Nothing illustrates the old adage “timing is everything” better than the tricky question of exactly when to deploy a hot new enterprise technology.
Jump in too early? You may end up with tech that’s a bit too unproven, costs a tad too much and works with far too little of your existing IT.
Wait too long? You run the biggest risk of all: missing the boat entirely.
Read on to discover which key stage of the AI journey your organization is in — and Herschel’s advice on what your next move should be.
According to Gartner’s recent survey of 3,138 CIOs, only four per cent have actively deployed and invested in AI at this point. What’s holding so many organizations back?
“The biggest challenge is internal,” Herschel said. “It’s not a lack of available [AI] technology. It’s a clear lack of understanding about what to do with that technology within the organization.”
If your enterprise is, indeed, among the four per cent that have already dived into AI, what should you do next? Herschel suggested taking stock by asking yourself how to measure and define AI success.
He also pointed out, however, the potential pitfalls of deploying AI way too early.
“Often what you see is that the technology stuff is further advanced than the organizational stuff: the ethics, the governance. So we need to make sure those are in place before we over-invest too much in the technology.”
Twenty-one per cent of surveyed organizations are actively experimenting with AI or in short-term planning for deployment. Herschel noted that much of this “early tinkering” with AI is “just traditional analytics.”
This approach is fine, he said, since it avoids “a classic issue with any new technology” of jumping in because everyone else is doing it, but “without understanding what skills will be required.”
By actively dabbling in AI, organizations at this stage are considering and planning for problems before they arise during active deployment.
“So it means really starting to think about things like, do you truly understand the situation? How do you begin to set [AI] expectations appropriately? It’s crucial at this stage.”
It also gives your organization time to develop a loose framework for AI governance and ethics issues Herschel alluded to earlier.
The bulk of surveyed enterprises (60 per cent) are either engaged in medium- to long-term AI planning or say it’s on their radar, but they have no specific action planned yet. And in Herschel’s mind, “that’s a perfectly acceptable place to be right now.”
Why is he okay with this wait-and-see approach?
“Even if you assume that, at some point, AI will be relevant to every organization — which is a legitimate discussion — it’s not ready for every organization right now,” Herschel explained.
“For you to have thought about AI and then to have made a conscious decision — either to invest or not to invest, to postpone that investment to wait for a higher level of technology maturity, to wait for a higher level of engagement from some of the partners in your value chain or from your customers — I think that’s perfectly acceptable.”
What’s the next step to take if you’re still in long-term planning for AI?
“Start asking what are the pitfalls and lessons learned from the four per cent who already deployed AI,” Herschel advised.
Which brings us to the 14 per cent of organizations who say they have “no interest in AI” and no plans for it whatsoever. Herschel had some pretty strong words for this group.
“Saying you’re not interested in AI [in 2018] is going to be like saying 20 years ago that you weren’t interested in the Internet,” he said.
If your enterprise is completely MIA when it comes to AI, what can you possibly do about it? Get out there and lobby the people who have power to change the tide, said Herschel.
Your CIO or chief data officer is the most obvious lobbying target, he said. But after them, try your CEO. Although Gartner’s survey shows CEOs initiate only 10 per cent of AI efforts, they wield approval power over 21 per cent of all AI funding.
Don’t forget to press for AI with line of business, though.
“I think this is the growth opportunity for AI in many organizations,” Herschel said. He believes that as AI starts to mature, business unit heads in marketing, supply chain, security, compliance and operations will be interested in pursuing AI to solve particular problems.
Herschel’s final tip, regardless of where your organization sits on the AI continuum: “AI is an immature technology. But don’t let that stop you from exploring its technology.”