AI is accelerating at lightning speed — how do you keep up?

As artificial intelligence evolves from high-school-level reasoning to surpassing PhD-level intelligence in just a few years, professionals everywhere are racing to understand how to stay ahead. To break it down, NDB alumni Arjun Mehta and Sophie Alvarado joined NDB Analyst Zara Kadeem for a candid conversation on emerging AI trends and the evolving nature of work.

Meet the Speakers

• Sophie Alvarado, former NDB Business Analyst turned co-founder at SignalSense Labs, a company specializing in AI-model evaluation and benchmarking.

• Arjun Mehta, former Engagement Lead at NDB and now a Partner at Apex Venture Capital, investing in next-generation AI SaaS companies.

• Zara Kadeem, current NDB Analyst who previously helped build multimodal AI tools in the startup ecosystem.

What Exactly Is an AI Model?

Sophie:
“The simplest way to imagine an AI model is like a massive offline dataset combined with endless matrix math — all to predict the next piece of information, or token. When you interact with any AI tool, the app is choosing which underlying model to use and how much context to give it. That context can include chat history, external tools, images, even code-execution abilities.”

As Sophie’s team benchmarks new entrants weekly, she notes that we’re now in a “global top-ten race,” where leading AI models compete intensely on reasoning, accuracy, and speed.

What’s Most Exciting about the Current AI Wave?

Sophie:
“The pace. The leap from GPT-4 to GPT-5 happened shockingly quickly. Models keep becoming smarter and cheaper at the same time — a rare combination in tech.”

Zara:
“One emerging breakthrough is the rise of interoperability through MCP – Model Context Protocol. It allows multiple agents, databases, and tools to talk to each other automatically. Instead of jumping across tabs or apps, your workflow becomes unified.”

She also highlights the explosion of multimodality — tools that accept text, voice, images, video, and produce all formats back.

Arjun:
“What excites me most is real-time voice AI. We’re finally at the point where speech-to-text, text-to-speech, and voice-to-voice interactions can occur fluidly. This unlocks entire industries.”

Real-World Impact: What Does This Look Like?

Arjun:
“In banking, for example, contact centers represent a huge portion of operational costs. Even before generative AI, banks were experimenting with voice automation. Now, voice AI can handle verification, fraud queries, and basic troubleshooting — 24/7, in any language.”

But the real disruption? Pricing.

“Traditional software was priced per user seat. With AI handling support tickets, the value shifts from per seat to per successful resolution. That upends the economic model of customer-service software entirely.”

Zara:
“This same model applies in healthcare. AI-driven diagnostic assistants, nutrition coaches, and wearable-based insights are pushing medicine into a personalized, proactive space.”

She also mentions the growing debate around “over-reliance on AI,” inspired by a viral MIT Media Lab report.

“Ultimately, AI should augment, not replace human decision-making.”

How AI Changes the Nature of Work

Sophie:
“When I was at NDB, the most exciting work was the big question-thinking — not the repetitive tasks. Future analysts and consultants will spend more time on strategy because AI will automate the slide-building, compiling, and grunt work.”

Research Areas Worth Following

Zara:
“One powerful frontier is simulation:

• How do we simulate human behavior?

• How do agent-to-agent interactions scale?

• How can AI systems learn through synthetic environments?”

She encourages users to examine model cards to understand limitations.

Arjun:
“The best path to staying competitive? Use AI every day. Code with it. Write with it. Test multiple models. There’s a moment when you move from ‘AI curious’ to ‘AI-native.’ And that shift puts you ahead of almost everyone.”

The Power of Strong Networks

The three actually connected through Zara — who met Sophie at an AI conference and later connected with Arjun through mutual friends.

Zara:
“Networks start with genuine conversations, not formal structures. Most of what I learn in AI comes from people just sharing ideas.”

Arjun:
“Through Zara, I ended up reconnecting with colleagues from my NDB days — some of whom I hadn’t spoken to in years.”

How Do They Personally Use AI?

Arjun:
“I’m obsessed with AI-assisted coding tools. Anyone can build custom software now.”

Sophie:
“Coding agents. Easily the most magical tech in my daily workflow.”

Zara:
“My pick is AI-powered smart glasses. A newly released pair can give real-time, speech-to-speech contextual guidance based on what you’re looking at.”

Favorite Frameworks They Still Use Today

Arjun:
“Have an agenda — then be ready to abandon it when needed.”

Sophie:
“Feedback frameworks. They shaped my approach to building healthy team culture.”

Zara:
“The classic two-by-two prioritization matrix, especially in product strategy.”