I, Human
summary
Covers concrete demos and failures: Microsoft Seeing AI is shown working—phone camera + cloud gives useful scene and person descriptions—while Watson fed Urban Dictionary learns to swear and Tay becomes toxic within hours, both deleted. Mentions Amazon Rekognition misidentifying US Congress as criminals and Google Translate gender bias; explains image and speech recognition, cloud infrastructure and compute needs, and simple brain-percentage stats (vision ~30% hearing ~3% speaking ~50%) so the claims, tools and real-world success or harm are clear.
AI-written summary · about
chapters (12)
- 0:00Human extraordinary nature
- 3:55Human origins and modernity
- 10:26Artificial intelligence concept
- 13:00Moravec's paradox in AI
- 16:14Seeing AI app revolution
- 18:24Face recognition overcoming limits
- 20:20Speech recognition evolution
- 24:35Voice assistants transforming life
- 26:44Sign language translation
- 28:09AI training unintended consequences
- 31:47AI amplifying human potential
- 33:41Closing call to human nature
related talks
38mHow Artificial Intelligence has changed your life
50mCognitive Services Extravaganza!
56mArtificial Intelligence? - more like Artificial Stupidity!
01h 03mIt's All About AI
42mIs your AI plotting to kill you? Why AI and Machine Learning is dangerous
44mThe age of applied AI
27mLive and Machine Learn
19mAI: The current state of play