January is always a month to remember because of Samsung. The smartphone brand usually launches the year’s Galaxy lineup. For example, the current Galaxy S25 series launched on January 22 this year. However, next year might be a different story. According to a new report, Samsung will hold its first Galaxy Unpacked event in late February.

Based on a report from the Korean outlet Money Today, Samsung is letting the Galaxy 26 series marinate a tiny bit more because of Galaxy AI. The brand is even hosting the event in San Francisco, the supposed center for AI technology.

The report also states that Samsung is evaluating its current lineup. This year, the brand launched the Galaxy S25 Edge. However, the new model’s sales have been reportedly lackluster, potentially prompting a return to the old model next year. The Galaxy S26 series might have the more traditional three-piece lineup consisting of the basic, the plus, and the ultra models.

Now, ahead of the new lineup’s launch, a lot of rumors are already going around about the Galaxy S26’s chipset. The brand is supposedly going all-in once again on Exynos. The upcoming Exynos 2600 can go up against Apple’s A19 Pro, if the reports are to be believed.

SEE ALSO: Samsung might bring a stronger Exynos to the Galaxy S26 series

HONOR has officially announced MagicOS 10 at its inaugural Global Developer Conference in Shenzhen, China.

MagicOS 10 is the world’s first AI agent operating system with self-evolving capabilities, signaling the industry’s shift from the OS era to the AI OS era.

In addition, HONOR also unveiled its 1×3×N ecosystem strategy alongside an upgraded HONOR AI Connect platform.

The framework opens HONOR’s AI stack to partners worldwide (the “1”), supports them with three enablement tracks — ecosystem, channel, and technology & brand (the “3”) — and scales across multiple industry clusters (the “N”).

These industries include education and office, smart home, audio and wearables, and even toys and pets.

From OS to AI OS

MagicOS 10 makes a multi‑dimensional leap from a conventional OS to a true AI OS. It pivots from basic device‑cloud coordination to an agent‑centric model, unlocking greater intelligence.

At the core of MagicOS 10 is a new and improved YOYO, now equipped with agentic capabilities to see, remember and execute, powered by HONOR’s proprietary MagicLM 3.0.

With one tap, users can ask YOYO to handle everyday tasks and creative workflows, such as AI‑powered color grading with Magic Color, complex command‑based image editing, and actions like food ordering, shopping, and ride‑hailing.

As multi‑device usage grows, MagicOS 10 is the first OS in the industry to fully break down platform boundaries. It enables seamless transfer of images, videos, and documents across HONOR, iOS, Android and Windows devices.

Moreover, OS also debuts the new Zero‑Gravity Transparency Design. This brings system‑wide transparent visual effects, from lock screen to desktop to in‑app views, with no increase in power consumption.

Additionally, the enhanced AI Deepfake Detection intelligently identifies fake voices, deepfake filters, and scam scripts to safeguard users.

HONOR AI Connect and 1×3×N Ecosystem Strategy

Launched in 2022, the HONOR Connect platform has already linked more than 30 million devices. At the conference, HONOR bared that t its upgrading tot he HONOR AI Connect platform, marking another leap.

HONOR AI Connect is built to lower partner onboarding barriers and reduce development costs. Developers can tap mature multimodal interaction capabilities (voice, vision, and more) without rebuilding complex base platforms, and access cross‑scenario AI collaborative services.

By systematically opening intelligent connectivity capabilities, the platform enables intelligence sharing, entry sharing, and ecosystem sharing.

HONOR has already partnered with more than 200 ecosystem partners and introduced a program that offers free IoT connectivity to deliver best‑in‑class value.

Through initiatives such as the HONOR Alpha Flagship Store and broader store access programs, HONOR provides more substantial resources and growth support to partners.

Looking ahead, HONOR remains committed to open collaboration. As part of its broad, long-term plan, the devices manufacturer is inviting global partners to co‑build the AI device ecosystem so intelligent technology benefits everyone.

Robotics company AgiBot has successfully deployed its Real-World Reinforcement Learning (RW-RL) system on a pilot production line with Longcheer Technology.

The company specializes in embodied intelligence, and the project marks the first application of RW-RL in real industrial robotics.

It connects advanced AI innovation with large-scale production, signaling a new phase in the evolution of intelligent automation for precision manufacturing.

Precision manufacturing lines have long relied on rigid automation systems that demand complex fixture design, extensive tuning, and costly reconfiguration.

Where AgiBot’s RW-RL system comes in is addressing such pain points. It enables robots to learn and adapt directly on the factory floor.

Within just minutes, robots can acquire new skills, achieve stable deployment, and maintain long-term performance without degradation.

The system also autonomously compensates for common variations, such as part position and tolerance shifts

Meanwhile, during line changes or model transitions, only minimal adjustments and standardized deployment steps are required. This dramatically improves flexibility while cutting time and cost.

Moreover, AgiBot’s system allows for flexible reconfiguration. Task or product changes can be accommodated through fast retraining; such solution exhibits generality across workplace layouts and production lines.

This milestone signifies a deep integration between perception-decision intelligence and motion control. And it represents a critical step forward unifying algorithmic intelligence and physical execution.

Unlike many laboratory demonstrations, AgiBot’s system also achieved validation under near-production conditions. It has completed a full loop from cutting-edge research to industrial-grade verification.

Moving forward, AgiBot and Longcheer plan to extend real-world reinforcement learning to a broader range of scenarios. These include consumer electronics and automotive components, with focus on developing modular and rapidly deployable robot solutions.