• Horizon AI
  • Posts
  • Use These AI Image Filter Prompts to Create Viral Photos 🔥

Use These AI Image Filter Prompts to Create Viral Photos 🔥

Alibaba launches Wan3.0 AI video model 🎬

Author

Gina Acosta
August 24, 2026

Sponsored by

Welcome to another edition of Horizon AI,

You can easily transform the look and feel of an ordinary image, but the right prompt is key to getting the effect you want. In today's issue, we're sharing some AI filter prompts for recreating popular photo styles, from nostalgic early-2000s digital cameras and night vision to cinematic looks and more.

Let’s jump into it!

Read Time: 4.5 min

Here's what's new today in the Horizon AI

  • Alibaba Releases Wan3.0 Video Model

  • Thomson Reuters Builds Its Own AI Model Using Chinese Open-Source Technology

  • AI Tutorial: Try These 10 Image Filter Prompts

  • AI Tools to check out

  • AI Findings/Resources

  • The Latest in AI and Tech 💡

AI News

ALIBABA

Alibaba Releases Wan3.0 Video Model

Alibaba officially rolled out Wan3.0, its latest AI video generation model, with major improvements over its predecessor, Wan2.5.

Details:

  • Wan3.0 can generate videos up to 30 seconds long, doubling the maximum duration of its predecessor. It can also recommend the best video length based on the user's prompt and extend existing clips.

  • The model supports multiple types of input, including text, images, video, audio, PDFs, PowerPoint presentations, and web pages. A single prompt can combine up to 10 images, five videos, and five audio clips.

  • Alibaba says Wan3.0 improves consistency and visual quality, helping preserve characters, objects, and spatial layouts from reference materials while reducing common problems such as visual distortion and visual drift.

  • The company is pitching Wan3.0 for a wide range of uses, including film and short-drama production, advertising, marketing, tourism, music videos, and social media content.

Wan3.0 is available in beta through the wan.video website, Alibaba Cloud Model Studio, and via API on Qwen Cloud, with Standard and faster Prime versions available at different prices.

TOGETHER WITH KODEKLOUD

Pass the Kubernetes exam on your first attempt.

RBAC. Networking. Storage. KodeKloud breaks these complex Kubernetes concepts down into hands-on practice, until you are ready to ace your exam.

Provision live clusters, drain nodes, debug failing pods, fix networking, set up storage. Every concept is explained in plain language before you type a command.

The curriculum is built backwards from the exam objectives and the work you will do on the job. Nothing extra. The scenarios you practice are the ones the exams test, and the ones you hit in production.

Challenges, hands-on labs, and crash courses teach beginners to think like a Kubernetes engineer from day one. Advanced labs go deeper into security, scaling, and complex troubleshooting. Every lab runs on real infrastructure with automated validation.

You're not memorizing answers. You're solving real Kubernetes problems. Starting from zero or sharpening what you have, this is certification prep that sticks.

THOMSON REUTERS

Thomson Reuters Builds Its Own AI Model Using Chinese Open-Source Technology

Thomson Reuters has launched its first in-house AI model, aiming to reduce its reliance on expensive third-party models while building AI capabilities around its expertise in legal and professional information.

Details:

  • Thomson-1 is based on Snowdon, an AI model created by adapting Alibaba's open-source Qwen model. Thomson Reuters worked with Imperial College researchers to modify the model for its own requirements.

  • The model was trained on the company's own data and by its domain experts, and it performs competitively with the strongest frontier models on the market, according to the company.

  • The new model will initially handle document review and other specialized tasks, some of which are currently powered by Anthropic's Claude, although Thomson Reuters says the goal isn't to replace Claude entirely. Instead, Thomson-1 will focus on areas in which the company has expertise.

Cost is a major factor behind the move. Building an internal model gives Thomson Reuters more control over its AI infrastructure and reduces the need to continually pay other companies for model access. Chief technology officer Joel Hron describes it as 'renting a house versus buying a house.' With an in-house model, he says, 'you are building equity in something that you own for the long-term, and that compounds over time.'

AI Tutorial

Try These 10 Image Filter Prompts

Go to ChatGPT (or another AI of your choosing), sign up, and try these:

  1. iPhone Blurry Effect 

Apply an aggressive, high intensity handheld camera shake effect as if the photographer took the shot while walking quickly or violently shaking the camera during a long exposure. Keep the main subject perfectly sharp, frozen, and photorealistic with NO distortion to the face, body, pose, outfit, or core structure. Only the background and trailing edges should show motion. Create chaotic left to right motion streaks across the entire scene, with long, thick luminous smears, doubled and tripled motion trails, jittery horizontal vibrations, wide bands of stretched highlights and reflections, and soft darkness pulled into bright streaks.

Simulate real aggressive camera movement by introducing micro jitters inside the streaks, layering multiple ghost gradients, and creating uneven intensity across the smear, as if the camera shook erratically. Blend multiple ghost gradients tightly around the subject's silhouette, with soft, semi transparent contour trails hugging the hair, shoulders, arms, torso, and clothing edges. The trails should feel like light being torn backward, not duplicated subjects. Allow the dominant color of the subject's outfit to softly bleed into the motion blur behind them.

Increase the glow and drag of all bright elements in the scene. Highlights should stretch into thick glowing ribbons, reflections should smear dramatically, and bright colors should bleed together into a continuous motion wash. Everything must remain fully blended, fluid, and organic, with no hard edges or cutout look. The background should feel like an explosive burst of motion and energy, while the subject remains perfectly crisp and untouched.

The final image must look like a real extreme long exposure photograph with aggressive handheld shake.

AI Tools to check out

✏️ FigEditor: Create publication-ready AI scientific figures from text, sketches, or references.

💬 Chatzy: AI platform for Customer Journey Automation.

✨ Pikes: Creative AI suite for CPG brands. Generate, edit, and ship photos, ads, and video from your real products.

📝 Video Transcript: Bring in a YouTube link, video file, or audio recording and create timed text for editing, research, reporting, translation, and accessible content workflows.

💥 Koji: AI-native customer interviews & research platform.

TOGETHER WITH KODEKLOUD

100 real-world tasks. 8 tool categories. 1 complete DevOps stack.

Master Git, Docker, Kubernetes, Linux, CI/CD, Terraform, and monitoring through structured tasks across real-job scenarios, and earn a shareable credential that proves you can apply what you’ve learned.

No lectures, no theory. Build a public portfolio as you go. Earn a verified badge. Free.

AI Findings/Resources

😬 A person representing themselves in court hid a prompt injection attack in a filing asking an AI system to side with them. Full article here

👀 A Paris-based tech founder created a viral open-source project, 'Watermarks Remover,' in response to Anthropic's AI watermark announcement

The latest in AI and Tech

Apple Music will introduce visible AI labels later this year to help users identify music that was created with significant use of AI.

Some of Nvidia’s largest customers have been told prices of servers containing its AI chips will rise by more than 15% in many cases with memory chip costs soaring.

The company is calling for changes to California’s SB 53 AI safety law, including monitoring frontier models during training and evaluation and stronger cybersecurity protections throughout development.

Nvidia is in talks to invest in Perplexity at a valuation above $30 billion. That would value the AI search company more than 50% higher than its previous funding round, while its annualized revenue has grown from $250 million to more than $750 million.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

We'd love to hear your thoughts on today's email!

Your feedback helps us improve our content

Login or Subscribe to participate in polls.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻