‘We Are in the MS-DOS Era of AI’

Today’s AI, despite its power, is still primitive in usability, much like the early command-line days of personal computing. “We are in the MS-DOS era of AI,” according to Abhishek Mathur, VP of engineering at Figma, a San Francisco-based software company that provides a cloud-based design and prototyping platform.

Mathur told AIM, “every prompt, every agent, every new modality is a design waiting to happen,” underscoring that we are only seeing the early stages of what AI can do and become. 

Users interact with AI through prompts the way people once typed commands into MS-DOS, effective, but far from intuitive or accessible. The real transformation will come when AI moves into its “Windows moment,” where visual, multimodal, and agent-driven interfaces replace text-based prompting and make AI truly seamless. 

Just as graphical operating systems unlocked the mass adoption of computing, the next phase of AI will be defined not just by better models, but by better design. 

Previously, chief product officer (CPO) at Microsoft, Aparna Chennapragada also told AIM that she thinks this is “the DOS-to-GUI moment of AI.” She added that Microsoft is pursuing two paths—integrating AI into familiar tools like Word, Excel and PowerPoint, and developing AI-first interfaces such as the M365 Copilot app and new AI editors.

The early days of computing were text-based, command-driven, and intimidating to the average user. Then came the graphical user interface (GUI), a breakthrough that made computing accessible to everyone.

Figma, Mathur believes, will play a similar role in AI-era creativity. “The next generation of designers will define how humans interact with AI. We’re only just getting started, ” he said.

Bengaluru as the Command Line 

With its world-class talent across engineering, product, and design, Bangalore is uniquely positioned to shape this future. The city needs to now translate this depth into global leadership in AI product design. 

The next generation of AI-native products will require multidisciplinary teams that combine model capability with deep understanding of workflows and user behaviour. 

Leslie Joseph, principal analyst at research and advisory firm Forrester, told AIM that if the ecosystem continues to invest in product leadership, design education, and cross-functional collaboration, the city can evolve from being a delivery hub to being a birthplace of world-leading AI-native products.

India is Figma’s second-largest market, home to one of its most vibrant user communities. “Every year, 35 million design files are created in India,” Mathur said. “That’s 35 million ideas, products, and innovations, and it’s growing exponentially.”

The Friends of Figma community, made up of designers, engineers, and product leaders, is thriving across cities like Bengaluru, Pune, and Hyderabad. 

Figma is constantly expanding its product ecosystem, from FigJam, the company’s collaborative whiteboard, to Figma Make, its latest prompt-to-application product. “Figma Make allows you to build applications inspired by your company’s design systems,” said Mathur. “So when you turn an idea into an app, it’s instantly on-brand, functional, and production-ready.”

The enthusiasm isn’t limited to design. India also leads globally in the adoption of Figma Make, the AI-powered product that turns text prompts into functioning prototypes. 

“The highest number of makes in any country have been created in India,” Mathur shared. India is increasingly a development hub for Figma. The company is hiring across design, engineering, and developer advocacy roles in Bengaluru, with plans to build global products from India.

Role of AI in the Next Phase

The next leap in AI isn’t just about better models—it’s about the rise of agentic AI, where systems don’t merely respond but act. In this phase, AI agents will extend the reasoning abilities of underlying models into real-world products

“What distinguishes meaningful products is the way they translate ambiguous human intent into structured tasks, guide users through decisions, and create trust through clear boundaries and feedback,” Joseph added.

Talking to AIM, Siddon Tang, SVP of engineering and product, chief architect, and GM of the Asia Pacific region for TiDB, an open source database, stated that “Great product design is what makes AI feel magical.”

Tang further added that as we move into the era of AI agents, design becomes even more critical. Agents may run on top of LLMs, but they still depend on high-quality APIs, guardrails, permissions, memory structures, and interaction models.
 

All of that is product design.

This is the reason why “Product-led growth (PLG) becomes even more potent in the AI era,” Tang said, adding, “think of products like Dropbox, Lovable, Manus, and others—they scaled because the experience was so good that users naturally shared them.”

The biggest breakthroughs ahead won’t come from slightly smarter models but from products that translate model intelligence into real-world capability, clarity, and confidence.

Tools like Canva, Figma, Notion, and a wave of vertical SaaS platforms are embedding AI everywhere. And as users bring more ideas, workflows, and creative inputs into these tools, AI itself improves.

AI’s role in design, Mathur argued, is twofold: it lowers the barrier to entry while raising the ceiling of what’s possible.

“As an engineer or product manager, you can now be an effective designer,” he said. “And as a designer, you’re freed from repetitive tasks, localisation, renaming layers, filling mockups — to focus on craft, quality, and vision.”

Figma, in his words, is not about replacing designers, but about amplifying creativity across roles. “Designers will now be in the business of codifying quality,” he added. “They’ll define the brand’s DNA in design systems — and everyone else can build on it.”

The post ‘We Are in the MS-DOS Era of AI’ appeared first on Analytics India Magazine.