The Best and Worst of AI in 2025

In partnership with

Welcome to VP Land! We're closing out 2025 with our first-ever AI awards, a major Hollywood coalition launch, and a wave of motion control and 3D breakthroughs.

In today's edition:

The Denoisies: Best and Worst of AI
Creators Coalition unites Hollywood on AI ethics
All the AI updates from the end of the year

The Denoisies: 2025's Best and Worst in AI

Over on Denoised, we took a look back at the year with our first-ever Denoisies, handing out 13 AI awards covering everything from Best Image Model to Funniest AI Fail.

A few highlights:

Best Image Model:

Addy's pick: Z-Image - The lightweight, locally runnable model wins for its customization via LoRAs.
Joey's pick: Nano Banana Pro - Joey went with Google's model for pure efficiency in getting the best results as fast as possible.

Best Video Model:

Joey's pick: Veo 3 - As the most definitive model of 2025, Veo 3 dominated in detail rendering and prompt adherence, setting the standard for the year.
Addy's pick: Kling 2.6 - Addy crowned it the "people's champ" for ending the year neck-and-neck with Veo 3.1 despite coming from smaller company Kuaishou.

Biggest AI Comeback:

Addy's pick: OpenAI Sora - Dominated the zeitgeist early in 2024, disappeared while competitors emerged, then came roaring back with Sora 2 and the meme-creation app that became the number one download that week. Despite eventually neutering the fun IP generation features, the comeback was undeniable.
Joey's pick: Google - Went from irrelevant video models at the start of 2025 to dominating with Veo 3, Nano Banana Pro, and Gemini across nearly every category.

Check out the rest of our picks

SPONSOR MESSAGE

Stop Planning. Start Building.

End of the year? Or time to start something new.

With beehiiv, this quiet stretch of time can become your biggest advantage. Their platform gives you all the tools you need to make real progress, real fast.

In just days (or even minutes) you can:

Build a fully-functioning website with the AI website builder
Launch a professional-looking newsletter
Earn money on autopilot with the beehiiv ad network
Host all of your content on one easy-to-use platform

If you’re looking to have a breakthrough year, beehiiv is the place to start. And to help motivate you even more, we’re giving you 30% off for three months with code BIG30.

Start building for 30% off today

Hollywood Creators Launch Anti-AI Exploitation Coalition

The Creators Coalition on AI launched as a cross-industry organization pushing for ethical AI implementation in entertainment. Founded by Daniel Kwan, Daniel Scheinert, Joseph Gordon-Levitt, and David Goyer, the coalition brings together hundreds of filmmakers, actors, writers, and producers to establish shared standards before AI further disrupts creative labor.

Key Details:

Four core pillars - The coalition demands consent, compensation, and transparency for training data; job protection plans for workers; guardrails against deepfakes; and protections for human creativity in storytelling. They're proposing an industry-wide AI Advisory Committee to set standards.
Notable signatories - Guillermo del Toro, Alfonso Cuarón, Cate Blanchett, Natalie Portman, Natasha Lyonne, Mark Ruffalo, Riz Ahmed, and Taika Waititi are among 500+ signatures. Members come from DGA, SAG-AFTRA, WGA, PGA, and IATSE, participating as individuals rather than official union reps.
"Fast vs. right" framing - The coalition explicitly states this isn't about rejecting AI but about drawing a line "between those who want to do this fast, and those who want to do this right." They're calling for collaboration with tech companies on accountability systems.
Training data criteria - AI models must meet four requirements: consent from creators, user controls, fair compensation, and transparent enforcement mechanisms. No specific policy proposals yet, just guiding principles.

The Final AI Updates of 2025

A rapid-fire look at all the AI updates to close out 2025:

Video & World Models:

Kling 2.6 Motion Control - Kling AI feature enabling precise, reference-image–driven control of character actions, expressions, and motion for up to 30 seconds.
LongVie2 - End-to-end autoregressive video world model generating up to 5 minutes with strong controllability and temporal consistency
Lucy Motion - Decart's image-to-video with path-based motion control where you draw trajectories and the model animates exactly along them
Odyssey-2 Pro - Real-time world model that generates every pixel frame-by-frame, simulating physics forward for minutes
Wan Vision Enhancer - fal's in-house video upscaler with 5 creativity levels from faithful restoration to artistic reimagining
DAAAM - Builds hierarchical 4D scene graphs as spatio-temporal memory, enabling agents to describe anything, anywhere, at any moment

Text-to-Image:

FLUX.2 [max] - Black Forest Labs' top-tier quality with grounded web search, up to 10 reference images for consistency, ranked #2 on Artificial Analysis
GPT Image 1.5 - OpenAI's API release with improved instruction following, better text rendering, and enhanced preservation of logos and faces

Image-to-3D:

Qwen-Image-Layered - Alibaba’s AI model that decomposes images into editable RGBA layers without external segmentation
SHARP - Apple’s feedforward system that converts a single 2D image into a photorealistic 3D scene in under a second, supporting real-time rendering at 100+ FPS
TRELLIS.2 - Microsoft's open-source 4B-parameter model producing up to 1536³ PBR textured assets with arbitrary topology support
Aurora - Creatify's image-to-avatar model creating ultra-realistic avatars from a single image with expression control and scalable quality

Audio Separation:

SAM Audio - Meta's first unified model isolating any sound from complex audio mixtures using text, visual, or span prompts

Surprise Entry in the AI Game:

Zoom AI Companion 3.0 - Achieved 48.1% on Humanity’s Last Exam—an exceptionally rigorous benchmark for expert-level knowledge and deep reasoning—marking a 2.3% gain over the previous state of the art through improved summaries and reasoning.

Join Addy behind the scenes as he breaks down his workflow for building a hybrid live-action/AI short, from shot consistency to post-production tweaks.