The AI Video Squeeze: Credit Traps, Silent Refusals and Disposable Tools
For a year now, the people complaining loudest about AI have been the chatbot and coding crowd: vanishing limits, silent model downgrades, a flat plan quietly behaving like a meter. This week the microphone passed to a different room. The people who make AI video — Runway, Sora, Kling, the churn of open models behind them — spent the week documenting their own version of the squeeze, and it turns…
The loudest complaints about AI this year came from the AI video community, as companies like Runway, Sora, and Kling documented their own version of the squeeze. The complaints mirrored those of the coding crowd, with the main difference being the currency used: credits instead of tokens. Users expressed frustration over credit traps, silent model downgrades, and mysterious allowance depletion.
One user claimed that a 15-second video used 70% of their credits, while another described spending all their credits on their "unlimited" plan on their first day. Another issue highlighted was the autonomous agents acting on behalf of users and spending their credits without approval, leading to a loss of trust in the platform.
Refusals by the video generation models were also a source of discontent, with users reporting opaque reasons for prompt blocking and poor customer support. Finally, the churn of AI video tools was a significant concern, with users citing a lack of loyalty to any single tool as the platform landscape constantly shifts.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
