|
Briefing
|
- Crowdsourced AI Testing – App testing platform Applause announced new solution on November 6, 2019, that leverages vetted human testers to train artificial intelligence (AI) algorithms and test if they are accurate, more human in response and unbiased
- Existing AI Problems – AI voice and facial recognition systems tend to be more accurate on white-skinned males than African-American females, have problems recognizing accents, as well as require more training to understand slang language
- Rigorous Testing – Algorithms are tested across devices, locations and circumstances to detect possible issues and provide user feedback in real-time
- Representative Community – Community of testers originate from various countries, ages, genders, races, cultures, political affiliations, ideologies, socioeconomic and education levels
- AI Solutions – Provides training data sets and crowdsourced testing for five AI technologies, including voice, biometrics and image recognition systems, as well as text or character identification and chatbots
- Benefits – Tests algorithms at all development stages to ensure AI provides quality experience to human users and produces products people will actually want and use
|
