Gartner Newsroom

“The reason why we emphasize on AI operationalization is because AI delivers business value only where it is productionalized.”

Trap #5: The new tool we buy will help us scale our AI initiatives. 

• Instead, use the right mix of hybrid/multicloud/edge/IoT and create recipe books.

Trap #4: Waiting to hire the perfect “unicorn” to operationalize AI projects.

• Instead, reskill existing teams with AI operationalization systems engineering skills.

Trap #3: What works in AI pilots will scale in production.

• Instead, complement your data science teams with diverse stakeholders from the I&O or EA teams, for example, so that they operate as a team. 

Trap #2: Taking a “set it and forget it” approach to AI.

• Instead, track changes in data sources to identify data drift.

Trap #1: “Ideating about AI” is more useful than actually “implementing AI”.

• Start with small AI projects that are driven by business use cases and requirements.

“We need to think about optimizing computing to be location-agnostic. Work is no longer a place I go to, work is a thing I do.”

“For today’s data center, minimize movement, automate as much as possible, minimize risk and maximize value.” 

“Start looking at off-premises as a way to move workloads, not because you want to get away from your data center, but because you can solve specific business problems and enhance business, and even customer experience, by doing that.”

“If you need to control the data, and you need to control the location, you can do it with a distributed cloud environment.”

“The way that we’re looking at I&O today is changing dramatically over time, and we need to change from managing infrastructure to managing software platforms.”

The promise of 5G is speed, massive IoT and low latency communications. The current version of 5G offers 1.2 gigabits per second, supports up to 1 million IoT sensors in a square kilometer, and delivers low single-digit mini second latency.

“What makes 5G more complicated than previous versions is that it has eight technology variations that operators can select and deliver. It also comes in multiple phases and releases.”

“Businesses will use 5G for high bandwidth IoT applications predominantly around video surveillance and analytics solutions. In the longer term, organizations will use the private 5G network to control drones, robots and manufacturing equipment in real time.” 

“Today if organizations have a pressing need, go with LTE, don’t wait. Near term, focus on 5G opportunities that need high bandwidth. Longer term, look for opportunities to exploit mobile edge compute.” 

“Don’t expect 5G to solve immediate enterprise problems, but long-term adoption is inevitable.”