Latest NewsOct 11: Continuous news-feed now available
Oct 2: New AI rants with Umair Khan, Andi Mann, Andrew Hillier, and Adam Carmi Sep 26: Report Launch Date: Tue, Oct 2, 2018: 6a.m. Sep 25: Machine Learning (not only) for Kids mini site goes live. Sep 15: EMA Speaking at DevOps Enterprise Summit (DOES) in Las Vegas: AI and DevOps Sep 11: Security Risks of Enterprise AI Sep 10: 6 Best Practices to Prevent Bias in Artificial Intelligence Sept 7: How to leverage AI for DevOps Sept 6: 6 Best Practices to Recognize and Prevent AI Bias Sept 3: EMA Speaking at DOES 2018: AI in DevOps Aug 21: How AI "thinks" versus how the human brain works (example) Aug 21: Added report launch timeline. Aug 17: Comprehensive Q&A in terms of report content and vendor selection. Aug 16: Naming the EMA Top 3 categories. Aug 16: FAQ page [BLOG POST]: The 10 Most Common Misconceptions in AI/ML in 2018 (August 15) EMA has started to schedule press briefings on the EMA Top 3 AI/ML decision guide for enterprise (August 14). First demo videos recorded with EMA Top 3 winners. Exciting! (August 13) |
AI Product Showcases and AI Rants
To prevent another AI Winter, we need to emphasize the tremendous business impact of what AI/ML can do today. This image shows IBM Visual Insights correctly checking an industrial part for quality issues, without ever having seen those exact defects before. This shows that today’s top-down approach of creating narrow inference models through providing large numbers of pre-labeled examples can exceed human performance in many cases.
|
AI/ML Breakthroughs
1993: The First Convolutional Neural Network by Yann LeCun
Yann LeCun's one minute demo of the first convolutional neural network (CNN) shows the vast potential of CNNs. Rule-based programming could never come close to recognizing everyone's individual handwriting under difficult conditions.
2003: The first CNN-based face detection software
In 2003, 10 years after LeCun's initial CNN breakthrough in handwriting recognition, his team successfully used CNN for facial recognition.
EMA AI Publications
How AI really Thinks
Don't be discourage by this piece, as our thoughts are purely directed toward setting realistic expectations. Knowing what you can and cannot expect from AI/ML is key to successfully leveraging these technologies. Read Article AI Project Cheat Sheet
Questions to ask software vendors when talking about AI. This will eventually be merged into the Top 3 report, but currently stands alone to provide basic guidance through the AI marketing jungle. Read Article |