The Five Questions Tech Executives Are Asking About AI

What a week! The theme of this year’s GDS Digital Innovation Summit was “Digital Frontiers: where trust, humanity, and innovation converge,” and last week saw that convergence in real time. It was great to be in the room where it was happening, but let’s not pretend everyone didn’t have one big thing on their minds: Generative AI.

We all want to use AI to disrupt the status quo and vault progress forward — but not at any cost. And nowhere is that more acute than in software development, which has become the epicenter for both AI innovation and anxiety. At Crowdbotics, we live in that epicenter. I had the pleasure of hosting two round tables at the event: “Exploring the Transformative Power of Generative AI – Efficiency, Innovation, and Beyond.” Teams told us they are piloting or investing in Gen AI to grow critical parts of the business, including:

Improving the speed of development
Improving operational efficiency
Modernizing legacy codebases
Centralizing company knowledge in AI-aware knowledge bases
Driving personalized experiences for customers and prospects

Throughout the unscripted, uncensored roundtables, it became evident that while Gen AI’s efficiency gains and innovative capabilities are broad, they are also vastly complex — with ethical implications, potential biases, and responsible deployment strategies.

5 Important Questions Executives are asking about AI

So what are those questions? Here’s what we overheard in the room and what we made of it. (All names have been withheld to protect the innocent.)

How Can AI Deliver in Adapting and Improving Existing Systems?
“We have so many legacy systems that can’t talk to anything; they’re all customized. How could AI help us translate all of that?”

AI can modernize and integrate legacy systems, but enhancing data quality and observability is essential before AI can derive strategic value from these systems. Our participants noted the potential value of AI in data analysis, though structured data in data lakes, data warehouses, etc., is required. AI should not be seen as a silver bullet for data problems; instead, companies should focus on educating leaders, setting clear expectations, and gradually integrating AI into legacy systems.

Shameless plug: At Crowdbotics, we use AI to understand and modernize codebases, providing strategic results. We can also help you create the infrastructure you need simply by organizing the code you already have.
How Do We Educate the Organization on the Benefits of Investing in AI?
“More than half of you mentioned bringing the company along with them, educating leadership, and trying to help them understand how you need to operate in this new world.”

Education is top of mind for everyone. There exists a landscape where countless innovative solutions emerge, each aiming to address various challenges. This abundance often creates an “early hype cycle”, characterized by a plethora of solutions in search of problems. Amidst this wave of possibilities, stakeholders find themselves carefully weighing which initiatives to greenlight. This uncertainty underscores the complex task of identifying and prioritizing the genuine problems that should be addressed, navigating through a sea of promising yet untested innovations. Bridging the knowledge gap among C-level executives is crucial for supporting AI investments. Establishing centers of excellence can serve as a resource for learning and testing AI applications, ensuring leadership understands AI’s value and supports its strategic implementation.
How Can We Quantify the Actual ROI of AI – And, How Is AI Also Better?
“There’s a lot of vendors out there that are using AI, but the value is unclear. What is the ROI in that?”

AI can significantly enhance productivity and streamline operations by automating mundane tasks and improving customer interactions. However, addressing data quality and observability is key to unlocking efficiency gains. Real-world examples include automating content creation and accelerating software development processes. With every new AI initiative, you should start with estimated ROI and measure ROI throughout the initiative.

Proof of Concepts (POCs), bake-offs, and centers of excellence are great ways to demonstrate value and get stakeholders comfortable with AI investments. Additionally, participants noted it was helpful to cite positive customer feedback to convince skeptical stakeholders investment is helpful. These tactics provide data to prove AI’s benefits, help manage change, and demonstrate tangible outcomes.
How Can We Get Beyond ‘AI for AI’s Sake’?
“How do we get beyond executive tourism? – i.e. doing something just to prove to execs that we’re doing something with AI.”

Moving beyond superficial AI projects to strategic, impactful implementations is essential. With AI being early in the hype cycle, many in the group noted they received an influx of AI investment requests from executives with questionable value. One participant called this “executive tourism”, which we thought was perfect. While organizational education is needed, participants recommended creating ROI estimates for all asks to keep focus on only the highest value initiatives.
In the Rush to AI, Are We Thinking Enough About Security?
“If the grid goes down as a critical system, that’s when people realize the importance of security in AI applications.”

AI’s integration into critical infrastructure—think communications, emergency, banking, energy grids, etc. —is posing major potential security risks, such as AI bots or system vulnerabilities. As AI-generated code becomes more prevalent, the automation of penetration testing and compliance becomes more crucial. Emerging threats include AI-generated voices and attacks by AI bots. Companies must develop robust AI-powered security systems, such as voice recognition security, to differentiate between human and AI-generated voices.

Big Takeaways: The Themes to Remember

Here’s the big thing: as much as Gen AI is top of mind, our biggest challenges with adopting it aren’t tech related — or even about AI at all. In the spirit of the Digital Innovation Summit’s tagline, they are about “trust, humanity, and innovation” converging. By far the most popular topic in both discussions was navigating the human problems inherent in new technology adoption early in the hype cycle.

Here are the key themes we noticed, that need to be solved for:

Poor Understanding of AI from Execs: Many execs push for AI adoption without grasping where it’s most impactful. This leads to mandates for AI in low-ROI areas instead of asking teams to recommend high-impact uses.
Spotting Snake Oil: It’s tough to discern vendors with real solutions from those selling AI snake oil. Many startups have flashy pitches but lack substance. Companies want AI, but it’s hard to separate genuine innovation from hype.
Fear of Replacement: A significant concern is job security. Some participants shared stories of teams refusing to use AI, fearing it would make their roles obsolete. This fear can be a major barrier to AI adoption.
Data Dilemma: AI needs structured, clean data for analysis. Some business leaders expect new AI tools to solve data silos and data hygiene problems magically. But without good data, even the best AI tools are hamstrung.

Addressing these challenges requires strategic solutions. POCs and bake offs can demonstrate real-world AI value, while customer success stories can build internal support. It’s crucial to start with solving real business problems rather than implementing AI for its own sake. Positioning AI as a helper, not a competitor, can ease internal resistance and foster a more welcoming environment for AI initiatives. Clear ROI metrics are essential to avoid superficial projects and ensure AI initiatives tie back to measurable outcomes.

Our Take

At Crowdbotics, we use AI strategically to enhance and uncover what is there, ease communication and organize innovation. Our CodeOps solution also prioritizes sustainability by reusing and modernizing vetted code — more sustainable practices for your business and fewer bottlenecks for your QC team. Our emphasis on leveraging vetted, tested code also ensures robust security in AI applications. Innovation is at the core of what we do, exemplified by our PRD-AI tool that democratizes innovation and enables more teams to contribute effectively. Our partnership with Microsoft further amplifies our ability to deliver cutting-edge AI solutions.

Last week made it super clear that everyone also wants a balanced, thoughtful approach to AI adoption. We need to focus on real-world applications and strategic outcomes rather than chasing hype. Promoting sustainability and addressing security concerns in AI applications should be prioritized. Finally, investing in IT infrastructure and data management will drive meaningful progress.

Go from Idea to Code with AI

Schedule a Demo