CES 2026: AI everywhere-Nvidia's Rubin, AMD's Ryzen AI PCs, Alexa+, robots, and the weirdest gadgets

AI headlines CES 2026: Nvidia Rubin, Ryzen AI 400, assistants in cars and homes, and robots headed to worksites. Plan for on-device, simulate first, and set clear data guardrails.

Published on: Jan 09, 2026
CES 2026: AI everywhere-Nvidia's Rubin, AMD's Ryzen AI PCs, Alexa+, robots, and the weirdest gadgets

CES 2026: What Actually Matters for Builders, IT, and Product Teams

CES 2026 is in full swing in Las Vegas. AI is front and center again, but the show still delivers fresh hardware and a few left-field ideas. Below is a fast, useful rundown of the biggest moves and why they matter for your roadmap.

Nvidia: Rubin architecture and open models for autonomous machines

Nvidia's Rubin architecture is set to begin replacing Blackwell in the second half of the year, with clear gains in speed and storage. The company also highlighted Alpamayo, an open family of AI models and tools that autonomous vehicles will use this year. The strategy lines up with turning Nvidia's stack into the Android for generalist robots.

  • Practical take: plan GPU capacity and storage upgrades ahead of Rubin-era model sizes. Build simulation-first pipelines for autonomy and robotics. If you're experimenting with digital twins, see Nvidia Omniverse for planning and validation.

AMD: Ryzen AI 400 and a push to mainstream on-device AI

AMD opened CES with partners across AI research and startups and put PCs at the center of its message. The Ryzen AI 400 Series focuses on expanding AI workloads on laptops and desktops.

  • For IT: expect mixed CPU/GPU/NPU workloads and new offline AI use cases. Update device policies for local model usage and data controls.
  • For developers: profile models against the NPU early; ship fallback paths for cross-hardware consistency.

Ford: AI assistant lands in the app first

Ford's AI assistant rolls out in the company's app, with a targeted 2027 in-vehicle release. It's hosted on Google Cloud and built on off-the-shelf LLMs. Details on the in-car experience are still light.

  • For product teams: scope guardrails now (voice, context, and drive-time safety). Plan an offline mode and clear data boundaries.

Caterpillar + Nvidia: Automated construction with a virtual testbed

The new "Cat AI Assistant" pilot for excavators pairs with a separate effort using Nvidia's Omniverse for construction planning and execution. Think: assistive autonomy on-site, with simulation upstream for planning.

  • Action: model your job sites in simulation first. Treat telemetry, labeling, and human-in-the-loop review as core features, not afterthoughts.

Boston Dynamics + Google: Atlas gets a new AI stack

Hyundai and Boston Dynamics are working with Google's AI lab to train and operate Atlas robots, including a new iteration shown on stage. The signal: more standardized, cloud-scale tooling in humanoid robotics.

  • For engineers: structure tasks as skills with repeatable evaluation. Expect more offboard training and on-device execution.

Amazon: Alexa+ and Ring step further into everyday workflows

Alexa+ is getting a full push, including a browser-based experience for early access and a refreshed app. Amazon also announced updates to Fire TV and Artline TVs, and Ring added features from fire alerts to an app store for third-party camera integrations.

  • Opportunity: new surfaces for voice + multimodal apps. If you're building home or office automations, revisit your Alexa integrations and Ring-compatible workflows. See Alexa developer resources for APIs and policy updates.

Razer: Project Motoko and AVA (AI in your periphery)

Razer's concepts this year center on ambient AI. Project Motoko aims to work like smart glasses without the glasses. Project AVA places an AI companion avatar on your desk.

  • Takeaway: plan for low-friction, "always-there" assistants. Keep latency under control and give users quick ways to mute, pause, or erase history.

Clicks Communicator: Keyboard-first phone with modern polish

Clicks Technology's $499 Communicator brings a physical keyboard back, plus a $79 slide-out keyboard for other devices. Early impressions point to a comfortable grip, contoured back, and a slightly elevated screen with a curved chin that helps protect the keys when set face down.

  • Signal for product teams: physical inputs aren't dead. If typing precision and screen visibility matter, hardware keys and smart ergonomics still win.

Skylight Calendar 2: Family planning with AI assist

The updated Skylight Calendar syncs calendars from different sources and can create to-dos from messages or photos, plus reminders and more. It's a clean example of AI acting as connective tissue between everyday inputs and structured tasks.

  • Pattern to copy: frictionless capture (text, images) → structured actions (tasks, reminders) with user confirmation.

Lego's first CES: Smart Play System with Star Wars sets

Lego previewed bricks, tiles, and Minifigures that interact and play sounds, with debut sets themed around Star Wars. It's tactile play enhanced by simple, clear feedback.

  • Idea: tangible interfaces + audio cues are great for learning and accessibility. Consider similar patterns for onboarding and education products.

Breakout session highlights

  • Palmer Luckey pushed retro aesthetics as a legitimate product strategy, not just nostalgia.
  • The "learn once, work forever" era may be over as models and hardware refresh faster; expect continual learning patterns.
  • Preview of "The Audacity," a Silicon Valley-based series, drew attention around tech culture and ethics.
  • Roku's $3 streaming service expansion suggests lean subscription models still have room.
  • Jason Calacanis offered a $25,000 bounty for an authentic Theranos device, a stark reminder of due diligence and proof in health tech claims.

The oddities

The quirky side of CES delivered, as always. Keep an eye on experimental form factors and ambient AI concepts-they often seed next year's mainstream features.

Action guide

  • General: Expect AI assistants in more places (cars, homes, desktops). Look for products that remove steps, not add screens.
  • IT & Development: Prepare for on-device inference on PCs (Ryzen AI 400). Define data retention and audit for assistants. Use simulation for robotics and autonomy before field tests.
  • Product: Scope AI features that start simple: summarize, schedule, suggest. Offer clear controls and transparency. Revisit hardware ergonomics-physical inputs can improve completion rates.

Keep leveling up

If you're aligning skills with what's shipping next, browse curated learning paths by role and tech stack here: AI courses by job and popular AI certifications.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)
Advertisement
Stream Watch Guide