General-purpose AI was the most common substitute — named by 40% of respondents and relied on most by 27% — but the authors frame this as an access gap, not a preference: the leading reasons for choosing it were instant availability (78%) and cost (74%), and many described these tools as lacking the memory, continuity, and structured reflection they valued in Ash.

The 26% who sought no support at all reported the worst outcomes, with 81% saying their access to support had worsened. Across the full sample, 60% reported worse access since losing Ash, and roughly nine in ten said they would use Ash again if it returned under a clear wellbeing framework.

The authors note important limitations: respondents were engaged former users, so selection bias likely inflates positive sentiment; all measures are self-reported and cross-sectional; and Slingshot AI is the product's developer — a conflict they disclose and address through transparent reporting, including users who found general-purpose AI adequate.