#spotifydesignforml search results

Big thanks to @spotifydesign @mluxsf for having me! I loved sharing out practical tips from the #PAIRguidebook and learning from the #mlux community in turn 🙌 Check out more UX of ML pointers from the #SpotifyDesignForML event 👉 spotify.design/articles/2019-…

Last month we hosted @spotifydesign, @facebookdesign and @GoogleDesign at our NYC #SpotifyDesignforML event. If you would like to find out how it went down, check out the event round up here to give you all the tea ☕️ spotify.design/articles/2019-…

How do you calibrate trust with users about the accuracy of ML predictions? Some design patterns suggestions here #spotifydesignforML

Thanks guys @spotifydesign, @maheen_sohail, @dqpdang! It was interesting 💚🖤#SpotifyDesignforML #mluxnyc

Lovely evening at ‘Design for ML’. Thanks @spotifydesign #mluxNYC #SpotifyDesignforML

Facebook’s @maheen_sohail talks about how Portal videochatting screens use classifiers to figure out what’s worth keeping in frame, but cinematographic technique to do the framing. #mluxnyc #spotifydesignforml

On getting more realistic avatars for getting more social context #mluxnyc #spotifydesignforml by @maheen_sohail

@maheen_sohail sharing how to help with social context using ML and VR, like using NLP and your voice! #mluxnyc #spotifydesignforML

Up next: @maheen_sohail sharing @facebookai and their approach for designing for the fufure! #spotifydesignforML #MLUXnyc

Actionable pointers!! 👇🏿👇🏽👇🏻👇🏼👇🏾 on 3 awesome chapters User needs, Mental models, Explainability & Trust-- there's also 3 more chapters!! #PAIRguidebook #mluxNYC #spotifydesignforML pair.withgoogle.com thank you @dqpdang 💖🤖✨@spotifydesign

What are some design patterns to help with explainability and trust? Glad you asked! Check out the #pairguidebook for more too! Allow for user control and surfacing other options if ml is wrong (which it will be!) @dqpdang #spotifydesignforml #mluxnyc

A guided tour from @dqpdang of the Mental Models and Explainability + Trust sections of @Google’s #PAIRguidebook (and a disappointing Spongebob popsicle). #spotifydesignforML #mluxnyc

Trust and explainability are really important in building ML products: consider what are the risks if your user over or under trusts the system? @dqpdang from @GoogleDesign at @spotifydesign #spotifydesignforML #mluxnyc

Does ML not always meet your expectations? You're not alone! Be are when you design your ML and AI powered systems you meet users mental models! #mluxnyc #spotifydesignforML @dqpdang

@dqpdang sharing the chapters of the #PAIRguidebook (see above 👆🏻👆🏽) #spotifydesignforML #mluxnyc

Up next: @dqpdang from @GoogleDesign sharing how Google approaches #humancenteredML #SpotifydesignforML #mluxNYC

The data curation team at @spotifydesign process: 1. Define structure 2. Reduce friction 3. Track progress So important when annotators are trying to help create data for ML teams to improve the experience #mluxnyc #spotifydesignforML

Up next - #datacuration team!! So important and integral for new ml features - including using human in the loop methods! #mluxnyc #spotifydesignforML

#Recsplanations! The system adapts and gives you explanations to recommendations ! Love it! Helps give reasons and context to recs @spotifydesign #mluxnyc #spotifydesignforml

Diane @spotifydesig sharing #UXwriting best practices keeping #ml explainable: 1. Set expectations 2. Don't over promise 3. Evolve with the user build trust with the user & help them understand why they might be recommended some things, updating the interface #spotifydesignforml

Thanks guys @spotifydesign, @maheen_sohail, @dqpdang! It was interesting 💚🖤#SpotifyDesignforML #mluxnyc

Lovely evening at ‘Design for ML’. Thanks @spotifydesign #mluxNYC #SpotifyDesignforML

Facebook’s @maheen_sohail talks about how Portal videochatting screens use classifiers to figure out what’s worth keeping in frame, but cinematographic technique to do the framing. #mluxnyc #spotifydesignforml

Mark from @spotifydesign sharing how they refine to make sure they are asking the right questions! #spotifydesignforml #mluxnyc

On getting more realistic avatars for getting more social context #mluxnyc #spotifydesignforml by @maheen_sohail

Top 3 questions from @spotifydesign ! 1. Ask the right Q's 2. Identify friction and automate it away 3. Go manual before going magical ✨🤖 #mluxnyc #spotifydesignforML

@maheen_sohail sharing how to help with social context using ML and VR, like using NLP and your voice! #mluxnyc #spotifydesignforML

Matt from the personalization team at @spotifydesign sharing how they perceive working with ml #spotifydesignforml #mluxnyc

@dqpdang sharing the chapters of the #PAIRguidebook (see above 👆🏻👆🏽) #spotifydesignforML #mluxnyc

Does ML not always meet your expectations? You're not alone! Be are when you design your ML and AI powered systems you meet users mental models! #mluxnyc #spotifydesignforML @dqpdang

Up next: @maheen_sohail sharing @facebookai and their approach for designing for the fufure! #spotifydesignforML #MLUXnyc

Up next - #datacuration team!! So important and integral for new ml features - including using human in the loop methods! #mluxnyc #spotifydesignforML

@spotifydesign 's #discoverweekly is a great example of how ML reduces friction and brings joy to the end users #mluxnyc #spotifydesignforml

#Recsplanations! The system adapts and gives you explanations to recommendations ! Love it! Helps give reasons and context to recs @spotifydesign #mluxnyc #spotifydesignforml

James sharing some killer examples of how ML engineering tackles problems in the #mlux space #mluxnyc #spotifydesignforml

The data curation team at @spotifydesign process: 1. Define structure 2. Reduce friction 3. Track progress So important when annotators are trying to help create data for ML teams to improve the experience #mluxnyc #spotifydesignforML

Up next: @dqpdang from @GoogleDesign sharing how Google approaches #humancenteredML #SpotifydesignforML #mluxNYC

Diane @spotifydesig sharing #UXwriting best practices keeping #ml explainable: 1. Set expectations 2. Don't over promise 3. Evolve with the user build trust with the user & help them understand why they might be recommended some things, updating the interface #spotifydesignforml

Actionable pointers!! 👇🏿👇🏽👇🏻👇🏼👇🏾 on 3 awesome chapters User needs, Mental models, Explainability & Trust-- there's also 3 more chapters!! #PAIRguidebook #mluxNYC #spotifydesignforML pair.withgoogle.com thank you @dqpdang 💖🤖✨@spotifydesign

"if you can't get to a good results with just heuristics, you probably won't get there with #ml" how @spotifydesign prototypes ML experiences and proving the user benefit as quickly as possible - it is important to go manual before you go magical! #mluxnyc #spotifydesignforml