GenAI FOMO? Careful, You May Get Burnt
- By Winston Thomas
- July 24, 2023
Getting into the generative AI hype? Being pressured to adopt and adapt GenAI products because you may be outplayed and outcompeted?
Well, in theory, that’s all true. The truth can be different.
Simply put, we are in the midst of an AI hype cycle, thanks to the immense interest generated by generative AI.
“But we’ve been here before; now, everyone’s talking about AI,” says Ross Solomon, head of product management for customer communications at Broadridge.
The U.S.-based fintech player builds industry-specific technology, solutions, data, and intelligence to drive their clients’ business transformation. Recently, it launched a generative AI product called Bond GPT.
Based on his experience, Solomon offers a word of caution. “I think going into the hype phase blindly without proper due diligence will see you throw away money,” he adds. And it is money that companies can’t afford to throw away in today’s uncertain times.
“So, I think it's better to take a step back, realize that the hype is warranted because it's real, but plan for it like you would any other business strategy,” he advises.
The elusive hype-busting plan
Before we get to a plan, it is good to understand why we need a generative AI. The trouble is that many companies point to their neighboring competitor as the reason.
Part of this blurring of the scope and goal of a generative AI project is that many more stakeholders are involved in AI; it has stopped becoming an IT science project overnight.
“A lot of other stakeholders are now involved in the hype. Yes, technologists have been looking at [AI] for over a decade at this point. But now business leaders are able to see the kind of disruption [they can do] and are getting involved,” says Solomon.
In some way, this is a good thing. As more business leaders want to build on the promises of generative AI, they help to create momentum, build alignment across the organization and break these projects out of the PoC cages.
Solomon calls this “an inflection point” for AL. But having one is not good enough if the projects fail to deliver. It is why Solomon calls for a more thought-out plan involving due diligence.
“The question everyone should be asking is whether [generative AI] helps anyone. Did they empathize with their client base, or are they solving a customer pain point? Did they use journey maps? Is this really solving a problem or creating efficiencies?” he questions.
Solomon used his company’s experience rolling out Bond GPT as an example. They used their vast trove of bond data to create a product but only launched it when they found the right use case.
“I think there are opportunities to find those good use cases,” says Solomon, but warned companies not to launch a generative AI product for the sake of it.
It’s easy to see generative AI products as the next best thing. Yet, for all its glitz and fanfare, it is sorely dependent on a single asset that companies still have trouble with: data.
For a decade, companies have struggled to get their internal proprietary data (the data that will give you a competitive differentiator) ship-shape. This is where Solomon feels that generative AI hype is widening the gap between expectations and actual results.
He notes that generative AI will fall flat or run into trouble if you “don’t have a successful data strategy, data governance and data infrastructure.”
“If your data can’t interact with each other, you’re not going to have good AI because you’re not going to mine it. And that’s a really important gap,” Solomon explains.
Play culture club again
Another essential thing to determine is whether your company has the right culture to build and use generative AI products.
“You need a pretty progressive culture to be successful at AI. Why? Because this is a very fast-moving technology. You need a culture that's open minded and open to change,” says Solomon.
He thinks AI will disrupt company cultures. And companies need to be ready for this kind of disruption if they want to succeed with generative AI. “So, if you have one of those rigid, top-down, hierarchical cultures, you're not going to be successful.”
Generative AI can also challenge established hierarchies that use information as power. With the right prompts, anyone in the organization can find usable information.
Whether the organization will allow them to act on this information in new ways or is more concerned with restricting them is yet to be played out on a grand scale. “AI is going to push organizations. And, I would make this for any new emerging technology; it doesn't just have to be AI, but anything that's moving fast.”
Leadership also needs an upgrade. With information at their fingertips, employees will make decisions quickly—but it needs a leader who can empower them.
The challenge, as Solomon points out, is many of these leaders have only immersed themselves in AI in recent years. So, how an organization educates its entire employee base and its leadership, enables decision making and becomes agile to new generative AI advancements will be telling in the near future.
That’s because governments and regulators are beginning to wake up to the realities of AI and introduce new laws and legislation. Companies must be aware of the potential legal challenges and biases hidden within their models so they aren’t blindsided. For this to happen, you need empowerment and lots of education.
Get your data priorities right
So, how does one start a generative AI project in a challenging market environment? After all, it can be time-consuming and resource-intensive, even though the benefits can be a valuable lead in the market.
Solomon, using his Bond GPT experience, advises companies to prioritize. And he feels that it is no different from the principles governing IT project prioritization.
“First, understand what your business objectives are and where your differentiators are,” said Solomon. After that, you discover how generative AI can improve or enhance your differentiators.
Solomon points to customer pain points as a starting point. “If your clients are struggling, can you use AI to address that need? So, it’s no different than product or feature prioritization.”
You can then do a feasibility assessment to understand where to begin and have the most impact. This is when you get realistic about your data quality and sources and whether your data integration is up to spec.
The only difference between an AI project and a typical IT project is the momentum. Things go really fast with AI, and “you need to make sure you have all your ducks in a row” from the start, says Solomon.
So what happens if your data strategy is only now catching up with your AI ambitions? Solomon still thinks companies can set up their AI journeys in parallel with their enterprise data strategy.
“But, if you're behind on your data, you're going to be behind on your AI project. I encourage any firm right now, if you’re looking to make a dent in AI, first make sure your data is set up for success.”
Winston Thomas is the editor-in-chief of CDOTrends and DigitalWorkforceTrends. He’s a singularity believer, a blockchain enthusiast, and believes we already live in a metaverse. You can reach him at [email protected].
Image credit: iStockphoto/Liudmila Chernetska
Winston Thomas
Winston Thomas is the editor-in-chief of CDOTrends. He likes to piece together the weird and wondering tech puzzle for readers and identify groundbreaking business models led by tech while waiting for the singularity.