Okay, AI motion graphics are getting pretty fun. đĽ¤
This NOVA SODA concept looks like a little summer drink commercial.
Bright colors, product visuals, animated typography.
And itâs the kind of content Magnific Move is designed to create.
#MagnificMove
And hereâs the cool part.
Donât like one element?
Just ask the agent to change it.
Move preserves the rest of the composition and saves each edit as a new version.
Much easier to explore variations without starting over.
Move now includes 97 templates with their prompts.
So whether youâre experimenting with product visuals or animated typography, thereâs plenty to explore.
Take a look: be.magnific.com/4jGfusB
#MagnificMove
Qoder has a surprise for the developer community: Qoduck. Alibaba's coding agent and agent harness comes from the same company behind Qwen. Individual users can try this newly launched stealth model free for a limited time. qoder.com/
A samurai, a cowboy, a disco dancer and a werewolf.
Not exactly the lineup I expected from an AI motion graphics showcase đ
But these Magnific One + Move experiments are genuinely creative.
The different collage styles are worth checking out.
Some of the motion graphics in this launch are seriously impressive.
And apparently you can build them by describing what you want.
Magnific has been cooking with Move
Interested in exploring AI-powered coding?
Qoder is offering DeepSeek-Flash with zero Credits within the daily free allowance through Oct 31. Download the app, sign in, and select DeepSeek-Flash in the model picker.
qoder.com/qoder
Ryan Scott retweeted
Editing a skincare UGC ad and still looking for the right soundtrack? đ
The CapCut plugin is coming soon to ChatGPT.
Open CapCut Webâs Video Editor, click the plugin icon, copy the installation prompt into ChatGPT desktop on PC, and ask for upbeat or calming music that fits your video.
Finding the right vibe could get a lot easier.
CapCut World brought leading creators and OpenAI together to explore AI-powered creative workflows.
Early access is available on ChatGPT desktop (PC only).
Mobile and web support is expected in mid-November.
#CapCutWorld #ChatGPT #CapCutDidThat #CapCutWeb
Ryan Scott retweeted
One second your Tesla is driving normally. The next, itâs climbing a wall and flipping through the sky.
The transition alone could make people replay this ad. đŽđ¨đ
The CapCut plugin is coming soon to ChatGPT.
CapCut World brought leading creators and OpenAI together to explore the next chapter of AI-powered video creation.
#CapCutWorld #CapCutDidThat
Ryan Scott retweeted
14 campaigns. 30 operational problems.
And I had no idea how much time we were wasting keeping track of everything.
Payment updates buried in conversations.
Missing information scattered across different messages.
Deadlines that required constant manual follow-ups.
So I decided to give our new AI employee one responsibility:
Find out whatâs slowing down our campaigns and organize a way forward.
I didnât explain every step.
I wanted to see how heâd approach the problem.
Hereâs what Viktor delivered:
â 14 campaigns reviewed.⨠â 30 operational issues identified.⨠â 6 urgent problems prioritized.⨠â $3,050 to $4,050 in completed work without confirmed payment.⨠â A complete campaign dashboard.⨠â A 9-sheet Excel tracker.
He also prepared payment follow-ups and organized the next actions requiring approval.
But the biggest result wasnât the dashboard.
It was the time we got back.
By centralizing campaign tracking, reporting, and payment follow-ups, Viktor has already should save us over 10 hours of manual work per week.
More than 10 hours we can now spend developing new partnerships and growing the business.
And thereâs something I really appreciate.
He didnât claim the $3,050 to $4,050 had been recovered.
He flagged the missing payment confirmations and gave us a clear plan to investigate them.
Every important action still requires human approval.
This changed how I think about working with AI.
The interesting part isnât how many answers you get.
Itâs how much responsibility you can give an AI employee and what he actually delivers.
Want an AI employee working toward your teamâs goals?
Hire @viktor_com and start free.
#AIemployee
Paid Partnership
đ¤ Paid partnership
Most big models trade speed for strength. Ornith-1.5 seems to be pushing both at once, which is the more useful direction right now.
Aloha!đş Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
â
SWE-bench Verified: 0.845
â
TB2.1: 0.813
â
CyberGym: 0.800
â
NL2Repo: 0.512
Agentic
â
HLE: 0.446
â
BrowseComp: 0.785
â
MCP-Atlas: 0.746
â
Toolathlon: 0.708
đ¤Hugging Face: huggingface.co/ornith-ai/OrnâŚ
We canât wait to see what you build with Ornith-1.5-122B.
Aloha!đş Meet Ornith-1.5-122B: No Tradeoff Between Strong and Fast
Stronger. Faster. Leaner. Ornith-1.5-122B delivers frontier-level capabilities with exceptional inference efficiency. With superior instruction following, stronger task decomposition, and deeper understanding of real-world workflows, it achieves state-of-the-art performance among open-source models of comparable size, while running on a single NVIDIA DGX card!!
Open weights under the MIT license.
Coding
â
SWE-bench Verified: 0.845
â
TB2.1: 0.813
â
CyberGym: 0.800
â
NL2Repo: 0.512
Agentic
â
HLE: 0.446
â
BrowseComp: 0.785
â
MCP-Atlas: 0.746
â
Toolathlon: 0.708
đ¤Hugging Face: huggingface.co/ornith-ai/OrnâŚ
We canât wait to see what you build with Ornith-1.5-122B.
Free access to DeepSeek V4.1-Flash-1017 on Qoder won't last forever. The model is set to use 0.0x credits through October 24 before returning to 0.1x.
qoder.com/qoder
Ryan Scott retweeted
1/ A polished final paragraph doesnât tell you how well a learner can think through an idea, question it, or explain their reasoning.
@gakkuai turns writing into a real-time verbal sandbox: gakku.ai
⌠Say it messy â Talk out loud without worrying about neat outlines
⌠See it mapped â Watch raw ideas lock into visual shapes instantly
⌠Shift gears freely â Interrupt, rephrase, or flip the plot whenever inspiration hits
Raku never ghostwrites. It simply gives kids the room to figure out their own minds.
Ryan Scott retweeted
Gakku is built for kids aged 6 to 16, but its UX is a masterclass for anyone building AI products at: gakku.ai.
Instead of a blank chatbox tapping its foot waiting for a prompt, Raku talks out loud while drawing a live visual map on screen.
Kids can interrupt, blurt out messy ideas, or rethink out loud as the AI listens and adapts.
It turns âI'm still thinkingâ from an awkward pause into the main event.
Ryan Scott retweeted
âiâm still thinkingâ belongs in the lesson.
thereâs room to work it out with Raku in Gakku: gakku.ai
kids can talk before they have a tidy answer. Raku can ask a question, help organise spoken ideas into a mind map, or bring a description onto the screen.
that gives the child something to react to. a missing detail, a reason that needs work, a picture they want to change.
the Lesson Map keeps the practice focused, while the conversation leaves room for an idea to become more interesting than it was at the start.
There are plenty of AI video tools to try, but Vidu Q4 stands out for its combination of 4K
10-bit visuals, cinematic camera motion, and prompt accuracy. Find it on @Flovaai:
flova.ai/


















