The Web needs a voice - literally, technically, politically

Michael Henretty, Ola Gasidlo, Cathleen Berger This talk will present ideas for how to tackle three dimensions of the future Web: voice recognition, web compatibility, and digital inclusion. When your latest Siri, Alexa or Cortana fail to answer your questions because they seem to simply not understand your instructions, that’s probably because you’re not a (white) male with a mainstream American accent. Voice recognition tools are only as good as their training data, if the data set doesn’t include all sort of accents, dialects, and dynamic adaptation, the systems we built on top of it will never fully reap their potentials. Creating an inclusive, and freely available database to train speech algorithms comes with its challenges - but they’re not unsolvable. And once we have that? That’s when we need to translate and connect these tools to our everyday web experience, writing code that is accessible, compatible, and adaptable across whichever platform you use to access the Web. We can already see a push to create a speech driven future of the Web, but while this is still a vision for most users, the technical questions underlying this development are being addressed now - which is why we need to make sure our answers contribute to a healthy environment. And once we have the training data and the standards? That’s when our political contexts will be yet again put to a test: how can we make sure our societies continue to grow in an open and inclusive manner? What do we need to pay attention to, if we don’t want technologies to become our oppressors? It’s in our hands to use the tools we create to make sure they give power to the people - because POP culture needs a voice.