Artificial intelligence investment has reached unprecedented heights, with $40 billion in global venture funding going to the AI sector in the second quarter of 2025. Yet beneath this surge lies a critical question: Where does genuine value creation diverge from speculative fervor?
Members of the Senior Executive AI Think Tank—technology leaders specializing in machine learning, generative AI and enterprise applications—offer insights that cut through the market noise. Their collective expertise spans decades of building AI-powered solutions across industries from healthcare to logistics, providing a grounded perspective on where AI investments deliver real, measurable returns.
“We’ve seen companies achieve an 80% reduction in manual effort by deploying an AI agent that does nothing but review and classify inbound legal contracts based on a defined playbook.”
‘Narrow AI’ Delivers Immediate Impact—but Foundational Factors Outlast Flash
Nikhil Jathar, CTO of AvanSaber Technologies, says the most compelling near-term AI opportunities exist in what he calls “narrow AI”—solutions targeting specific, high-cost business problems. His team leverages AI, virtual reality and the cloud to build enterprise solutions that optimize supply chain and inventory management and other business operations.
“We’ve seen companies achieve an 80% reduction in manual effort by deploying an AI agent that does nothing but review and classify inbound legal contracts based on a defined playbook,” he says. “This isn’t glamorous, but its ROI is immediate and massive.”
However, Jathar’s “smart investment” horizon extends far beyond these immediate wins. “The most durable, long-term value won’t come from the models themselves, which are becoming commoditized,” he explains. Instead, he advises investors to focus on three areas:
- Companies with unique, proprietary datasets that will be valuable for training specialized models.
- The foundational “picks and shovels”—hardware and cloud infrastructure providers.
- Vertical SaaS companies that deeply embed AI into an indispensable industry workflow.
“If I were placing long-term bets, my money wouldn’t go to the flashiest apps,” Jathar concludes. “It would go to the infrastructure builders and the ‘boring’ vertical SaaS companies using AI to dominate a niche like agriculture, logistics or drug discovery.”
“The future belongs to AI that knows your business better than your best employee.”
Domain-Specific Intelligence Trumps General Models
With a Ph.D. in Data Science and over 20 years of experience in AI, Sarah Choudhary, CEO of ICE Innovations, says AI’s long-term value proposition lies in domain-specific intelligence systems. Her company develops AI-powered mobility and automation solutions and intelligent systems.
“AI’s short-term wins lie in infrastructure like NVIDIA and verticalized copilots that wrap models in workflow-specific value,” Choudhary observes. “But the long-term value? That’s in domain-specific intelligence—AI built not to be general, but to deeply understand industries like healthcare, supply chains or legal systems.”
The productivity, cost and human benefits of domain-specific systems like the UK’s Surgery Hero can be enormous, making them not only invaluable assistants for professionals in a variety of sectors, but also an appealing option for investors with an eye on social impact.
“My bet is on companies building context-rich, trustworthy systems with human-in-the-loop design and access to real-world data,” Choudhary continues. “Forget flashy prompts; the future belongs to AI that knows your business better than your best employee.”
“I’d bet on vertical AI, open foundation models and infrastructure firms powering scalable, secure AI deployment.”
Proprietary Data, Edge AI and Open Models Offer Lasting Value
Suri Nuthalapati brings deep expertise in AI, data and cloud technologies to his role as Data and AI Leader for the Americas at Cloudera. He believes that investors who are looking for long-term value in AI need to look past the “hype.”
“Short-term wins include AI-native productivity tools and inference infrastructure,” Nuthalapati says. “However, durable value will come from proprietary data plus AI, autonomous agents and edge AI.”
Nuthalapati’s work gives him unique insight into the challenges of building scalable, secure enterprise AI. He recommends investors keep an eye on open foundation models—large AI models whose architecture, parameters or training data are openly available, allowing enterprises to adapt them for specific needs without surrendering control to a third-party provider.
“I’d bet on vertical AI, open foundation models and infrastructure firms powering scalable, secure AI deployment,” Nuthalapati concludes.
AI’s Exponential Value Lies in Augmenting Human Potential
The most transformative AI investment opportunity may lie in fundamentally reimagining how artificial intelligence integrates with human workflows. As CEO of r.Potential, an enterprise intelligence company that helps businesses plan and prepare for workforces where human and digital workers coexist, Greg Shewmaker offers a unique perspective.
“As we automate work, we must ensure we don’t dehumanize workers,” Shewmaker emphasizes. “We have an opportunity to create advanced AI that elevates human achievement, rather than simply being a low-cost alternative.”
Shewmaker says that task-focused AI initiatives may be time-savers—but they’re not game-changers.
“Currently, companies focus on building AI agents as ‘smart tools’ for discrete tasks like closing tickets or generating code, which creates linear value: more output, same thinking,” he explains. “The real transformative potential lies in digital workers who understand your business reality and participate in it.”
Shewmaker says the most effective AI systems don’t just execute tasks; they model scenarios and anticipate outcomes. This, he says, is where the true exponential value of AI lies—as a “force multiplier” for humans across an organization, helping them make more informed, higher-quality decisions faster than ever before.
“We are betting on digital workers that learn to think, collaborate and execute as true co-workers—grounded in context, transparent in reasoning, accountable by design and amplifying the characteristics that make humans extraordinary,” Shewmaker concludes.
What Investors and Execs Need to Know
- There’s immediate ROI in narrow AI solutions that address specific, high-cost business problems. Deployments targeting defined workflows can deliver significant reductions in manual effort and clear, measurable returns.
- “Flashy” apps may not have legs. “Foundational” AI assets—proprietary datasets, hardware and cloud infrastructure—and vertical SaaS companies that dominate niche industries offer longer-term value.
- Domain-specific intelligence systems deeply understand individual industries. AI applications with specialized knowledge of healthcare, supply chains or legal systems command premium pricing and create defensible competitive advantages.
- Speed is nothing without security and scalability. Vertical AI and open foundation models can help companies achieve adaptable and safe AI solutions.
- The best solutions amplify human potential. Systems that model scenarios, anticipate outcomes and compress decision timelines create exponential value by supporting human intelligence and insight.
Prioritize Depth Over Dazzle
The AI investment landscape is flush with opportunity, but not all plays are equal. Short-term wins are emerging in targeted applications, verticalized copilots and productivity tools. However, long-term value will come from infrastructure, proprietary data, vertical SaaS integration and AI systems designed as collaborative co-workers.
For investors and executives alike, the lesson is clear: The future belongs to those who build AI into the very fabric of their industries. The best systems are grounded in context, powered by unique data and supported by scalable, secure infrastructure—and centered on helping humans chart the best path forward.
