AgentPulse | AI Agents Directory
Posts
#28: Agents on the Move: Google Styles You, Cursor Follows You, and Meta Steals Your Team

#28: Agents on the Move: Google Styles You, Cursor Follows You, and Meta Steals Your Team

AgentPulse: Your Weekly Dose of AI Agents News.

O P
July 03, 2025

In partnership with

Welcome back, AI Agent Enthusiast!

In today’s Agent Pulse:

📢 Top AI Agents News
⚔️ Agent Arena
Featured Agents
Free courses
Papers

📢 TOP AI Agents News

👗 Doppl by Google Labs: AI Outfits, Animated

Google just launched Doppl, an experimental mobile app that uses generative AI to transform how we shop and style—by turning your photo (or an outfit screenshot) into a personalized, animated try-on reel.

Key Features

Upload a full-body photo to see yourself in any outfit from your camera roll.
Doppl generates short AI-powered videos, letting you view the outfit move.
Save and share your virtual looks perfect for styling feedback or social sharing.

What to Watch

Still early-stage: expect glitches like texture hiccups or floating edges.
U.S.-only availability and feature scope is limited (tops, bottoms, dresses).
Includes invisible SynthID watermarks and explicit privacy-focused disclaimers.

Why It Matters

Doppl shows how AI agents are evolving into creative companions, not just functional tools.
By adding animation, it bridges the gap between static e-commerce and experiential retail boosting shopper confidence and cutting returns.
It’s a bold playground for agentic styling, hinting at future integrations—imagine an assistant that suggests outfits based on weather, calendar events, or past purchases.

Google’s Doppl is a fun, creative leap into the world of personalized, animated styling assistants. It's early and imperfect, but suggests a future where agents become our personal visual, dynamic, and emotionally expressive stylists.

💻 Cursor Mobile & Web: AI Coding in Your Pocket

Cursor just expanded its powerhouse agent experience beyond the desktop IDE—launching Cursor Agents on web and mobile browsers via PWA, plus seamless Slack integration .

Start coding tasks anywhere: From your phone or tablet, ask agents to fix bugs, scaffold features, or query your codebase—and let them run in the background.
Continuity across devices: Tasks started on mobile can be reviewed, edited, and merged back in the desktop IDE—context and history flow end-to-end api.
Team collaboration in-sync: Share agent runs or PRs with teammates, get Slack updates, and compare multiple agents with different prompts to optimize results

Why It’s a Big Deal

Code-anywhere workflows are now real: no need to wait for your desktop—Cursor agents can do the heavy lifting while you’re offline or on the go.
Fracture-free team flow: Start a task in Slack or mobile, finish it in your IDE—without losing CRT or context.
Agent-level experimentation: Launch multiple agents using different models (e.g. GPT‑4 vs Claude) and compare outputs directly from mobile or web.

Cursor web and mobile agents cement a new reality: AI assistants are untethered—working behind the scenes as you move through your day, syncing seamlessly with your workflow and your team. It’s a striking leap from “modal AI coding” to pervasive, context-aware agent collaboration.

⚔️ It’s Poaching Season

The AI agent world isn’t just competing on model quality anymore - it’s now a full-blown talent war, and both tech giants and startups are hunting in packs.

Meta fired the opening salvo, quietly recruiting a full squad of top researchers from OpenAI’s Zurich office. This wasn’t a drip of hires it was a deliberate group lift. The recruits, including Lucas Beyer, Alexander Kolesnikov, Xiaohua Zhai, and Trapit Bansal, bring deep expertise in multimodal and vision models. Their work laid much of the foundation behind OpenAI’s image-understanding capabilities.

Reports suggest offers exceeded $100 million in total comp, as Meta pushes to accelerate its “Superintelligence” team. And it’s not just pay - Meta is also dangling autonomy, compute, and research freedom.

But the poaching wave isn’t limited to Big Tech.

Cursor, the fast-growing AI coding agent startup, has also joined the race hiring two key leaders from Anthropic’s Claude Code team: Boris Cherny (chief architect) and Cat Wu (head of product). These hires bring first-hand experience from building one of the most advanced agentic dev tools in the space. Cursor’s strategy is clear: skip the catch-up and recruit the blueprint.

Why This Matters:

Talent signals roadmap: When Cursor brings in Claude Code’s core team, it’s a clear bet on more autonomous, dev-native agents - likely beyond code suggestion into full codebase orchestration.
Team lifts > solo hires: Meta’s group approach gives them tight research alignment and velocity from day one. The agent space moves fast, and cohesion matters more than ever.
The stack is still fluid: Foundational design decisions on workflows, UX, autonomy are still being made. The people making them now will shape how agents look and behave for years.

Takeaway for builders: If you want to understand where AI agents are headed, don’t just follow the models. Follow the talent. In a space where product, infrastructure, and LLMs collide, the team you build might be your most defensible moat.

Learn AI in 5 minutes a day

This is the easiest way for a busy person wanting to learn AI in as little time as possible:

Sign up for The Rundown AI newsletter
They send you 5-minute email updates on the latest AI news and how to use it
You learn how to become 2x more productive by leveraging AI

AArena: All-in-One AI Platform

Bring your task. Get best results.

TOP 3 AI on July 2:

ChatGPT-4o
Llama 3.3 70B Instruct
Llama 4 Maverick 17B

👉 Enter the Arena
👉 Battleboard

Other News Headlines

LVMH & Google explore “agentic AI” for luxury brands – LVMH is testing internal and customer-facing AI agents (via Google’s Vertex AI) for tasks like clienteling and trend forecasting, while startups like Alta leverage autonomous agents for personalized styling. (Source)
Honor’s Magic V5 smartphone debuts with integrated “Yoyo” AI agent – The new foldable phone can autonomously create presentation slides and manage ride-hailing across multiple apps, with future agentic features through an Alibaba partnership. (Source)
Agentic AI deployments triple to ~33% of organizations – KPMG reports one-third of firms have moved AI agents into full production, though Gartner warns 40% of these projects may be canceled by 2027. (Source)
Databricks launches Agent Bricks beta for domain-tuned agents – Built using Mosaic AI Research, this tool auto-generates synthetic data and benchmarks to streamline enterprise agent creation in areas like ML pipelines and document extraction. (Source)
Ubisoft introduces NFT-linked AI agents in new blockchain game – In Captain Laserhawk: The G.A.M.E., AI agents govern in-game decisions (voting, proposals) on behalf of NFT owners, blending human control with autonomous agent actions. (Source)