Human-First AI: Because AI Needs The Human Eye
summary
Claims an open-source Human-First AI framework that jumpstarts industrial ML where data is scarce by combining rule/physics-based data generators, an execution graph and a reusable human-first workbench that wraps TensorFlow/PyTorch models. Uses Ray to parallelize synthetic data generation and model tasks. Decomposes problems so a two-stage detector cuts false positive rate by two orders of magnitude, enables faster parallel development, and reports provenance/explainability for deployed industrial systems.
AI-written summary · about
chapters (7)
- 0:00Industrial AI challenges
- 2:07Human-First AI concept
- 3:11Three failure patterns
- 4:13Automotive cyber security demo
- 6:41Profit problem solution
- 8:29Trustworthy AI features
- 10:14Long-term human role in AI
related talks
22mThe Imperative for Open Source AI - The Path Forward to Take Down the AI Oligarchs
46mCan you trust your AI?
19mDeveloper Centric AI
54mThe Human Potential and AI
19mAI in the Field: Humans at the Heart of Value
08mAgents and Robots: The Human Effort Putting AI Into Production
16mUsing AI and Computer Technologies to Enhance IT Infrastructure
39mThe Search for Transparency and Accountability in the Age of...