Top 5 Takeaways from Luminate AI’s 2023 Data & AI Leadership Exchange

Data and AI teams–assemble! 🐚

We were thrilled to connect with some brilliant minds in the AI space at Luminate AI’s recent kickoff event in New York City, where we got to chat about the future of AI and return with new learnings about this ever-so-hot field.

Luminate AI, a new start-up led by chief data and analytics officers, hosted fireside chats, networking opportunities, and panel discussions at its Data & AI Leadership Exchange, which underscored the vast potential of the D&A space, particularly across the generative AI landscape. From social impact to business strategy to concerns over regulation, here are some of the hot topics that stood out.

We're just getting started

Remember how mind-boggling the idea of the first smartphone was—the now defunct Blackberry—many moons ago? Or even the very first iPhone? (You mean I can combine the power of my iPod mini, Motorola Razr, digital camera, and Tom Tom into one? Say no more.) 

Now think of where we are today with the iPhone 14. We’ve certainly come a long way. (If you had told us in 2007 why we would need not one but THREE cameras on our phone…)

We’re at the same place with the rise of generative AI, an analogy brought up by prominent AI strategist Mark Minevich, who said we’re really only at the “age of Blackberry” when it comes to generative AI. In the same way we’ve come leaps and bounds with smartphone advancements, think of where we’ll be with generative AI in the years to come. 

Minevich also pointed out the “exponential and experiential” nature of AI in general. Helping to “lead your organization into the future,” he said, you simply have to try it, recommending pilots and POCs as a way to get started if you haven’t already done so.

There's room for more at the table

Here’s a major advancement in D&A that’s helping more users unlock the power of AI, particularly when it comes to driving AI-powered insights: self-service capabilities.

The data experts will undoubtedly always retain a critical role in an organization’s decision-making and success. However, with the increasing emergence of more business-user-friendly analytics tools, organizations are making use of self-service capabilities to enable more users to access and analyze data.

Self-service analytics—a theme prevalent in our one-on-one conversations with D&A leaders at the event—empowers business users to perform queries and generate their own reports without IT or analyst support. You don’t need to be a BI architect; you don’t need to write any SQL code. With the right self-service tools, all you need is business acumen to uncover the right data-driven insights.

Networking opportunities at Luminate AI’s Data & AI Leadership Exchange

Governance is key

While unlocking the power of AI for more users is awesome, governance and safety are always going to be major discussion points.

For example, you’ve probably heard of generative AI hallucinations—i.e., the propensity for large language models (LLMs) to “make up facts”—but have you heard of prompt injections? 

Sameer Maskey, founder and CEO of AI solutions provider Fusemachines, explained this example of a newer, more concerning generative AI advancement, the ability (well, vulnerability) for people to inject prompts into a query to make an LLM override previous instructions. Basically, people can inject prompts into a system to get it to do things it shouldn’t exactly do, Maskey warned.

If your organization is considering implementing generative AI in your workflows, it’s imperative to work with an established AI-powered analytics vendor that has put proper guardrails and governance in place. 

Additionally, here are more key recommendations we heard around mitigating security concerns:

• Don’t put your customer data into ChatGPT and other natural language AI chatbots—yet. Anonymize all of your data. (Bharat Krish, board advisor at Fusemachines)
• Establish an AI governance at your organization, focusing on explainability, transparency, responsibility, and accountability. (Mark Minevich)
• Consider running your own open-source LLMs. Although this can be expensive, it maintains data privacy and enables you to fine-tune the technology for your domain. (Sameer Maskey)
• Help drive the conversation forward around establishing regulation. (Mark Minevich)

AI for good

There’s no doubt that AI is helping humans carry out some truly remarkable—and noble—feats. Minevich noted that he’s seeing AI help bridge the education gap by developing curriculums in underserved school areas, as well as augmenting research in healthcare in an effort to eradicate diseases—just to name a couple use cases.  

Ori Carmel, founder and CEO of Sowen—a consultancy that helps organizations use data and technology for social impact—pointed out the “incredible potential” for AI to transform the world. 

During the event’s “Unlocking Social Impact with Data and AI” session, Carmel suggested organizations set up a task force to create a set of standards to ensure their efforts are moving in the right direction. 

“We all agree that organizations should be doing good in the world. But not all of them are able to do this effectively,” he said.

Instead of just checking a box on your completed efforts, Carmel explained, you also need to measure their impact. Make sure you’re investing in measurement tools to assess the true outputs of your efforts—not just on how much you’re spending—and focus on building the right leadership to drive your initiatives.

It’s not enough to only pay lip service to sustainability and equality, Minevich added. You also need to show return on investment of impact in these areas.

AI: a means to the end

Lastly, Carmel and Minevich rounded out the “AI for good” session with an important point to keep in mind: AI is not necessarily the end goal. It’s more of a means to the end.

“People see data and AI as the end,” Carmel explained. “But this is a failure. These are tools to drive an end goal.”

In other words, instead of thinking first and foremost how to get your organization’s D&A capabilities up to snuff, think about what you’re trying to accomplish in the first place. If you’re a CPG company, are you trying to better understand consumer behavior patterns? If you’re in marketing, are you trying to get a handle on customer segmentation so you can optimize your campaigns? If you’re a company trying to evolve from reactive to proactive ahead of the next big change in the economy, are you hoping to cut down analysis time to be agile? 

Nate Rackiewicz, chief data officer and co-founder of Luminate AI, put it this way: 

“Deciding ‘I guess it’s important because it’s complicated’ doesn’t cut it.” 

Instead, make sure your organization recognizes that D&A and AI are mission-critical priorities, and always measure the impact and value-capture of these investments on the back end. “Keep your eye on the ‘so what?’ of it all,” Rackiewicz said.

Whatever your goal is, make sure you know what you want your technology to solve for. AI might be the next shiny object—albeit a very shiny and very transformative one—but don’t forget the big picture.

Chris Walker (left) and JD Costantini (right) from Tellius’ Northeast team

Learn more about GPT analytics

Tellius is excited to kick off this exciting partnership with Luminate AI as they bring their thought leadership and expert advisory services to clients across industries. Stay tuned for more events and opportunities in the AI space.

To learn more about Tellius’ new GPT-integrated product enhancements, check out Tellius Copilot.