Latest news with #GopikrishnanAnilkumar
Forbes
5 days ago
- Business
- Forbes
Top Infrastructure Upgrades Needed To Support AI And Quantum Growth
Emerging technologies like artificial intelligence and quantum computing have the potential to reshape nearly every industry—but only if global infrastructure evolves fast enough to support them. From intelligent energy systems and ultra-fast connectivity to new governance models and quantum-safe security, foundational upgrades are essential. The approaching tech revolution requires more than just faster chips or bigger data centers; business and government leaders must rethink how they manage power, information, regulations and talent. Below, members of Forbes Technology Council highlight the most critical infrastructure shifts needed to unlock the full potential of next-generation innovation. 1. Time-Aware Infrastructure What we need is time-aware infrastructure—systems that don't just process data fast, but also understand when to act, defer or self-correct. Next-gen tech isn't just about power; it's about precision, timing and trust. Without temporal intelligence, scale becomes noise. - Adam Zachary, Mobiz IT 2. Expanded Global Connectivity To support AI and quantum tech, the world needs much faster, more stable and more secure internet (like 5G, 6G and quantum internet). Without strong global connectivity, these technologies can't transfer huge amounts of data quickly or work effectively for health, research and industry. - Gopikrishnan Anilkumar, Amazon Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify? 3. Quantum-Ready Talent The critical upgrade isn't hardware—it's human capital. We're building quantum-ready infrastructure while our workforce still struggles with basic cloud concepts. The biggest bottleneck will be finding engineers who understand both legacy systems and quantum principles. Infrastructure is useless without operators who can manage it. - Anna Turos, Lighthouse HQ 4. Modern Power Grids We need a power grid glow-up—energy infrastructure is the bottleneck. AI and quantum suck juice like V8s with turbos. Without smarter, decentralized grids—think microgrids and edge energy—tech will stall. I saw this at GM data centers. Future tech isn't just code—it's current. No watts, no wonder. - Tim Bates, Godfather of Tech 5. Sustainable Compute Practices The AI gold rush is overwhelming infrastructure that was never built for this scale. Power grids are strained, data centers are bottlenecked and companies keep throwing more hardware at the problem. That model is broken. Without a shift to workload efficiency, longer asset life and sustainable compute practices, AI's growth will collapse under the weight of its own demand. It's time to rethink the foundation. - Tomas O Leary, Origina 6. Agentic AI Functionality One critical upgrade is building infrastructure that supports agentic AI, or AI that acts on behalf of users. To unlock its full potential, global systems must prioritize performance, scalability and seamless integration, enabling AI to transform everyday users into power users by automating repetitive tasks and enhancing productivity. - Adam Lieberman, Finastra 7. Global Ethical Protocols We need global standardization of ethical protocols. To harness AI and quantum safely, nations must establish interoperable ethical frameworks and audit standards. Without harmonized regulations on data use, algorithmic bias and quantum encryption, cross-border innovation risks fragmentation and public mistrust, stifling the very technologies meant to advance humanity. - Cristian Randieri, Intellisystem Technologies 8. Data-Centric Architectures To support next-gen tech like AI and quantum, global infrastructure must shift to data-centric architectures—scalable, low-latency systems that move data closer to compute. Speed, locality and parallelism will define success as workloads outpace traditional network and storage designs. - Sven Oehme, DataDirect Networks 9. Increased Bandwidth Accessibility I do not believe global infrastructure can handle the bandwidth necessary for AI and quantum to grow substantially. Currently, 1 Gbps symmetrical service is difficult to find, and when it is available, the price is often out of reach, especially for small businesses and tech startups. The lack of available bandwidth and the cost of bandwidth will slow the adoption of AI and quantum technologies. - Robert Martin, Oil City Iron Works, Inc. 10. Upgraded Energy Distribution Systems A critical shift is modernizing power infrastructure. AI and quantum workloads demand immense, stable energy. Without investment in clean, high-density energy grids and smarter distribution, innovation will bottleneck at the plug. Energy is the new scalability frontier. - Luis Peralta, Parallel Plus, Inc. (dba ParallelStaff) 11. Data That's Decoupled From Applications A critical shift is decoupling data from applications. Today, data is linked to specific apps, limiting AI's ability to access and learn from it. Unlocking next-gen tech like AI and quantum requires treating data as a universal asset, accessible across platforms and systems. - Yogesh Malik, Way2Direct 12. HPC-Grade Infrastructure To support AI and quantum computing, enterprises must shift to HPC-grade infrastructure built with accelerated processors, high-speed networking and scalable storage. This upgrade enables faster data processing, low-latency hybrid workloads and quantum-safe security, making it essential for next-gen innovation across IT. - Sai Sandeep Ogety 13. More Resilient, Smart-Routing Infrastructure Survivability is critical for global infrastructure as AI and quantum continue to boom. Dense, distributed workloads mean zero tolerance for downtime. Single-region, monolithic systems won't cut it. We need infrastructure built to expect failure, reroute around it and stay consistent when regions or providers go dark. If you can't guarantee consistency, you aren't ready for what's coming. - Spencer Kimball, Cockroach Labs 14. Data Interoperability A critical shift is data interoperability. The future of AI depends less on raw compute power and more on whether systems—especially in healthcare—can speak the same language. Without harmonized, structured data, even the best models are flying blind. - Zameer Rizvi, Odesso Inc. 15. Post-Quantum Cryptography A critical shift will be the widespread adoption of post-quantum cryptography (PQC). As quantum computers advance, they'll threaten current encryption methods. Global infrastructure must rapidly transition to quantum-resistant algorithms to protect sensitive data and critical systems across all industries, from finance to healthcare, before a 'quantum doomsday' scenario. - Dennis-Kenji Kipker, 16. Deterministic Networking AI and quantum don't just need speed; they need precision. One failed deployment showed us it's not bandwidth that breaks systems, but timing. A few milliseconds of network jitter, and the model collapses. The next wave of infrastructure must prioritize deterministic networking and tightly synchronized edge computing. Without timing, intelligence stalls. - Aditya Vikram Kashyap, Morgan Stanley 17. Energy-Resilient Compute Fabrics To support AI and quantum at scale, global infrastructure must shift toward energy-resilient compute fabrics—integrated systems that co-optimize compute, cooling and clean power. These workloads demand not just speed, but also sustainability. Without upgrading to intelligent, decentralized and carbon-aware compute grids, the physical backbone for next-gen tech will become a bottleneck. - Giridhar Raj Singh Chowhan, Microsoft 18. Modernized Data Centers One key upgrade will be modernizing global data centers to meet the huge demands of AI and quantum computing—both in terms of processing power and energy use. This means adding advanced cooling systems, faster networking and hybrid setups that combine classical and quantum technologies, all while shifting to more sustainable energy sources to keep up with rapid data growth. - Ajit Sahu, Walmart
Forbes
25-07-2025
- Business
- Forbes
Enterprise AI Isn't A Feature—It's A New UX Paradigm
Gopikrishnan Anilkumar is a Principal Product Manager at Amazon, where he leads product management for multiple AI products and services. The rise of generative AI is creating a powerful shift in enterprise technology. From customer support to analytics, organizations are integrating AI into critical workflows, expecting it to enhance productivity, automate repetitive work and drive strategic decision making. But although the technology has matured rapidly, the user experience (UX) layer (i.e., how users interact with and trust these systems) often lags behind. Too many AI-powered enterprise tools fall short not because of model limitations but because they misunderstand how users work, decide and trust. Generative AI isn't just a back-end engine. It represents an entirely new kind of interaction. Designing for it requires rethinking user experience from the ground up. AI Isn't A Button—It's A Behavioral Shift In most AI implementations, product teams treat intelligence as a bolt-on feature, like a button that triggers a model or a prompt box that feeds into a text generation service. But generative AI introduces uncertainty, autonomy and fluidity. It doesn't just return results, as it even participates in decisions. This requires a shift from static feature design to behavioral design. AI systems should be context-aware, act only when useful and adapt to user preferences. In regulated environments, even more care is needed. Each action must be auditable, suggestions explainable and the AI must know when to step back. Prompting Isn't A Scalable UX Pattern Many enterprise tools built around large language models (LLMs) assume users will communicate with AI via typed prompts. Although this may work for developers and early adopters, it's rarely intuitive for business users. Professionals in healthcare, finance or operations don't have time to construct ideal prompts. They want accurate outcomes based on minimal input. Effective AI UX provides structured interactions: smart defaults, fill-in-the-blank suggestions, clickable intents and domain-specific controls. These reduce the friction between user need and AI action. Trust Is A UX Outcome, Not Just A Technical Goal Trust in AI is often discussed as a matter of model quality or data governance. Although these are critical, trust is also a function of how the system presents itself to the user. A confident answer without evidence can erode trust. A transparent answer with sources, confidence scores and options for review builds it. Good UX makes uncertainty visible. It gives users the tools to verify claims, retrace logic and even reject an answer when appropriate. In this way, trust becomes an experience outcome, not just a back-end aspiration. For example, an AI assistant may show which documents were retrieved from a knowledge base to generate an answer, cite time stamps and internal sources, indicate confidence levels with visual cues, and offer a fallback when the model is unsure or the input is ambiguous. These elements transform AI from a black box to a trustworthy product. Enterprise Context Demands Systems Thinking Enterprise workflows are complex. They span systems, roles, data policies and compliance boundaries. A generative model may provide a correct answer in isolation but fail when deployed in a broader system that requires traceability, authorization or multistep context. To realize AI's full promise, product leaders must move away from "feature thinking" (the idea that AI is just another button or tab). Instead, they must embrace "system thinking"—designing AI as an intelligent, adaptive layer that spans data, logic, UX and decision making. This means designing: • With human-in-the-loop interactions, not blind automation • For progressive disclosure, not information overload • With fail-safes and fallbacks, not brittle dependencies • For explainability and auditability, not just speed Conclusion The next generation of successful enterprise AI tools won't be defined by how advanced their models are. They'll be defined by how well they integrate into human workflows with clarity, control and credibility. As product and technology leaders, our responsibility isn't just to deploy powerful models but to ensure they are usable, understandable and trustworthy. That starts with UX. Generative AI isn't a feature to layer on top of legacy interfaces. It's a paradigm shift in how enterprise software behaves. The sooner we embrace this, the faster we'll build systems that truly empower the people who use them. Disclaimer: The opinions in this article are those of Gopikrishnan Anilkumar and not representative of any organizations he has worked for. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
