2023-06-30 13:30:00
In episode 21 of The Cutting Edge Podcast, Leonard Lee discusses generative hybrid artificial intelligence (AI). He defines hybrid AI and explains why more organizations are spreading AI workloads across multiple environments.
This episode is sponsored by “Selling to the New Executive Buying Committee,” an Acceleration Economy Course designed to help vendors, partners, and buyers understand the shifting sands of how mid-market and enterprise CXOs are making purchase decisions to modernize technology.
Highlights
00:38 — Hybrid AI is similar to hybrid cloud, as workloads are spread across the cloud, on premises, at the edge, and on devices.
01:50 — Factors that have shifted where AI workloads and applications are distributed include edge AI technologies such as TinyML, edge AI processors, and engines that support machine learning (ML) workloads; edge cloud technologies that enable AI model training outside hyperscale data centers; and emerging distributed AI technology architectures for federated and collaborative learning and inference across the edge.
02:39 — Hybrid AI came into the mainstream through smartphone technologies like Apple‘s A-series and Qualcomm’s Snapdragon processors that put AI capabilities into people’s hands. Another factor was Intel pushing computer vision for industrial and enterprise applications running on their CPUs for on-premises and on-device deployments.
03:27 — Security, privacy, confidentiality, data sovereignty, cost, and sustainability are the main reasons AI systems and workloads are making their way into local private edge environments and on devices.
04:52 — As innovators find useful applications for edge generative AI, generative AI applications will become more specific to an organization’s purpose and data. Some of these applications will not require massive computing power, rather, they will be trained locally, privately, securely, and confidentially under enterprise or consumer control.
05:37 — The cost of querying a generative AI application is a concern. For example, a generative AI web search can cost 10 times that of a search engine. This will force AI systems and workloads to be deployed where they can run most economically and sustainably: hybrid and distributed.
06:56 — There are still many important challenges to overcome before we get to great AI. C-suite leaders should abide by Apple CEO Tim Cook’s comment regarding generative AI strategy: “It’s important to be deliberate and thoughtful.”
1688145124
#Hybrid #Dominant #Workload #Model #Support #Apps