In partnership with |
 |
|
|
xAI just gave Grok Imagine a serious upgrade: creators can now lock a character's face and voice across scenes, generate straight from a text prompt, and render in native 1080p. It's the biggest jump for the tool since its multi-reference feature launched, and it's rolling out to everyone this week. |
The bigger shift here is control. AI video used to mean rolling the dice on a prompt and hoping the character stayed consistent from shot to shot. If voice-matching and multi-reference hold up outside the demos, that gamble mostly disappears. |
Today in AI: |
xAI's Grok Imagine adds voice and image references Autodesk's Flow Studio brings 3D control to AI video OpenAI's Astra model cracks 10 old math problems |
The GTM Playbook HubSpot Had to Acquire |
|
Warmly ran pipeline, outreach, and lead scoring on autopilot for hundreds of startups - before a single sales hire. |
HubSpot acquired them for it. Now the cofounders are walking you through the exact system, live, before they disappear into product. Join the Builder Session on August 12. |
Leave with an agentic GTM stack you can replicate this week. Plus HubSpot Credits when you join HubSpot for Startups. |
Register Free Today |
|
|
|
What's new? xAI expanded Grok Imagine Video 1.5 with image and voice references, letting creators lock a character's face and voice across multiple generations, plus text-to-video creation and native 1080p output. |
What matters? |
Multi-reference support anchors up to seven separate elements - faces, products, locations - independently, so creators can swap a character into a new scene while everything else in the shot stays the same. Voice references pair a character's photo with a voice sample, keeping the same face and voice consistent across every scene in a video without rebuilding cues for each new clip. The update is live now on Grok's web, iOS, and Android apps, with image references, text-to-video, and 1080p also available through the xAI API for developers. |
Why it matters? |
Consistency has been the hardest part of AI video to fake convincingly, and locking both face and voice across a scene closes one of the last obvious tells. This is also one of the rare upgrades that's usable today instead of sitting behind a waitlist. |
|
GUIDE |
|
|
|
What's new? Autodesk launched 3D Editor + Canvas inside Flow Studio, letting creators block scenes, direct camera moves, and frame shots in a true 3D environment before AI handles the final render. |
What matters? |
The 3D Editor lets creators set blocking, camera paths, and composition directly, instead of describing a shot in a text prompt and hoping the output matches what they pictured. Canvas is a node-based 2D workspace for refining and iterating the final look with Flow Studio's AI image and video models once a scene is blocked out. Creators can combine Flow Studio-generated characters with outside AI environments and choose fast 2D generation or precise 3D direction depending on what a shot needs. |
Why it matters? |
Prompt-only AI video has always struggled to give filmmakers real directorial control over camera and blocking. Pairing a 3D workspace with AI rendering pushes Flow Studio closer to a tool studios can build previs and pitches around, not just generate demo clips. |
|
SPONSORED BY ELEVEN LABS |
The best voice models, now with full orchestration. Build real-time voice and chat agents on one low-latency stack: any LLM, your tools and knowledge, testing, Guardrails, and omnichannel deployment. |
Start building free |
|
|
|
|
What's new? OpenAI said an internal version of its next major model, Astra, resolved 10 long-standing open problems in mathematics and theoretical computer science, publishing machine-checkable proofs alongside the results. |
What matters? |
The problems spanned eight fields, including group theory, quantum complexity, and lattice cryptography, with some unsolved for decades before Astra's internal run. OpenAI published Lean 4 certificates for every result on GitHub, a machine-checkable format meant to address concerns that AI-generated proofs are hard to verify independently. The total compute cost to find the solutions came to roughly $2,000 at Sol API rates, and OpenAI hasn't said whether Astra ships as GPT-5.7, GPT-6, or another name. |
Why it matters? |
Math proofs that can be checked by machine, not just skimmed by a professor, is what turns "AI helped with research" from a claim into something verifiable. Astra's results suggest OpenAI is positioning its next model around research-grade reasoning, not just chat. |
|
Everything else in AI |
Midjourney released V8.2 as its new default model, focused on bolder aesthetics, fewer low-quality generations, and personalization that better learns a user's taste. |
ByteDance unveiled Seedream 5.0 Pro, a professional image model with region-precise editing that lets creators recolor or swap just one part of an image while leaving the rest untouched. |
Selfie Lab published a 7-step guide to Grok Imagine's multi-reference feature, showing how to build consistent AI character sheets and scenes in about 30 minutes. |
GodOfPrompt bundles prompt libraries, courses, and templates into one pack for creators looking to turn AI skills into a side income. |
|
Essential AI Guides - Reading List: |
|
|
Let us know! |
What did you think of today's email?
Before you go, please give your feedback to help us improve the content for you!
|
|
|
Work with us |
Reach 100k+ engaged Tech Professionals, Engineers, Managers and decision makers. Join brands like MorningBrew, HubSpot, Prezi, Nike, Ahref, Roku, 1440, Superhuman, and others in showcasing your product to our audience. Get in touch now → |
|
