IPh.D. linguist and Eastern Europe domain expert passionate about technology and communication with a desire to build better language technology for everyone.
Learning about effective conversation design, or how to climb the tree and not have it climb back at you (that's a quote from a GU professor, no credit is mine), is the art of "What could possibly go wrong?" when language and technology intersect. #CxD
My favorite comment today. 💡
“It’s super weird when you start thinking about where language-knowledge ends and world-knowledge begins” by Ellen Lucast.
#language#knowledge#ontology#AI#GPT3#NLProc
I'm looking to hire a contractor PM and three contractor linguists for three months (April 1--June 30) as part of a new initiative on the fairness of large-scale language technologies.
Designing my first conversation with a chatbot.
.
.
.
The linguist in me thinks: Don't forget about all the possible variations!
The user in me thinks: Don't overdo it. Language is the tool, not the end goal.
#CxD#chatbots#language#conversation#design#dialogue#flow#bots
As an interface, voice gained another category of users so traditional human-to-human assumptions no longer apply.
Designing conversational AI and learning to communicate with machines becomes similar to acquiring a new language. And it sucks to fail at language.
#Voice#AI#UX
This is interesting.
Failure to communicate with voice technology unsettles humans whose voice communication skills have never been evaluated to that extent before. #CxD#ConversationalAI#VoiceFirst https://t.co/tL3FeuJ9E5
Back in the late 1990s when I was in elementary school, we'd get a chance to spend an hour or two on those old, DOS-based computers. I would type up questions for the computer and hit Enter. No answers were given.
Now that's possible and it's called conversation design. #CxD
Learning about effective conversation design, or how to climb the tree and not have it climb back at you (that's a quote from a GU professor, no credit is mine), is the art of "What could possibly go wrong?" when language and technology intersect. #CxD
“Everybody that has ever written a story knows this: your first draft sucks.”
It took me an incredibly long time to realize that good texts aren’t written once, but rewritten many times. And that the first draft is just that: they first draft. #CxD
https://t.co/fnL2X8zjwv
The cited example of a nurse is a typical culprit. I’m also thinking about languages in which some professions/occupations lack equivalents for another gender, or those equivalents are very recent without wider adoption. When social need outpaces language, problems will happen.
Machine translation is hard. Gender bias in machine translation is even harder. @gedinu
“One clear area of bias in machine translation… is gender stereotyping when translating from a language with ungendered nouns to one with gendered nouns.” https://t.co/aufB4GlnyA
🚨🚨🚨
The FATE group at @MSFTResearch NYC is hiring a postdoc whose work focuses on the societal implications of AI&ML!
📢📢📢
Does that sound like you? Come join us!
Group: https://t.co/OO9GPXD5Xk
Job Ad: https://t.co/Hvq1h3Yqhc
Happy to answer any questions you may have!
🗣️ Actions on Google now supports the <lang> tag.
Use it to change the spoken languages throughout your response. Learn more here → https://t.co/UGj0XR36GW
Speech Recognition has progressed tremendously over the last few years, but what key problems are still being tackled in the industry today? Join our Chief Scientist for a discussion on this topic tomorrow morning! https://t.co/6oFa1QjNjK #VoiceFirst#VoiceUI
📅 Happening today: NLP at Microsoft Research: "Ask Me Anything" with Alexandra Olteanu (Microsoft Research Montreal) and Hanna Wallach (Microsoft Research NYC)
Register here: https://t.co/0QO3OsSBAA
#NLP#NLProc#Microsoft#MSFT#WiNLP2021