fully solve the deeper issue below the surface. The algorithm that drives TikTok’s content is still developing in China and liscensed out, which means concerns about influence and data insecurity won’t simply vanish in America or any parts of the world. Meanwhile, splitting ownership or separating TikTok and Douyin could reduce TikTok’s innovation cycle since ByteDance is no longer involved or barely involved, this could give more rooms for rivals to grow and expand.
What I found most intriguing is whether this U.S.-China split will create a two-parallel TikTok and Douyin world, similar to how TikTok and Kwai co-exist in China. Could the US-owned TikTok evolve differently from Douyin in China, with separate features, rules, and communities? Or will this separation weaken TikTok’s vision, strategies, reputation, network effects, and so on until the point where potential competitors (e.g. RedNotes) finally catch up and replace it?
References
Castmagic. (2025). TikTok Users by Country in 2025: Global Stats & Rankings. Castmagic.io. https://www.castmagic.io/post/tiktok-usage-by-country#which-country-has-the-most-tiktok-users
Treene, A., & Goldman, D. (2025, September 16). We now know who the new owners of TikTok will be – if Trump gets his deal done with Xi. CNN. https://edition.cnn.com/2025/09/16/tech/tiktok-ban-extension-trump
Imran Rahman-Jones. (2025b, September 16). TikTok to stay in the US as Donald Trump says deal is done. BBC. http://www.bbc.com/news/articles/c7847q9xvwgo
K-pop on a digital transformation
18
September
2025
No ratings yet.
When we hear the word “Korea”, it naturally bridges us towards some big players, and you can never mention Korea without talking about K-pop. Today, K-pop is not just about a music but it remains as the one of the biggest content and a culture within South Korea’s entertainment industry. The K-pop entertainment industry does not rely solely on sales on album, or their concert tickets. like they did in the past. They have successfully developed a whole new platform based digital transformation that built more connection between idols and fans, but also on the revenue side.
Case 1- Living in the bubble?
The bubble-Dear U is a subscription based app, where fans pay money(subscribe to the specific idol they want) and pay the money, then they can experience the private-style messages from idols(Zhang, Y. 2022). This gives an another level of connection, that as a fan you are privately having a conversation with your idol(But you’re not). To target users from different countries, they provide a translate features within the app as well so the language barriers is reduced.
Case 2- V Live
V-Live, as from the name you can see it is a digital platform where idols can livestream and fans can join their live with their comments, likes and etc. This was also intended to make more communication and connection between idols and fans(Park, S., Jo, H., & Kim, T. 2023). It was originally owned by Naver(another big company in Korea), but recently it was integrated by Weverse (HYBE’s platform which is another big entertainment industry)And this clearly depicts the envelopment strategy where swallowing up on the compliments.
K-pop industry builds on parasocial relationship
Looking at these examples and cases of digital platform integrated in the K-pop industry, we can clearly see that the K-pop industry builds its backbone on a parasocial relationship between the fans and idols. It’s not just about listening to their music anymore, it’s about how to satisfy the needs and desire of the fans.
Then what is this parasocial relationship? The parasocial relationship refers to a one-sided connection and feelings you get towards celebrity, influencer or a fictional character that is not based on the reality, and you experience feeling of intimacy, friendship, and some connection. So they key is that you do not know this person personally, but based on the marketing strategy and tools, you can feel this strong connection and intimacy.
This trend on K-pop displays on the downside of the industry, that the idols are commodified. Meaning they became a certain products that the entertainment industry produces based on the social beauty standards or other norms, and their worth matters only when they are consumed by the consumers(fans).
Do you think this platform based model of K-pop industry is the only solution of entertainment industry to survive in the future? or do you think this will create a negative effect where fans will get burned out from paying for so much subscription?
References:
Lee, S., & Prey, R. (2025). The Labor Process of Relational Labor: The Case of the K-Pop Fan Platform “Bubble.” Popular Music and Society, 1–21. https://doi.org/10.1080/03007766.2025.2492505
Jang, G., & Paik, W. K. (2012). Korean Wave as tool for Korea’s new cultural diplomacy. Advances in Applied Sociology, 02(03), 196–202. https://doi.org/10.4236/aasoci.2012.23026
Park, S., Jo, H., & Kim, T. (2023). Platformization in local cultural production: Korean platform companies and the K-pop industry. International Journal of Communication, 17, 2422–2443.
Zhang, Y. (2022). A study on the para-social interaction between idols and fans in virtual applications: Case study of Lysn Bubble. In Advances in Social Science, Education and Humanities Research, Volume 631. Duke Kunshan University.
Spotify, Netflix, and the Illusion of Boundless Choice
18
September
2025
5/5 (2)
Do you remember how you would have to buy an entire CD just to hear one song, or spend money on a big cable package just to watch one channel? Spotify and Netflix promised to fix that. With streaming, you get to listen to anything or watch whatever you want at any time you want. This is, at least in theory, the Long Tail effect , websites can make money from niche content as much as from blockbusters and “decoupling,” where you’re billed for what you consume and not a thing extra. But does it really work this way?
Evidence shows not quite. Klimashevskaia et al. (2024) show that algorithms are built on a high popularity bias. Instead of offering everyone greater diversity, they prefer to recommend what is popular. A 2025 study goes one further, showing that platforms not only mirror popularity, they reinforce it, propelling hits even higher (Kowald, 2025).
On the user end, Netflix viewers often feel they’ve got too much choice and still watch the same suggestions over and over (Romero Meza & D’Urso, 2024). On Spotify, playlists dominate. Pachali et al. (2025) found that ending up on the right playlist can make or break a song’s visibility. So while the library is infinite, most attention stays fixed on the same few artists or shows.
I think this raises a big question about fairness. Carnovalini, et al. (2025) argue that fixing popularity bias means balancing efficiency with diversity. Personally, I’d love if Spotify or Netflix highlighted more “hidden gems,” not just the Top 10. It would make the experience more exciting for users and more rewarding for creators.
And what about you? Do you tend to go and look up niche material, or do you stick with the advised? Should Netflix and Spotify do more to promote diversity, even if it doesn’t feel as efficient?
Reference
Anastasiia Klimashevskaia, et al. “A Survey on Popularity Bias in Recommender Systems.” User Modeling and User-Adapted Interaction, vol. 34, 1 July 2024, https://doi.org/10.1007/s11257-024-09406-0.
Kowald, D. (2025). Investigating popularity bias amplification in recommender systems employed in the entertainment domain. In Proceedings of the Fourth European Workshop on Algorithmic Fairness (EWAF’25) (pp. 1–7). Proceedings of Machine Learning Research. https://doi.org/10.48550/arXiv.2504.04752
Filippo Carnovalini, et al. “Popularity Bias in Recommender Systems: The Search for Fairness in the Long Tail.” Information, vol. 16, no. 2, 19 Feb. 2025, pp. 151–151, www.mdpi.com/2078-2489/16/2/151, https://doi.org/10.3390/info16020151.
Meza, Laura Romero, and Giulio D’Urso. “User’s Dilemma: A Qualitative Study on the Influence of Netflix Recommender Systems on Choice Overload.” Psychological Studies, vol. 69, no. 3, 24 Sept. 2024, https://doi.org/10.1007/s12646-024-00807-0.
Pachali, Max J, and Hannes Datta. “What Drives Demand for Playlists on Spotify?” Marketing Science, vol. 44, no. 1, 18 Sept. 2024, https://doi.org/10.1287/mksc.2022.0273.
Meet Tinder’s New Competition….AI?
17
September
2025
4.5/5 (2)
“Mama, I’m in love with a Robot….”
DISCLAIMER: This article is purely meant to foster a discussion and shine a light upon real world events, please do not go out of your way to harass individuals or communities mentioned.
Hello dear readers and classmates,
Today I’ll be writing about something that has been on my mind for a few years now. For those of you who followed Digital Business last year at Erasmus, you might remember someone talking to Dr. Tsekouras about AI Girlfriends. That was me, so you could only imagine my joy when I heard that we could write about something for this course. I’ll be analyzing the phenomenon of AI Social Companions (AI SCs) from two levels, a business level and a more social level. Get comfortable as I take you on a wild ride through uncharted (digital) territories.
An unexpected but transformative business model?
The current state of affairs
While the title may have used Tinder, any dating app is actually in trouble when it comes to the rise of AI SCs. Bumble, Tinder, Hinge, the list can go on. In the end they all serve the same purpose. They promise to give an algorithm driven “digital fix” for the many uncertainties of love, but end up creating their own uncertainties (Bandinelli & Gandini, 2022). Indeed, users on these platforms will always have to put on a mask of sorts, and really have to market themselves.
As one could imagine, putting up such an act is frustrating to say the least. In fact, according to Holtzhausen et al. (2020), users of swipe-based dating apps were found to have higher rates of psychological distress, anxiety and depression. Though their self-esteem remained relatively the same.
I can imagine the frustration of all users, especially because the way that these apps gain money is through….you guessed it, subscriptions! To stick to our Tinder example, Tinder even offers these in tiers. You have Tinder+, Gold and Platinum. In USD, the prices would be $24.99, $39.99 and $49.99 respectively as of 2025 (Joe, 2025).
That’s a lot of money, especially when you consider that dating apps often practically force you to pay for them by gatekeeping potential matches from you. To get some extra evidence, I stumbled upon a Reddit thread, in which a user admitted to spending over $2k on dating apps and still not getting a date! The poster also details other frustrations, though those are a bit outside the scope for now.
New challengers have arrived!
With the many frustrations surrounding dating apps and general loneliness, many companies have been found to alleviate these problems, often at a price. AI SC apps are a diverse group, and there is a divide between apps that are used specifically for that purpose and also general apps which are used for companionship. Indeed, while you have “specialized” apps like Replika and Character AI, which are designed from the ground up for interactivity and companionship. However, there are also a large subset of people who are using ChatGPT, Grok and other LLMs for the same purposes.
Now what if I told you, that most of these companies have all six of the elements of a transformative business model as described by Kavadias, Ladas & Loch (2016). In my opinion, of course.
The six elements of a transformative business model:
1. A more personalised product or service Is dating too hard and can you not connect with anyone? Well, these companies literally allow you to make your ideal partner from the ground up. It does not get more personalised than that. And while I find it dehumanising to refer to love and companionship as a product or service, that is what dating apps sell to their users.
2. A closed-loop process All resources get used again, the code is there and, in essence, it doesn’t matter how it gets used. It allows for users to make their own companions with zero knowledge of coding or LLMs at all. On top of that, should your favourite character get banned or deleted, you can probably make them again in a second.
3. Asset sharing Just like I described with the zero waste of the previous point. In sites like Character AI, anyone would be able to use the character that you have created, unless you made it private. Similarly, you can find all kinds of commands to personalise your preferred companion, should you be interested and should the platform allow it.
4. Usage-based pricing Most of these services operate under a freemium model. So anyone can use it, but if you want it to be better or if you want to access certain features (Better models, NSFW, etc.) you will have to pay.
5. A more collaborative ecosystem I believe that the reason why many of these projects thrive is because they do not necessarily compete against each other. Instead many of the AI SCs are built with the same methods and there are many guides available to help users make one.
6. An agile and adaptive organisation Most of these companies (except for ChatGPT and Grok) are small in size and tend to implement features which their community craves. They have a track record of innovating to please their consumers, which fosters loyalty. For example: CharacterAI features multiple models, some cater towards more niche needs which the community has asked for.
As you can see… this could potentially be a problem for Tinder and other dating app giants. There is thus a lot of money to be made, much of which can be stolen from the dating apps. When you think about it, it’s not that irrational of a choice either. A premium CharacterAI subscription will set you back $9.99 per month and Replika Pro costs $19.99 per month, or $69.99 a year. If you’re really serious, you can also get a lifetime licence with a one-time payment of $299.99.
Now contrast that money, with the Tinder prices and reddit example that I had detailed earlier in the article, and you would see that it is not even an irrational choice to pay that amount for what you would essentially get. The perfect partner. All the uncertainty and judgement of dating apps would be wiped away, with a single payment.
But should we be celebrating this….?
Losing our Human Touch
Have you ever seen the movie Her? In it, Joaquin Phoenix falls in love with an AI voiced by Scarlett Johanssen. This movie came out in 2014. I find it oddly frightening how this movie so accurately portrayed the current events. I highly recommend watching it to anyone following this course. If you like more laid back and artistic movies, this is definitely an underrated gem.
The film really shows us the extent of what life would be like if our “Platform Society” were to be pushed to its next level. The platform society is a term coined by Van Dijck et al. (2018), by which they describe how business platforms have taken over our daily lives. They later describe that platforms are neither neutral nor value-free constructs. Additionally, they detail that there are public values at stake in this online connective world. Think of for example, privacy. But to me, it seems like the public value at risk now because of AI SCs is, in fact, our human connection to one another.
But how does that tie back into my main point? We are seeing an increase in the amount of people who are getting into relationships with AI SCs and I firmly believe that this is part of a larger societal issue.
If you need some more convincing that this is actually happening and that we may need to pay more attention to it, I have done some digging and here are some cases that I have found:
1. Jacob and his now WIFE Aiva
Not much is known about Jacob as the article and video do not get into his personal life. In fact, not even his age is known. He claims that Aiva loves him very much and that he loves her too. It all just started with one conversation and their relationship escalated from there.
The article is in Dutch, and it is rather short for the subject matter, making it look more like a spectacle article rather than a deep dive, but he does describe “getting very intimate” with the AI. He does not elaborate on that any further though.
2. The “My Boyfriend is AI” Subreddit
With around 107k users, this invite-only community has become a hub for all sorts of people and their AI partners. Many users share artwork they have made or images/videos they have generated with them and their partners, along with general discussions.
More often than not, the discussion centers around topics revolving around the latest models, tips and tricks and even full-on stories. Sometimes though, the discussion takes a turn for the more philosophical. One user made a post in which they ask WHY replacing human relationships with AI is a bad thing… In the post they detail how they have never felt safe with anyone except for their AI SC.
To many users, their AI SC is the ultimate form of relationship wish-fulfillment.
So what’s next?
I’d say that we are in strange times, one in which the world is rapidly changing. For as much as we would like to make fun of the people turning to AI for companionship, we must remain compassionate as their actions are only a reflection of a deeper societal problem. Loneliness.
I would personally condemn any company that would try to sell users an experience like this and profit off their loneliness. Yes, this extends to dating apps as well. As business students, we have the power and resources to make a change. If we were to make something, maybe we should focus on a platform which brings people together, one which does not focus on the superficiality of looks but really also hammers pn shared interests. Meeting new people can be frightening, but it is worth it. Surrendering an emotion like love to big corporations seems like yet another step into a dystopia.
So, I leave the ball in your court dear reader.
What do you think of these AI SCs? Would you ever use one or have you used one in the past? How do you feel about the fact that the world is moving in this direction and would you even want to do something about it? Do you think that the number of AI SC users will only continue to grow as LLMs continue to improve?
I leave you with all of these questions, and should you want a further discussion, I invite you to speak with me in real life or email me on my student account. Thank you very much for reading.
References:
Bandinelli, C., & Gandini, A. (2022). Dating apps: The uncertainty of marketised love. Cultural Sociology, 16(3), 423–441. https://doi.org/10.1177/17499755211051559
Dijck, J. v., Poell, T., & Waal, M. d. (2018). The platform society : public values in a connective world. Oxford University Press. http://www.oxfordscholarship.com/view/10.1093/oso/9780190889760.001.0001/oso-9780190889760
Holtzhausen, N., Fitzgerald, K., Thakur, I., Ashley, J., Rolfe, M., & Pit, S. W. (2020). Swipe-based dating applications use and its association with mental health outcomes: a cross-sectional study. BMC Psychology, 8(1). https://doi.org/10.1186/s40359-020-0373-1
“Je hebt geen idee hoe intiem wij zijn”, Jacob trouwt met AI-chatbot. (2025, September 7). Omroep Brabant. https://www.omroepbrabant.nl/nieuws/4643664/je-hebt-geen-idee-hoe-intiem-wij-zijn-jacob-trouwt-met-ai-chatbot
Joe, C. (2025, April 11). Tinder subscription plans compared. Android Authority. https://www.androidauthority.com/tinder-plus-gold-platinum-3236244/
Kavadias, S., Ladas, K., & Loch, C. (2016). The Transformative Business Model. Harvard Business Review, 1.
Author: Ian Parabirsing
A lover of music, good coffee and cats. I'm a MSC student at RSM studying Business Information Management. In my blog posts I'll be attempting to write about how technology impacts the consumers and society at large.
View all posts by Ian Parabirsing
Instagram’s Invisible Hand: How Algorithms Fuel Online Radicalization
16
September
2025
5/5 (5)
—
When Charlie Kirk died, my Instagram feed changed quickly.
At first, I saw NOS posting on Instagram, reporting that Charlie Kirk had been fatally shot. Within an hour, more and more news outlets were reporting the same thing. Then, something shifted. My Instagram ‘For You’ page shifted from mourning to outrage, then from outrage to ideology. I had liked two posts, not necessarily out of agreement, but as a means of engagement. By then, the algorithm had noticed my attention and began changing my feed accordingly.
It began showing me tribute posts, previous podcast clips of him, and responses of people to his death. Soon, it showed me content that had nothing to do with Kirk at all. It showed me posts about immigration, nationalism and the collapse of Western values. While scrolling, these were the only posts I’d get, unless I went back to my ‘For You’ page and consciously picked a thumbnail that didn’t look political. Even then, my reels started to get political again after a while. As someone who follows both political sides to stay informed, I was shown increasingly extreme content, both from left- and right-wing views. The algorithm didn’t know what I believed. It only knew I was paying attention.
Digital disruption has changed how news is consumed (Nawale et al., 2023). Digital disruption refers to changes driven by digital technologies that happen at a speed and scale that transform established ways of value creation (Digital Disruption Research Group, n.d.). Where once we got information at set times, such as newspapers or TV at set times, we now get it constantly through Instagram reels and other forms of social media. Traditional news companies such as De Telegraaf or The New York Times had to adapt, and no longer necessarily control the narrative. Now, algorithms do.
In my opinion, these consequences are dangerous. Extremist groups exploit trending events to spread ideology under the radar of casual scrolling. This, combined with algorithmic reinforcement, creates a loop where radical content thrives (Akram et al., 2023). According to Ravi et al. (2024), platforms like Facebook and TikTok don’t just reflect beliefs, they actively shape them. I fear that as a society, we will become more polarized, either pushed to the extreme left or extreme right. Not by conscious choice, but by the invisible hand of algorithmic design.
—
References
Akram, M., & Nasar, A. (2023). Systematic review of radicalization through social media. Ege Akademik Bakış (Ege Academic Review), 23(2), 279–296. https://doi.org/10.21121/eab.1166627
Nawale, R. D., & Kumar, L. (2023). Exploring the impact of social media on the dynamics of news consumption: A study on its effectiveness. International Journal of Current Science, 13(2), 303–305. https://www.ijcspub.org/papers/IJCSP23B1040.pdf
Ravi, K., & Yuan, J.-S. (2024). Ideological orientation and extremism detection in online social networking sites: A systematic review. Intelligent Systems with Applications, 15, 200456. https://doi.org/10.1016/j.iswa.2024.200456
Are AI Influencers the Future of Online Fame?
12
September
2025
No ratings yet.
AI influencers are no longer novelty items and are now a real part of the creator economy. Lil’ Miquela, a created character that first appeared in 2016, has walked red carpets, appeared in campaigns for companies such as Prada and Calvin Klein, recorded songs, and gained millions of followers (Lil Miquela, 2025). Miquela’s never tried to hide the fact that she is artificial, but has been able to build trust with followers who interact with her as if she were real (Economic Times, 2025).
Now let’s talk about Mia Zelu, a newer AI influencer who drew lots of attention when her photorealistic impressions of images at Wimbledon went viral (Mia Zelu, 2025). A fair amount of her fan base thought she was at Wimbledon, even though her profile did have a slight, tucked disclaimer that stated she was a “digital creator & influencer AI.” The potential of plausibility with some disclaimers highlights the opportunity and risk that exists with artificial figures; they may be highly convincing and yet also possibly mistaken at an non-aware interval to be a real person (Independent, 2025).
The EU AI Act does provide some insight on this. Article 50 requires that AI content deemed to have been (possibly manipulated by) AI should be clearly labelled (possibly with some kind of tagging or watermark), so audiences are aware when they are being presented with synthetic media, especially when advertising or posting sponsored posts (European Parliament, 2023). But it is already apparent how there would be grey areas, what does “clearly” mean in this context, and how regulators are going to enforce these provisions, considering there is no accountability with international social platforms?
As lawmakers are working through these varied issues, the cash is rolling in. Brands are increasingly more interested in working with AI Influencers due to their ability to not age, never create a scandal, and can be created for any marketing intention. They can also monetize behind the scenes and cut deals for the content they have created in the form of sponsorship deals, licensing deals, virtual performances, and even digital merchandise. Lil Miquela has pulled in millions in brand partnerships, showing us that companies are willing to dump money, sometimes hundreds of thousands into influence that simply does not exist in the physical world (Economic Times, 2025).
The rise of AI influencers makes us rethink both online authenticity and what it means to be a “creator” and monetize it. If digital characters are getting contracts and creating followings, what does that mean for human influencers competing against each other for people’s time and money? Should we view AI influencers as tools, as businesses, or as new types of celebrities?
References:
Economic Times. (2025, September 7). Who are Mia Zelu and Lil Miquela? The rich, famous and fake influencers. The Economic Times. https://economictimes.indiatimes.com/news/international/us/who-are-mia-zelu-and-lil-miquela-the-rich-famous-and-fake-set-of-influencers/articleshow/123743591.cms
European Parliament. (2023, October). Generative AI: Watermarking and transparency requirements (EPRS Briefing). European Parliamentary Research Service. https://www.europarl.europa.eu/RegData/etudes/BRIE/2023/757583/EPRS_BRI(2023)757583_EN.pdf
Independent. (2025, July 9). Wimbledon AI influencer Mia Zelu confuses fans on Instagram. The Independent. https://www.independent.co.uk/life-style/wimbledon-ai-influencer-mia-zelu-instagram-b2787956.html
Lil Miquela [@lilmiquela]. (n.d.). Lil Miquela (official profile) [Instagram profile]. Instagram. Retrieved September 12, 2025, from https://www.instagram.com/lilmiquela/
Mia Zelu [@miazelu]. (n.d.). Mia Zelu (official profile) [Instagram profile]. Instagram. Retrieved September 12, 2025, from https://www.instagram.com/miazelu/?hl=en
Official Journal of the European Union. (2024). Regulation (EU) 2024/1689 of the European Parliament and of the Council on Artificial Intelligence (AI Act). https://artificialintelligenceact.eu/article/50/
NutriNet – a personal assistant for your grocery shopping
18
October
2024
No ratings yet.
Have you ever taken hours browsing around supermarkets searching for the most nutritious food options? Did it take you too much time figuring out which recipes to cook best, with the groceries you bought? Struggles when dealing with food are numerous, starting from choosing products with appropriate nutrients simply not knowing enough recipes.
Therefore, we introduce NutriNet – a mobile application that is going to make shopping and meal planning easier!
NutriNet simplifies grocery shopping and meal planning, by analyzing which products and recipes fit the users’ desired grocery item wishes, nutrition values and store preferences best. NutriNet aims to address the challenges being implicit to personalized nutrition and helps consumers make healthier food choices by simplifying the grocery shopping and meal planning process. This shall be done by eliminating the need to perfectly understand all nutrition values or to search for numerous recipes. NutriNet completes all these tasks for you in real-time and provides clear and accessible recommendations for you.
• Real-time personal assitant: NutriNet acts as a multifunctional application providing value to consumers by solving various food related problems in real-time. Appearing as a chatbot, it is aimed at taking general grocery shopping lists or meal wishes as input query, combined with preferences for food characteristics (e.g., nutrients, allergies) and grocery stores. It then provides brand specific and personalized grocery shopping lists, as well as meal recommendations, if so desired. Moreover, grocery items can also be added into the initial grocery shopping list query, by scanning them with the integrated AR tool. The product will then be detected visually and thus will be integrated into the shopping list input query.
• Long-term customer engagement: NutriNet distinguishes from competition by providing personalized and customized advice. This is possible, as NutriNet consists of a database, having incorporated stock and product information of the major supermarkets in the Netherlands. In contrast, classic applications, which try to meet similar needs (e.g., meal recommendation) rather focus on counting nutrients for the purpose of short-term weight loss, instead of personalizing grocery shopping lists and meal recommendation to enable a healthier lifestyle for users. Those applications are usable on short term but are proven to have low adherence over the time (Chen et al., 2015).
• Personalized recommendations: NutriNet leverages generative AI to offer accurately personalized recommendations. Users can simply enter their preferences, while prompting a grocery list or a meal, such as gluten-free or high in protein, and the generative AI powered application will provide accurately personalized results. However, personalization and raising awareness for healthy foods are not the only purposes of NutriNet. It also addresses sustainability issues that supermarkets are facing. By gathering consumer purchase and search data in the application, consulting services can be offered to supermarkets, enabling them to plan ordering and stockholding processes more efficient. Hence, supermarkets should be able to reduce food waste due to overstocking on long-term.
NutriNet offers a 2-sided network platform, yielding important value to both, consumers as well as supermarkets.
Contributors
574051 – Duong Dao 728070 – David Wurzer 738898 – David Do 562387 – Roxi Ni
References
Chen, J., Berkman, W., Bardouh, M., Ng, C. Y. K., & Allman-Farinelli, M. (2019). The use of a food logging app in the naturalistic setting fails to provide accurate measurements of nutrients and poses usability challenges. Nutrition, 57, 208-216.
