Good morning, AI enthusiast.

TikTok's parent company, Bytedance, has entered the high-stakes generative video race. Its new model, Seedance 1.0, is showcasing impressive capabilities that put it in direct competition with giants like OpenAI's Sora and Google's Veo.

The model demonstrates remarkably smooth motion and character consistency across different scenes, two key hurdles in AI video. With another major player now in the mix, what does this mean for the future of high-end AI video tools and their accessibility for creators?

In today’s AI recap:

• Bytedance's new Sora-rivaling video AI

• Midjourney’s upgraded Style Reference feature

• OpenAI overhauls prompt management for developers

Bytedance's Video Challenger

The Recap: TikTok's parent company, Bytedance, has unveiled Seedance 1.0, a new text-to-video model with impressive capabilities that positions it as a direct challenger to OpenAI's Sora and Google's Veo.

Unpacked:

• The model generates video with remarkably smooth motion and physical realism, handling everything from subtle expressions to large-scale action scenes.

• It natively supports multi-shot storytelling, maintaining character and style consistency across different scenes and camera angles.

• Creators can generate content in various styles, from photorealism to illustration, with precise control over complex actions and camera movements.

Bottom line: Bytedance's entry intensifies the competition in the high-stakes generative video market. This signals that top-tier AI video creation tools may soon become more accessible, pushing the boundaries of digital content.

Midjourney's Style Upgrade

The Recap: Midjourney just rolled out a major update to its Style Reference feature, introducing the V7 system for more precise style transfers. The new release also includes a command for greater creative exploration.

Unpacked:

• The V7 system enables you to transfer artistic styles and create mood boards with more precision than previous versions.

• A new --sref random command introduces significantly more variation, giving you a powerful tool for discovering unexpected visual styles.

• While the new style system is the default, you can fall back to the old system at any time by simply using the --sv 4 parameter in your prompt.

Bottom line: This update gives creators finer control over their visual outputs, streamlining the process of achieving a consistent aesthetic. It marks another step toward making AI image generation a more intuitive and directed creative tool.

OpenAI's Prompt Overhaul

The Recap: OpenAI is streamlining developer workflows by introducing Prompts as a new API primitive, allowing prompts to be versioned, managed, and reused across its platform.

Unpacked:

• The new Prompt object acts as a reusable resource within the API, which can be preconfigured with specific models, tools, and messages.

• In the Playground, a new 'Optimize' button helps refine prompts for the API, and developers can now save and share prompt configurations, replacing the old presets system.

• This update allows developers to centrally manage prompts, ending the need for manual copy-pasting between services like the Playground and API.

Bottom line: This change significantly reduces friction for developers building complex applications, making prompt engineering more systematic. It signals a move toward treating prompts as durable, version-controlled assets rather than disposable text.

Where AI Experts Share Their Best Work

Join our Free AI Automation Community

Join our FREE community AI Automation Mastery — where entrepreneurs, AI builders, and AI agency owners share templates, solve problems together, and learn from each other's wins (and mistakes).

What makes our community different:

• Real peer support from people building actual AI businesses

• Complete access to download our automation library of battle-tested n8n templates

• Collaborate and problem-solve with AI experts when you get stuck

Dive into our course materials, collaborate with experienced builders, and turn automation challenges into shared wins. Join here (completely free).

The Shortlist

Amazon announced a $13 billion investment to build and expand its AI and cloud infrastructure in Australia, marking the largest-ever tech investment in the country.

Moonshot AI open-sourced Kimi-Dev-72B, a new coding model that it claims is state-of-the-art among open-source models after achieving a 60.4% score on the SWE-bench Verified benchmark.

Salesforce found in a new research paper that top LLM-based AI agents struggle with basic Customer Relationship Management (CRM) tasks, with success rates dropping to 35% on multi-step problems and showing low awareness of data confidentiality.

Singapore will use predictive AI on national health records and other data to help doctors forecast disease risk years in advance as part of its Healthier SG initiative.

Signing off,

David, Lucas, Mitchell — The Recap editorial team