Avatar Sweeps VES Awards

In partnership with

Welcome to VP Land! AI agents have been getting all the buzz — but Perplexity and Anthropic just said hold my beer.

In today's edition:

Avatar's seven-win sweep at the VES Awards
Perplexity and Anthropic strike back at AI agents
Adobe Firefly Quick Cut and the blank page problem
Google Flow's image-to-video upgrade and Quiver's native SVG model

Avatar Takes Seven VES Awards; Sinners Wins Supporting VFX

Avatar: Fire and Ash dominated the 24th Annual VES Awards with seven wins, including Outstanding Visual Effects in a Photoreal Feature and the Emerging Technology Award for the Kora Fire Toolset — Weta's proprietary pipeline innovation for the film's fire simulations.

Supporting VFX: Ryan Coogler's Sinners took Outstanding Supporting Visual Effects in a Photoreal Feature over bigger-budget productions.
Animation: KPop Demon Hunters led with three wins — Outstanding Animation, Outstanding Character (Rumi), and Outstanding Effects Simulations in an Animated Feature.
Real-time: Ghost of Yōtei won Outstanding Visual Arts in a Real-Time Project; Prehistoric Planet: Ice Age took Outstanding VFX in a Photoreal Episode.
Honorees: Jerry Bruckheimer received the VES Lifetime Achievement Award; Sir Richard Taylor of Wētā Workshop received the VES Visionary Award.

Read the full breakdown

SPONSOR MESSAGE

Trust-First AI, Built Into Your Browser

Agentic workflows are everywhere. Real trust is still rare.

Norton Neo is the world’s first AI-native browser designed from the ground up for safety, speed, and clarity. It brings AI directly into how you browse, search, and work without forcing you to prompt, manage, or babysit it.

Key Features:

Privacy and security are built into its DNA.
Tabs organize themselves intelligently.
A personal memory adapts to how you work over time.
This is zero-prompt productivity. AI that anticipates what you need next, so you can stay focused on doing real work instead of managing tools.

If agentic AI is the trend, Neo is the browser that makes it trustworthy.

Try Norton Neo and experience the future of browsing.

Download Norton Neo

AI Agents Have Been Getting All the Buzz. Perplexity and Anthropic Just Hit Back.

AI agents have been getting all the buzz — but Perplexity and Anthropic just said hold my beer. Both companies dropped updates that push squarely into agentic territory — persistent memory, scheduled automation, mobile access, multi-model orchestration.

Perplexity launched Perplexity Computer, a new product that orchestrates 19 specialized AI models — Claude Opus 4.6 for reasoning, Gemini for research, ChatGPT 5.2 for context recall, Grok for lightweight tasks, Nano Banana for images, Veo 3.1 for video — connected to 400+ apps with persistent memory and real file system access. It handles long-running jobs and sandboxed code execution. CEO Aravind Srinivas calls it an "orchestra" versus a single-model system. $200/month for Max subscribers.

Anthropic shipped three Claude features in 48 hours:

Claude Code Remote Control: Mobile and browser access to local coding sessions via QR code or claude.ai/code. Execution stays on your machine — mobile is just a window. Pro/Max only.
Claude Cowork Scheduled Tasks: Automate recurring workflows via prompts and plugins on hourly, daily, or weekly cadence. Desktop app only. 15 new plugins added.
Claude Code Auto-Memory: Remembers project context, debugging patterns, and preferred approaches across sessions — no more re-explaining your stack every time.

The reality: memory, scheduling, mobile access, and multi-agent orchestration are the core features dedicated AI assistant platforms are built around. The big labs just decided they want that market too.

Adobe Firefly's video editor now includes Quick Cut, an AI feature that assembles raw footage into a structured first-draft edit. Describe what you want in natural language — "lengthy interview," "product demo," "vlog" — set aspect ratio and pacing, and Quick Cut takes it from there.

Scene detection: Automatically identifies visual and audio changes and finds optimal cut points.
Smart shot selection: Analyzes focus, composition, and action to pick the most polished takes.
Audio analysis: Detects key dialogue and sound cues to drive narrative flow.
B-roll integration: Upload or AI-generate supporting footage; Quick Cut organizes it into tracks with transitions.

The takeaway: Quick Cut solves the blank page problem for video editors — the part where you're staring at hours of raw footage and don't know where to start. We previously covered Firefly's AI video expansion; Quick Cut is the next step, automating the editorial workflow itself, not just the generation.

Google Flow Gets Image-to-Video Upgrade with Nano Banana 2

A few AI tool updates worth knowing about this week — two from Google, one from a new a16z-backed startup.

Nano Banana 2 (Gemini 3.1 Flash Image) is Google's new default image model, replacing Nano Banana 1 across Gemini app, Google Search, Flow, and Google Ads.

What's improved over Nano Banana 1:

Resolution: Up to 4K output (vs. 1K max in NB1), plus new extreme aspect ratios (4:1, 1:4, 8:1, 1:8)
Speed: Faster generation at Flash-class inference
Consistency: Maintains up to 5 characters and 14 objects across a workflow — a significant jump from NB1's handling
Intelligence: Uses real-world knowledge from web search for accurate subjects, text rendering, and photorealistic edits
Watermarking: SynthID + C2PA interoperability on all outputs (new)
Availability: Default everywhere — Nano Banana Pro still accessible via menu for specialized tasks

Google Flow got a major update built around Nano Banana 2 as the image engine feeding into Veo 3.1 video.

Ingredients to Video: Generate or refine images with Nano Banana 2, then pass them as reference frames into Veo 3.1 — anchors video output to specific colors, framing, and mood without long text prompts
Image generation in Flow is free — iterate on visual references at no cost before committing to video
New asset grid: Searchable, filterable canvas for organizing images, videos, and project elements
Enhanced editing controls: Extend clips, add or remove segments, select camera motion styles

Quiver AI raised $8.3M from a16z for Arrow-1.0, a multimodal model that generates clean, editable SVGs natively from text, images, or sketches. Most models convert pixels to vectors as a post-processing step — Arrow skips that entirely, treating SVGs as "visual code" based on founder Joan Rodríguez's StarVector research. Strong for logos, icons, technical drawings, and layered structures. Free beta at app.quiver.ai.

An ex-Google Maps PM built WorldView using AI agents and public data—live satellites, ADS-B aircraft, CCTV, seismic and traffic feeds on Google 3D tiles—sparking viral comparisons to Palantir and even a response from its co-founder about data fusion.

Stories, projects, and links that caught our attention from around the web:

🎬 TechCrunch asks what AI actually delivers for indie filmmakers — faster and cheaper, yes, but also lonelier production pipelines.

🔧 Nuke 17.0 lands with native Gaussian splat support and a new USD-based 3D system — Foundry going all-in on the formats that matter for virtual production.

🎥 DGA president Christopher Nolan laid out his AI position: use it as a creative accelerator, but mandatory disclosure and creator rights are non-negotiable heading into AMPTP negotiations.