At the Human Clarity Institute, we’re seeing early signs that the convenience of AI carries a quieter price — on human energy and on the world that powers our feeds.
This is an emerging pattern we expect to grow in the coming years. As AI systems generate more of what we watch, read, and react to, our time online can start to feel efficient yet strangely exhausting. Friction disappears, but so does the sense of pace and pause that protects attention.
There’s also a wider cost we rarely feel in the moment: the power required to train and run the models behind the content. When the stream never ends, neither does the demand for energy. Personal depletion and planetary strain begin to rhyme.
What our 2025 data shows
- Emotional dissonance: When people find out content was AI-made, top feelings include “annoyed”, “disappointed”, and “cheated”.
- Trust deficit: 49.55% say they do not at all trust big tech companies to use AI responsibly.
- Growing unease: Roughly two in three think about AI’s impact on their daily life at least occasionally.
- Value concerns: Half worry that AI will make human creativity less valued (28% “quite a lot”, 22% “very much”).
- Bothered by synthetic content: Most respondents report at least slight to very strong discomfort when what they consume isn’t human-created.
Source: HCI Digital Life 2025 dataset (n = 1,003).
Why this matters
AI promises speed, scale, and ease. Yet our attention and emotions still move at human pace. When feeds accelerate, many people report feeling more drained, less grounded, and less certain about what to trust.
At the same time, every effortless interaction sits on real infrastructure. The more we rely on synthetic content to fill gaps in entertainment, learning, or decision-making, the more invisible energy we consume in the background. Personal fatigue and environmental load may be two sides of the same behavioural habit: endlessly asking machines for more.
This is not a call to disengage from AI — it’s an invitation to use it deliberately: slower when it helps, paused when it doesn’t, and transparent about the costs we prefer not to see.
Future signals we’re watching
- Rising reports of “AI fatigue” — tiredness after long sessions with generative tools or AI-curated feeds.
- Growing public conversation about the energy footprint of everyday AI use.
- Design patterns that re-introduce healthy friction (session limits, slower defaults, clearer labelling of AI-made media).
We expect these discussions to move from niche forums into mainstream policy, design standards, and personal wellbeing habits.
Have you noticed AI tools making your work or leisure feel efficient but more tiring? We’re collecting early experiences to inform future HCI studies on AI fatigue and behaviour. Share your story with us.