Between 1971 and 2020, ‘Natural’ spaces in urban areas in Hong Kong, which include green areas, planting trees, and ornamental landscape features, increased in quantity at an average annual rate of 0.26% within the urban area of Hong Kong (World Data 2020). These provide residents with the opportunity to connect with and experience nature (Dunn et al., 2006; Palliwoda et al., 2017). The health benefits of this are increasingly being recognised (Soga and Gaston 2016; Mell 2017), with evidence of an enhancement in both the physical and mental health of people living near parks and natural spaces (Lovell et al., 2014). Although a small number of countries have created policies and strategies to increase opportunities for people to visit natural spaces, these tend to be at a macro, national level rather than considering in detail biophilic design, as in design for the love and appreciation of nature, at a local and neighbourhood level (Mell 2017; Ten Brink et al., 2016). This is a vital component, when considered in the context of the fundamental disconnection to nature that is occurring for so many who join the continuing migration pattern from the rural to the urban (Soga and Gaston 2016). Within this context, this paper uses the small urban pocket parks in Hong Kong as a testing ground, assessed using predictive Artificial Intelligence (AI) as a methodology to to assess and understand the extent to which the character of these spaces may influence a better connection to nature within urban areas.