Personalized deep learning equips robots for autism therapy
news.mit.edu
news.mit.edu
No they don't. Being physically animated makes a difference in how robots feel to be around. Aspects of their movement, sound and appearance are interpreted as affect. Researchers are able to roughly manipulate the type and amount of affect and other social cues displayed by robots, which is what makes them useful for research and therapy with autistic children.
This isn't my area, but it's a well-established niche in robotics research.
Of course, many (mental health workers) are amazing, even seemingly super-human, some of them. No question about that.
But, actively caring for and defending such a marginalized and misunderstood population are never congruent with an astute professional career path in a capitalist society. I am from the US, and this may be different in other countries.
In US social mental health programs, we have been implementing scorched earth policing policies and enforcing them in the mental health clinics. One frightening issue at the moment involves state policies that give mental health facilities quotas for drug testing their patients. This is effectively leveraging mental health needs to criminalize and deny services. It’s most common in Republican states but rampant everywhere. Professional mental health workers are by and large non-critical and uninspired to take action.
So how does this effect autistic patients? In the same way that autism leads to issues with hiring processes and, confrontations with law enforcement, housing, etc. Autism falls victim to prejudice of mental health professionals according to a bias at the level of communication.
The policies are of course described as safety measures. Despite their likelyhood of encouraging suicide, the rapid rise in the US suicide rate + opioid crisis is currently being leveraged as the justification for such policies. Nevermind that opioids are not used in mental health treatments and that both issues are geographically strongly correlated with trends in economic displacement. Nothing new under the sun.
How is it possible to generalize to a population of 320 million?