Keeping your enterprise AI expertise up to speed

AI is such a significant force, you can’t afford not to stay on top of the ever-changing developments

Keeping your enterprise AI expertise up to speed
Thinkstock

Artificial intelligence (AI) could not be a more strategic enterprise technology. As we move into the ’20s, the most disruptive business applications will be those that incorporate machine learning, deep learning, and other forms of AI.

AI has become the brain driving cloud-native enterprise applications. Developers everywhere are embedding AI microservices to imbue cloud applications with data-driven machine learning intelligence. Increasingly, there is no substitute for the sophisticated AI that performs high-speed inferencing on sensor-sourced data and on data acquired from applications, clouds, hub gateways, and other online resources.

Staying abreast of AI trends, technologies, and applications is fundamental to success in modern business, even if you’re not a data scientist or machine learning specialist. Companies that innovate with AI will dominate their industries for decades to come.

Build a top-notch enterprise AI competency

Considering how ubiquitous and dynamic the AI industry has become, keeping your enterprise AI expertise up to snuff can be a challenge. If nothing else, you should make sure that your company implements a far-reaching program of AI skills, processes, tools, platforms, and methodologies that covers the following major points:

  • Teams: Assemble dedicated teams of data scientists and other developers to build, train, deploy, and manage AI applications as a standardized operational process across all business functions.
  • Platform: Deploy an integrated, open, and trusted platform for data science, machine learning, data engineering, and application building across the multicloud.
  • Data: Combine hybrid data from on-premises platforms and public clouds when building, training, deploying, and managing machine learning, deep learning, and other AI models.
  • Workloads: Deploy a fast, scalable, hybrid data environment to manage both data at rest and data in motion for myriad AI workloads.
  • Automation: Adopt cloud-based AI devops tools that incorporate popular modeling frameworks, automate model management and hyperparameter tuning, accelerate AI workloads across distributed GPUs and other compute nodes, and enable developers to access pretrained models from libraries across public clouds, private clouds, and on-premises systems.
  • Deployment: Distribute and scale AI inferencing, training, modeling, and data preparation workloads across public clouds, private clouds, and on-premises systems.
  • Tools: Adopt robust tools for data integration, security, governance, lifecycle management, devops, and orchestration across all AI initiatives, projects, applications, and workloads.

Keep up with enterprise AI trends

Wrapping your head around AI’s complexities is a never-ending task. As an industry analyst, I do this for a living, and even I often feel like I’m simply treading water.

The AI industry is evolving so fast that all of us—enterprises, teams, individuals—need a mental map of trends in both business and technical topics. What follows is a cheat sheet of top AI trends that I’ve compiled from the last two years’ worth of my own year-end predictions, all of which—near as I can tell—are still in full force:

  • Governance: AI regulations are coming fast in most modern countries and every industry. Enterprise chief legal officers are mandating end-to-end AI transparency. AI risk-mitigation controls are becoming standard patterns available in data science pipeline tools. AI data science team workbenches are enabling downstream reproducibility. AI deepfakery is working its magic—benign and otherwise—more deeply into our lives.
  • Acceleration: GPUs are dominating AI acceleration. GPUs are expanding their footprint in immersive AI applications. AI systems-on-chip are dominating the hardware-accelerator wares.
  • Tools: AI development frameworks are becoming interchangeable within an open industry ecosystem. More data scientists are buying certified high-performance AI algorithms, trained models, and training data from online marketplaces. AI modeling frameworks are converging on a two-horse race.
  • Platforms: SaaS-based AI solutions are reducing enterprise demand for data scientists. More labeling of AI training data is being automated through on-demand cloud services. Kubernetes-orchestrated containers are becoming integral to the AI pipeline. Reinforcement learning is becoming a mainstream AI approach. Client-side training is moving toward the AI mainstream. Blockchain is feeling its way into the AI ecosystem.
  • Performance: Dominant AI development frameworks are being re-engineered for superior cloud-to-edge performance. AI benchmarking frameworks are crystallizing and gaining enterprise adoption. Industry-standard AI benchmarks are becoming a competitive battlefront.
  • Processes: AI is automating AI developers’ core modeling functions. Automated end-to-end AI devops pipelines are becoming standard practice. Enterprise AI is shifting toward continual real-world experimentation. AI is becoming an industrialized operational business function. AI is driving closed-loop IT operations management.

For enterprise professionals, your strategy for staying prepared and abreast of all of this should include participation in AI industry and professional confabs taking place this year, such as the Nvidia GPU Technology Conference. Also, going back to school to earn an AI certification would be a good idea. And it couldn’t hurt to engage with friendly AI industry analysts for an expert’s reality check on all this.

Copyright © 2020 IDG Communications, Inc.