“There is no AI without data, there is no AI without unstructured data, and there is no AI without unstructured data at scale,” he said. Chet Kapoor President and CEO of data management company DataStax.
Kapoor was kicking off a 2024 roundtable on “new data pipelines” in the context of modern AI applications, where they were also vanessa Larco partner at the venture capital firm NEA; and George Fraser CEO of data integration platform Fivetran. While multiple topics were covered, such as the importance of data quality and the role of real-time data in generative AI, one of the big takeaways was the importance of prioritizing product-market fit over scale in what are really still the early days of AI. The advice for companies looking to jump into the fast-paced world of generative AI is simple: don’t be too ambitious at the beginning and focus on practical, incremental progress. The reason? We’re really still figuring out the full scheme of things going forward.
“The most important thing for generative AI is that it all comes down to people,” Kapoor said. “The SWAT teams that are actually going out and building the first projects aren’t reading a manual, they’re writing the manual on how to build generative AI applications.”
While it’s true that data and AI go hand in hand, it’s easy to feel overwhelmed by the sheer amount of data a company may have, some of it possibly sensitive and subject to strict protections, and perhaps even stored in countless locations. Larco, who works with and sits on the board of numerous startups across the B2C and B2B spectrum, suggested a simple yet pragmatic approach to unlocking true value in these early days.
“Work backwards to achieve what you’re trying to achieve: What are you trying to solve and what is the data you need?” Larco said. “Find that data, wherever it is, and then use it for this purpose.”
This is in contrast to trying to spread generative AI across the enterprise from the start, throwing all your data at the large language model (LLM) and hoping it eventually produces the right thing. That, Larco said, will likely create an expensive and inaccurate mess. “You have to start small,” he said. “What we’re seeing is companies starting small, with internal applications, with very specific goals, and then finding the data that matches what they’re trying to accomplish.”
Fraser, who has led “data movement” platform Fivetran since its inception 12 years ago, racking up big-name clients like OpenAI and Salesforce along the way, suggested that companies should focus narrowly on the real problems they face today.
“Just solve the problems you have today; that’s the mantra,” Fraser said. “The costs of innovation are always the 99% of things that were built that didn’t work, not the things that worked that you wish you had planned for on a big scale in advance. While those are the problems we always think about in retrospect, those are not 99% of the cost that you assume.”
Much like the early days of the web and, more recently, the smartphone revolution, early generative AI applications and use cases have glimpsed a powerful new AI-enabled future. But so far, they haven’t necessarily been game-changers.
“I call this the Angry Birds era of generative AI,” Kapoor said. “It’s not completely changing my life — no one’s doing my laundry yet. This year, every company I work with is putting something into production — small, internal, but they’re putting it into production because they’re actually solving the problems, about how to build the teams to make this happen. Next year is what I call the year of transformation, when people will start building applications that really start to change the trajectory of the company they work for.”
