It’s 2025, and we live in a world where artificial intelligence has optimized nearly everything. Music, movies, dating, workouts, grocery shopping, therapy — if there’s a process, an algorithm is working behind the scenes to make it faster, cheaper, and more efficient.
And yet, somehow, life feels… flatter.
Where’s the thrill? The magic? The messy serendipity that used to make things worth doing?
We’ve entered what I call the era of Algorithmic Marginal Utility — a world where every aspect of our experience is so optimized that the difference between good and great has become vanishingly small. It’s like music on shuffle forever: no skips, but also no bangers that hit you in the chest.
So the question is: If everything gets better, why does it feel worse?
What Is Algorithmic Marginal Utility?
In classical economics, marginal utility is the added satisfaction from consuming one more unit of a good or service. The more you have of something, the less you feel its impact.
Now apply that idea to your digital life. AI recommendation engines flood you with hyper-relevant content: the next perfect video, the next perfect song, the next perfect snack. You’re always just “one more scroll” away from another dopamine hit.
Image courtesy own creation via Alex Pawlowski (ChatGPT)
But over time, you stop feeling anything at all.
That’s because AI doesn’t care about emotional arcs or delayed gratification — it optimizes for now. It’s good at reducing pain, but terrible at creating meaning.
AI and the Tyranny of Optimization
Think about Spotify in 2025. You open the app and it knows your mood, time of day, and even your recent search history. It delivers you a perfectly personalized playlist — a vibe cocktail. It’s pleasant. It’s chill. It’s… forgettable.
Now contrast that with the feeling of discovering a mixtape your crush made in high school — chaos, awkward transitions, weird track choices. But real. Personal. Emotional.
AI is making everything smooth. And in doing so, it’s sanding off the edges where meaning used to live.
Same goes for:
- Netflix: You don’t choose movies anymore — the algorithm autoplays something you’ll probably like.
- Amazon: You buy what’s suggested before you even realize you needed it.
- Dating apps: Your matches are statistically optimized for you… yet first dates feel more like job interviews than sparks flying.
Behavioural Economics: The “Why” Behind the Flatness
AI’s effect on human experience is best understood through the lens of behavioral economics:
- Hedonic Adaptation: The more you consume pleasure, the less intense your reaction. AI serves you an endless buffet of mid-tier joy.
- Paradox of Choice: While AI narrows your options to reduce decision fatigue, it also robs you of exploration and ownership.
- Satisficing vs. Maximizing: AI assumes we want to find “good enough” fast. But humans often enjoy the search, the frustration, the learning curve.
We’re built for narrative, not efficiency.
Culture in the Age of Algorithms
What’s happening to culture when everything is data-driven?
- Music: Genres blur into algorithm-friendly homogeneity. Playlists optimized for productivity, focus, or sleep — not passion, rebellion, or heartbreak.
- Film & TV: AI-generated scripts are now a real thing. Formulaic plots tested for audience retention. Zero risks, zero soul.
- Fashion: Microtrends birthed by TikTok and killed by recommendation engines before they become actual movements.
We’re seeing the death of the subculture. Anything interesting gets surfaced, monetized, and flattened in record time.
And with generative AI tools like Midjourney and Sora now in everyone’s hands, anyone can make anything — but fewer people care to.
Welcome to the Feedback Loop
AI optimizes what works. It studies what you like… based on what you previously liked. So if you liked indie music last year, it’ll keep you on a slow drip of adjacent content. It won’t introduce you to Mongolian throat singing or 1980s Japanese city pop unless you ask — and even then, it’s probably a sanitized version of it.
This is the trap of comfort.
Over time, your world narrows. The peaks and valleys flatten. Your curiosity atrophies. And the systems interpret that as success.
Real World Case: Spotify vs. the Live Scene
Take the ongoing debate in 2025 about live music vs. streaming. While Spotify continues to dominate audio consumption, we’re seeing a counter-movement: a resurgence in live shows, vinyl, and “slow listening.”
Fans crave the rawness of a live set, the imperfection of a cracked voice or a missed note. That’s where memory lives. That’s where soul lives.
Artists, too, are fighting back. Björk, for instance, has been deliberately releasing “unstreamable” tracks — experimental, non-optimized, and long-form — as a rebellion against algorithmic aesthetics.
What Can We Do About It?
We’re not anti-AI. This isn’t a luddite rant. But we need new design principles for digital life — ones that reintroduce friction, play, and emotional spikes.
Some ideas:
- Intentional friction: Spaces that don’t track you, don’t recommend, don’t autocomplete.
- Human-first curation: Bring back tastemakers — DJs, critics, weird friends.
- Savor over speed: Value the time it takes to love something. Make effort part of the experience.
- Build for awe, not just ease: Tech should make life better, not just more convenient.
Final Perspective: Are We Optimizing Ourselves Out of Being Human?
In the quest to make everything smoother, we may be sanding down our humanity.
We are creatures of chaos, longing, mystery, contradiction.
AI, in its current form, is none of those things. It offers consistency, not catharsis.
Prediction, not surprise. Satisfaction, not joy.
The challenge of our generation won’t be whether AI takes our jobs.
It will be whether we let it take our wonder.
About: Alex Michael Pawlowski is an advisor, investor and author who writes about topics around technology and international business.
For contact, collaboration or business inquiries please get in touch via [email protected].
Source:
[1] Ahmad, Z. F., Lelis, L. H. S., & Bowling, M. (2020). Marginal Utility for Planning in Continuous or Large Discrete Action Spaces. Proceedings of the 34th Conference on Neural Information Processing Systems (NeurIPS 2020).
[2] Khan Academy. (n.d.). Total utility and marginal utility. Retrieved from https://www.khanacademy.org/economics-finance-domain/microeconomics/choices-opp-cost-tutorial/marginal-utility-tutorial/a/lesson-overview-total-utility-marginal-utility-and-utility-maximization
[3] The Motley Fool. (2025). What is Marginal Utility? Overview and Examples. Retrieved from https://www.fool.com/terms/m/marginal-utility/
[4] FasterCapital. (2025). Marginal utility: The Theory of Price and Maximizing Customer Satisfaction. Retrieved from https://fastercapital.com/content/Marginal-utility--The-Theory-of-Price-and-Maximizing-Customer-Satisfaction.html
[5] ResearchGate. (2020). Diminishing Marginal Utility of Wealth. Retrieved from https://www.researchgate.net/figure/Diminishing-Marginal-Utility-of-Wealth_fig1_339173853
[6] SlideTeam. (n.d.). Marginal Utility Curve PowerPoint Template Slide. Retrieved from https://www.slideteam.net/marginal-utility-curve-powerpoint-template-slide.html
[7] Simplus. (n.d.). Infographic: AI + Energy/Utilities. Retrieved from https://www.simplus.com/infographic-ai-energy-utilities/
[8] Smart Energy Consumer Collaborative. (2024). Computing Consumer Interest in AI Infographic. Retrieved from https://smartenergycc.org/computing-consumer-interest-in-ai-infographic/