Artificial Intelligence: The Future of Healthcare
summary
Concrete examples and concrete fixes are here: an Amazon resume filter that learned to downgrade women, a Roomba turned off its camera because of a poorly specified reward, and traffic sign attacks that flip stop into speed limit. Tool names appear—TensorFlow and Keras plus scikit learn Ray and Unity—and claims include need for roughly 100k examples for good neural net performance, HIPAA/cloud limits, human‑in‑the‑loop fixes, and API integrations with transcription and EMR platforms.
AI-written summary · about
chapters (16)
- 0:00AI and its types
- 3:30Ethical concerns in AI
- 5:19AI adoption barriers
- 7:03Reinforcement learning risks
- 9:58Computer vision challenges
- 12:49Cybersecurity and job security in AI
- 14:17AI data collection state
- 15:20Why computers don’t think
- 17:28Reducing AI risks
- 20:18Human-in-the-loop benefits
- 23:31AI augmenting healthcare
- 25:37Clean data and AI interfaces
- 27:43Building AI systems
- 28:47Over-reliance on AI problems
- 31:42AI in healthcare delivery
- 35:11Q&A
related talks
58mHealthcare with AI of the Future
36mArtificial Intelligence: The Future of Software
22mLeveraging AI/ML for Transformative Software Solutions in Healthcare
47mThe Future of Healthcare and AI
56mArtificial Intelligence? - more like Artificial Stupidity!
26mArtificial Intelligence: The Recent Past
32mMen and Machine: An Artifical Love Affair?
38mHow Artificial Intelligence has changed your life