✨ From vibe coding to vibe deployment. UBOS MCP turns ideas into infra with one message.

Learn more
Andrii Bidochko
  • Updated: March 18, 2026
  • 6 min read

Carl Pei Forecasts End of Smartphone Apps as AI Agents Take Over

Carl Pei Says Traditional Smartphone Apps Will Vanish as AI Agents Take Over


AI‑powered smartphone vision

In a bold SXSW interview, Carl Pei, CEO of Nothing, announced that the era of standalone mobile apps is ending – AI agents embedded in the next‑generation smartphone will handle tasks that today require dozens of separate applications.

Pei’s Vision for an AI‑First Device

During the 2026 SXSW conference in Austin, Pei painted a picture of a device that “knows you so well it can anticipate your intentions and act without you having to tap an app.” He argued that the current app‑centric model is a relic of the pre‑iPhone era, likening it to the old Palm Pilots and PDAs that forced users to juggle multiple tools for a single goal. Pei’s statement aligns with Nothing’s recent About UBOS narrative, which emphasizes AI‑driven experiences over conventional UI paradigms.

The AI‑First Device Concept Explained

The core of Pei’s proposal is an AI‑first smartphone that integrates large‑language‑model agents directly into the operating system. Instead of launching a “weather app,” the device’s AI would automatically surface a forecast when you glance at the lock screen. When you say “I need coffee,” the AI would coordinate messaging, map navigation, ride‑hailing, and calendar entries in a single, frictionless flow.

Key technical pillars include:

  • Persistent user intent modeling: Long‑term memory that learns habits, health goals, and work patterns.
  • Agent‑centric APIs: Interfaces designed for AI agents rather than human touch, enabling “zero‑click” interactions.
  • Secure, on‑device inference: Privacy‑first processing that keeps personal data local while still leveraging powerful models.

These components echo the capabilities showcased in UBOS’s OpenAI ChatGPT integration, where developers can embed conversational agents into custom workflows without building the underlying infrastructure from scratch.

Implications for the Traditional App Ecosystem

If AI agents become the primary execution layer, the role of the app store will shift dramatically. Instead of a marketplace for isolated binaries, developers will create “agent plugins” that expose functionality to the AI core. This mirrors the emerging Workflow automation studio, where users stitch together services via visual blocks rather than downloading separate apps.

Potential outcomes include:

  1. Reduced fragmentation: One AI layer can mediate between services, eliminating duplicate UI patterns.
  2. New revenue models: Subscription‑based “agent bundles” could replace per‑app purchases.
  3. Accelerated innovation: Developers focus on core intelligence rather than UI polish.

For startups, this shift opens a fast‑track to market. The UBOS for startups program already offers a low‑code environment to prototype AI‑driven experiences, positioning new ventures to ride the wave of agent‑centric design.

Memorable Quotes from Pei’s SXSW Interview

“The current way we use phones is very old‑school. It’s pre‑iPhone…you have lock screens, home screens, apps. It hasn’t really changed for like 20 years.”

“If I know your intention, I just do it for you, instead of having you jump through four different apps to grab a coffee.”

“The future isn’t the agent using a human interface. You need to create an interface for the agent to use. That’s the more future‑proof way of doing it.”

How This Differs From Today’s Smartphone Landscape

Today’s phones rely on a layered stack: operating system → app store → individual apps → user interaction. Pei’s AI‑first model collapses the stack into two layers: OS + AI core. The difference can be visualized in the table below.

Current Model AI‑First Model
Multiple apps per task Single AI agent per task
User‑driven navigation Agent‑driven execution
App store as revenue hub Agent marketplace & subscription bundles
Static UI/UX patterns Dynamic, context‑aware UI generated by AI

These contrasts illustrate why Pei believes “apps are going to disappear.” The shift is not merely cosmetic; it redefines how value is delivered on mobile devices.

Future Outlook: Industry Impact and Timeline

Analysts predict a three‑phase rollout:

  • Phase 1 (2026‑2027): AI assistants that execute single commands (e.g., “book a flight”).
  • Phase 2 (2028‑2029): Persistent intent models that proactively suggest actions.
  • Phase 3 (2030+): Full agent‑centric OS where third‑party developers ship “agent plugins” instead of apps.

Companies already experimenting with Phase 1 include the ChatGPT and Telegram integration, which lets users trigger complex workflows via simple chat commands. Meanwhile, the Chroma DB integration provides the vector‑search backbone needed for fast, on‑device memory retrieval – a cornerstone of Pei’s long‑term vision.

From an enterprise perspective, the Enterprise AI platform by UBOS is already positioning itself as the backend for large‑scale agent ecosystems, offering compliance‑ready data pipelines and role‑based access controls.

What This Means for You – Take Action Now

If you’re a developer, marketer, or business leader, consider these steps to stay ahead of the AI‑first wave:

  1. Explore Web app editor on UBOS to prototype agent‑driven interfaces without writing extensive code.
  2. Leverage the UBOS templates for quick start, such as the AI SEO Analyzer or AI Article Copywriter, to embed intelligent features into existing products.
  3. Join the UBOS partner program to gain early access to upcoming agent APIs and co‑marketing opportunities.
  4. Review the UBOS pricing plans to select a tier that supports scalable AI workloads.

Conclusion

Carl Pei’s bold claim that “apps will disappear” is more than a headline—it signals a paradigm shift toward AI agents that anticipate and fulfill user intent without the friction of traditional app navigation. As the industry moves from Phase 1 command‑based assistants to fully agent‑centric operating systems, platforms like UBOS are already laying the groundwork with integrations, low‑code tools, and enterprise‑grade AI infrastructure.

For tech‑savvy consumers and professionals, the message is clear: adapt now, experiment with AI‑first tools, and position your products to thrive in a world where the smartphone becomes a seamless, conversational partner rather than a collection of isolated apps.

Read the full interview on TechCrunch:
Carl Pei predicts AI agents will replace apps.

Explore more AI‑driven innovations on the UBOS homepage and discover how the future of mobile experiences is being reshaped today.


Andrii Bidochko

CTO UBOS

Andrii Bidochko is an AI entrepreneur and researcher focused on AI agents, reinforcement learning, and autonomous systems. He writes about the technologies shaping the future of machine intelligence, from frontier models and agent architectures to real-world AI applications.

Sign up for our newsletter

Stay up to date with the roadmap progress, announcements and exclusive discounts feel free to sign up with your email.

Sign In

Register

Reset Password

Please enter your username or email address, you will receive a link to create a new password via email.