Not just measuring sound, but recognizing events.

Believable Agents - Woggles

Russell’s theory - Core affect

Machine learning & technology tools may often stem from disability context

• Deaf -> speech recognition
• Emotion -> autism

It is funny how we’re using emotionally handicapped computer as an assistance to people who we deemare emotionally handicapped.

Affective computing.

Productive parallelism between the ways that autism mitigation programs view autistic individuals and the ways that surveillance capitalism views platform users, e.g, call center employee evaluation

Humans must validate their humanity to AI systems - > reverse Turing test

Response to The Native Ear - Michelle Pfeiffer

In the recent years, I assume that the global has largely adopted the “English American” accent given the rise of social media and globalization of Western content. How would gen alpha (who probably has adapted to the English/American accent) be adapting to the accent test? How can accent still be valid as identification if our world converges into singularity?

There’s also a case where speaker accentuates their accent even more to ensure people (and maybe themselves) that they come from that accent origin. For example, I, despite being as a native Bahasa Indonesia speaker, have spent the recent years speaking more English than before. However, sometimes I found myself fabricating my Indonesian accent into the English sentences so that I made it clear to listeners (who are not native Bahasa speaker) that I’m not a native English speaker. As mentioned by Pfeiffer, the accent lies in the listener, when the listener doesn’t fully understand the natural accent of Bahasa Indonesia, how can they differentiate whether I fake accentuate my Bahasa Indonesia accent or just use my natural English tone (with a subtle underlying Bahasa Indonesia accent)? “The listener’s role then becomes to be able to hear those signs of “putting on an accent (Pfeiffer)..”

Accent becomes testimony.

Mechanistic Interpretability.