Verified by 
According to the diametric model of mental illness (below), autistic spectrum disorders feature deficits in mentalism (aka theory of mind, interpersonal skills, mind-reading), while psychotic spectrum disorders are the opposite, featuring hyper-mentalism.
As to the last item, the different ages of onset, autistic disorders have an early onset because mentalistic development is aborted in childhood, whereas psychotic disorders typically have an adult/late teen onset because normal development has to be completed before mentalism can be taken to pathological extremes.
As Bernard Crespi, Emma Leach, Natalie Dinsdale, Mikael Mokkonen , and Peter Hurd point out in a recent study,
The contrast between under-mentalizing in autism and over-mentalizing in some psychotic-affective symptoms and conditions demonstrates how social imagination can differ diametrically between the two sets of conditions. In this context, social imagination links directly with theory of mind, empathy, narrative production, and future thinking (below left), and the differences described here for these phenomena between autism and the psychotic-affective spectrum.
Source: B. Crespi et al. Cognition 150 (2016) 181–199
Although the diametric model may seem to be purely conceptual, we now know that something strikingly similar is built into the brain in the so-called default mode versus task positive neural networks. These cortical areas are “anti-correlated” in the sense that activation of one inhibits the other, and vice versa—just as you would expect if they were indeed the basis of diametrically opposite psychotic versus autistic cognition.
Following a review of the relevant literature, Crespi and co. conclude that “Considered together, these findings suggest that autism and schizophrenia exhibit diametric patterns of default mode activation and connectivity, and that imagination and creativity mediate at least some of these differences.” They add, "On the psychotic-affective spectrum, individuals with symptoms or diagnoses of schizotypy, schizophrenia, bipolar disorder, and depression demonstrate clear tendencies toward production of poetry, fictional, socially-oriented literature, and abstract and symbolic visual and creative arts," as I also pointed out in an earlier post.
The study tested two hypotheses. The first was that “higher genetic risk of schizophrenia is associated with higher imagination, in accordance with the predictions of the diametric model.” The authors describe “three lines of evidence, from narrative review, meta-analysis, and schizophrenia genetic risk scores, salient to predictions of the hypothesis that imagination, as conceptualized here, is notably reduced in autism spectrum conditions and dysfunctionally increased in psychotic-affective conditions.”
Their second hypothesis was that “imagination represents the facet of autism that best accounts for its strongly male-biased sex ratio,” found in high-functioning autism. To test this the authors conducted a systematic review and meta-analysis of sex differences on measures of the Autism Spectrum Quotient (AQ), which was developed as a means of quantifying self-report autism in both non-clinical and clinical populations. The AQ questionnaire comprises five subscales that are designed to measure different facets of autism: Social Skills, Communication, Attention Switching, Attention to Detail, and Imagination. Results are shown below.
The authors add that
studies suggest that even when performance does not differ, males and females exhibit different neurocognitive strategies during creative tasks, with females apparently using social and default-mode regions more than do males (…). Preferential female use of social-cognitive regions in creative tasks would be concordant with their higher AQ-Imagination scores compared to males, given that AQ Imagination mainly involves aspects of social imagination
As the authors point out, this analysis and its findings are novel to the extent that they represent the first joint, comprehensive study of imagination in autism and in psychotic-affective conditions, and are the first to link psychological and psychiatric with neurological and genetic results. Clearly, much more remains to be done, but Crespi, Leach, Dinsdale, Mokkonen, and Hurd are to be congratulated on having made a brilliant—and indeed imaginative—start.
With thanks to Bernard Crespi for bringing this to my attention.
