Contact Information

7154 N University Dr #95,
Tamarac, FL 33321, USA

We Are Available 24/ 7. Email Us.

As enterprises move beyond the pilot stage to scaling and operationalizing artificial intelligence, one tech giant is changing the way its AI operations are organized within the company. Meta (Facebook’s parent) announced in early June that it would decentralize AI at the company, distributing ownership of it into Meta’s product groups, according to CTO Andrew Bosworth.

“We believe that this will accelerate the adoption of important new technology across the company while allowing us to push the envelope,” Bosworth wrote in his post announcing the change.

The announcement signals a shakeup of how AI is organized at Meta, with the VP of AI Jerome Pesenti leaving the company and other changes such as the consolidation of several separate AI teams.

The changes at Meta beg the question for other forward-thinking enterprises across industries: ‘Is Meta’s AI reorg the example to follow? How should we think about structuring our own artificial intelligence research and operations?’

How Enterprises Structure Initial AI Practices

Often, enterprise organizations get their start with AI as an initiative driven by a single business unit. For instance, marketing organizations within enterprises have been using AI techniques for a long time now, says Gartner’s lead AI analyst Erick Brethenoux. Then, organizations may distribute their AI practice to business units or product groups, as Meta has just said it will do, with the goal of accelerating adoption across the business.

“That’s not new, right? We’ve seen it over and over again,” Brethenoux says. “People shift from centralized to decentralized to centralized to decentralized — and not just with AI, by the way. They’ve done that with all kinds of other capabilities and competencies within the enterprise.” HR is one example, he says.

A Better Approach: Hybrid

But Brethenoux was surprised to hear that Facebook was moving to a decentralized AI model going forward.

“They should be one of the most advanced, mature companies,” he says. “I was surprised to see that they are doing something that my clients have done before but have come away from.”

Instead, these enterprises that have tried and abandoned the approach taken by Meta — Brethenoux calls them his most mature clients — are operating under a model that’s a hybrid of centralized and decentralized AI.

How Hybrid AI Works

Here’s how he describes how they organize the practice. These enterprises typically start their AI practice under a particular business unit and then that is evolved to find a way to syndicate the AI knowledge to a centralized location (physical or virtual), often called a Center of Excellence, an AI Lab, or a Data Science Lab. But instead of just leaving this AI Lab to operate on its own, these mature companies also establish an executive committee — a steering committee — that has real authority to decide on the projects for this AI Lab.

This AI Lab then reports into a corporate function, not a business unit. Why? Brethenoux says this reporting structure establishes two important things. The first is neutrality among different business units. The second is that it ensures that the projects that are chosen are in alignment with the company’s overall strategy.

That might sound just like a centralized approach. But these companies don’t stop there, Brethenoux says. Next, they take the AI experts from the AI Lab and rotate them through different business units. These experts spend 6 to 12 months in business unit one, then move to business unit two and spend the same amount of time there, and so on. After a full tour, they go back to the AI lab for 3 to 6 months before they return to the rotation again.

“They learn from the field as the AI expert is confronted with the reality of each business unit to understand what is really happening on the ground,” he says. What’s more, “They propagate the knowledge.” The rotating AI experts take the solved problems of one business unit to other business units that may be experiencing similar issues.

“When [organizations] have that model in place where they centralize the knowledge somewhere but have the people rotating across the business functions, they realize that it boosts retention,” Brethenoux says. “Because AI experts are exposed to and are solving a lot of different problems, and the knowledge sharing is intensive. That helps in the retention of people who are normally curious, and AI experts are normally curious people.”

This is the approach that Brethenoux now recommends to his clients, large and small, who are looking for the optimal setup of AI within an organization. It may look a little different depending on the industry you are in — telecom will be different than automotive, and automotive will be different from pharmaceutical. But the skeleton of the setup is the same across all industries, he says.

The multiple crises of the pandemic and all the after-effects of the pandemic — supply chain disruptions, remote work, and more — have accelerated organizations’ move to this kind of setup for the artificial intelligence practices, Brethenoux says, just like other technology initiative timelines have been accelerated.

For IT organizations looking to maximize the value of their AI programs across the organization, the hybrid approach may be the answer.

“People are starting to focus on the outcome of what AI can produce and less on the technology itself,” Brethenoux says.

What to Read Next:

Priorities of Highly Successful Chief Data Officers

AI Set to Disrupt Traditional Data Management Practices

Why Your Company Should Evolve From Data-Driven to Decision-Driven



Source link

Share:

administrator

Leave a Reply

Your email address will not be published.