Doesn't do to understimate the conservative attititude that exists in many of the potential application areas for "AI at the edge" solutions. Not sure the lengthy proof of concept, trial period approach is going to keep up with the pace of technical change as AI disrupts so many markets.
I agree. I think questions about how organisations can keep up - and what the right strategies are - at a time of profound and rapid tech disruption, are very much worth addressing. This is a particular issue for organisations that start with large amounts of tech debt. See my article in Heywood Quarterly https://heywoodquarterly.com/why-ai-exposes-our-digital-shortcomings/
I think you have it covered! During the 2010s I ran a company that I had co-founded delivering real-time AI analytics to airports. We deployed a cloud-edge solution. Key challenges were resilience and security at the edge (what to do when something pops in an airport ceiling or someone tampers with it), consistency of assumptions and the complexity of getting security accreditation for such a distributed architecture. It takes an end-to-end complex systems mindset.
Doesn't do to understimate the conservative attititude that exists in many of the potential application areas for "AI at the edge" solutions. Not sure the lengthy proof of concept, trial period approach is going to keep up with the pace of technical change as AI disrupts so many markets.
I agree. I think questions about how organisations can keep up - and what the right strategies are - at a time of profound and rapid tech disruption, are very much worth addressing. This is a particular issue for organisations that start with large amounts of tech debt. See my article in Heywood Quarterly https://heywoodquarterly.com/why-ai-exposes-our-digital-shortcomings/
I think you have it covered! During the 2010s I ran a company that I had co-founded delivering real-time AI analytics to airports. We deployed a cloud-edge solution. Key challenges were resilience and security at the edge (what to do when something pops in an airport ceiling or someone tampers with it), consistency of assumptions and the complexity of getting security accreditation for such a distributed architecture. It takes an end-to-end complex systems mindset.
The security accreditation - assurance - issue is particularly fascinating. Thanks for this insight.