• Historical outcome of Moore's Law is actually accelerating
• Relative cost of processing is falling like a rock
• Network & storage are becoming expensive bottlenecks
• Networking & storage will become distributed & decentralized 
• Processing is getting pushed to the edge where the data is created
• Most edge data (99%) won't move back to the cloud - example Tesla - most of the data won't go back to the cloud
• AI will move from modelling to inference - "Inference is the deployment of the model, taking real-time data from sensors, processing data locally, applying the training that has been developed in the cloud and making micro-adjustments in real time."
• Most AI will be bought and applied, not built
• Data becomes the new development kit
• Implications for hardware, software and services

Image source: Elements