A presentation at Frontend Conference in in Zürich, Switzerland by Desiree Garcia
e g a l l i v a s e k a t It t o b a e s i a r to Desiree García @thedezzie
Conversational systems and emerging AI
"Just add water"
"One size fits all"
Engineers couldn't do it alone.
Why do people want conversational systems?
“Don’t play what’s there, play what’s not there.”
Machines still have a hard time with language http://bit.ly/ai-language-problems
Simple conversational systems that are great
Complex conversational systems that are great
Time + Money
Is it convincing?
Is it convincing effective?
Content first
Content last
Intents Entities Dialog builder #greeting Conversation starts Condition #greeting Watson responses (0) No responses added Edit . Order pizza flow and some … Order pizza flow and some … Condition Condition No conditions set No conditions set Watson responses 20 We got ya down for a $topping pizza. Anything else Watson responses 20 We got ya down for a $topping pizza. Introduction
@location . @restaurant @cuisine
"In just minutes a day"
Getting people to train a conversational system is as easy as getting people to start exercising .
. http://www.r2d3.us/
Initial training ingestion employee manual Initial training improvement apprenticeship ongoing training refinement mastery
Content matters, because training matters
Deploy your conversational system app Assess its performance using human judgment Optimize its operating thresholds Retrain machine learning model with golden examples from humans 5. Lather, rinse, repeat
Deploy your conversational system app Assess its performance using human judgment Optimize its operating thresholds Retrain machine learning model with golden examples from humans 5. Lather, rinse, repeat
"Shut it down, shut it down!"
In AI, everyone is learning as they go along
They built teams.
developer domain expert sme content strategist ux designer
developer domain expert
Intents Entities Dialog builder #greeting Conversation starts Condition #greeting Watson responses (0) No responses added Edit . Order pizza flow and some … Order pizza flow and some … Condition Condition No conditions set No conditions set Watson responses 20 We got ya down for a $topping pizza. Anything else Watson responses 20 We got ya down for a $topping pizza. Introduction
@location . . #recommend @restaurant @cuisine
Empowering the creators
developer business domain expert data scientist
domain expert
"Ther e's w ater i n my room !" e k i l u o y d l "Wou e m o s u o y bring us to ice?"
They're closest to the content
moar content engineer data scientist domain expert business
How's your relationship to content?
moar content engineer data scientist domain expert business
"It's just content"
When you appropriate a domain or discipline that you don't know, you are unaware of your biases
"It's just content"
...what if you don't need AI?
Yay!
Inspiration for this talk
Would you put a baby in charge of customer support?
How about letting a toddler recommend medical treatment?
Or getting financial advice from a teenager?
It takes a village to raise a chatbot
Thank you! These slides live here: bit.ly/raising-chatbots An visual introduction to machine learning: r2d3.us A review of chatbot platforms bit.ly/chatbot-platforms An excellent guide to conversational design Conversational Design by Erika Hall (A Book Apart) Sparkle and Spin - A Book About Words, by Ann Rand and Paul Rand Twitter: @thedezzie Email: dezzie@automattic.com
Your job isn't going anywhere. You do not need to be a machine learning engineer to work with AI. Former IBM Watson designer Dezzie Garcia takes you through the journey of designing effective ways for people to be able to build their chatbots, from the assumption that engineering is at the center of the work. Maybe you'll be surprised, or maybe you won't–that's good! This talk is intended to get audience members to reflect on how their roles have evolved in the industry, and how this may be happening again.
for free. You
can too.