Exploring the future of AI , AI Tools Tutorials • Helping you work smarter with AI Learn. Create. Automate. DM for paid collab 📧[email protected]
united states
Joined January 2023
- Tweets45.2K
- Following8.2K
- Followers18.3K
- Likes102K
UMER AI retweeted
Drop 1/15: Introducing Joy, a dual-channel conversational dataset.
Real, unscripted conversations between two people, each on their own track. Built to teach voice models how people actually talk: when to listen, when to jump in, when to say “hmm”.
100,000 hours.
50,000 speakers.
10 languages.
The flat speaker distribution is the detail I'd point to if someone asked whether the data was collected carefully.
One year for you, one year to gift. Your choice. flova.ai/?refCode=3BFU8P57
Creative block? Sometimes you don’t need another idea—you need a place to start. ✨
Dreamina’s first month is $0.29, and from Oct 20–30, CapCut Desktop is 60% off Pro or 40% off Ultra. Each product has its own subscription and pricing. Use Dreamina to generate visuals that get your concept moving, then bring them into CapCut PC to trim the clips, adjust the rhythm, add music and transitions, and polish the final details. Start with something, then make it yours.
🔗 CapCut offer: tinyurl.com/2v23rn22
🔗 Dreamina: tinyurl.com/5xbtv4x6
#CapCutDidThat #Dreamina #Smallbusinessmarketing #CapCutSale
UMER AI retweeted
More and more, image models get ranked by human vote. A vote can miss that your font changed.
Today @DavidZats and I are releasing VybBench, an open benchmark that scores design fidelity with numbers instead of votes. Ask a model to change one headline and it can nudge your font, shift your blue and redraw a corner of your logo too. Each edit changes a little more, until the design isn't yours.
VybBench scores each piece:
145 fonts, matched pixel by pixel against the real font
100 hex codes, scored on how far each color drifts
200 small edits, scored on how many pixels changed, with a judge checking the edit was made
We are not including model scores in the post, but you can see how each model scores on the Hugging Face. Every prompt, output, judge verdict and the scorer is on Hugging Face, link in the reply.
Full Disclosure: At @cocoailabs we also build Vyb - our first design model built to preserve design fidelity and taste.
UMER AI retweeted
I hired a retention researcher and gave him 12 brands to look at.
One request: please come back with something better than "send them 15% off."
His name is Viktor. He's an AI employee, and I gave him the assignment in Slack.
I wanted to understand what brands were doing to give customers a reason to buy again. Refill flexibility, product education, guarantees, something beyond another coupon with a countdown timer.
He came back in about 10 minutes with a Second Order Library and three experiments to consider. Saved 10+ hours of doing the same research myself.
• A replenishment reminder based on when someone should actually run out.
• A guarantee that makes trying a second product feel less risky.
• A specific gift on order two, introduced while the customer is still excited about order one.
Each proposal included who should get it, what to compare it against, and what to measure, with margin called out where it could bite. We do enjoy keeping some of the money.
The part I appreciated: he separated what he found on brand websites from what he thought we should test. He also flagged what he couldn't verify.
Seeing another brand do something doesn't mean it works. It gives you a hypothesis.
Now I have three test plans to evaluate and the research behind them, in one place.
Not bad for a morning's work.
If you have a research backlog of your own, hire @viktor_com for your team. Free to start, no credit card.
#AIemployee #Retention
Paid Partnership
🤝 Paid partnership
I like the guarantee idea. If someone enjoyed the first product but feels unsure about trying something new, removing a little of that risk could make a real difference.
I hired a retention researcher and gave him 12 brands to look at.
One request: please come back with something better than "send them 15% off."
His name is Viktor. He's an AI employee, and I gave him the assignment in Slack.
I wanted to understand what brands were doing to give customers a reason to buy again. Refill flexibility, product education, guarantees, something beyond another coupon with a countdown timer.
He came back in about 10 minutes with a Second Order Library and three experiments to consider. Saved 10+ hours of doing the same research myself.
• A replenishment reminder based on when someone should actually run out.
• A guarantee that makes trying a second product feel less risky.
• A specific gift on order two, introduced while the customer is still excited about order one.
Each proposal included who should get it, what to compare it against, and what to measure, with margin called out where it could bite. We do enjoy keeping some of the money.
The part I appreciated: he separated what he found on brand websites from what he thought we should test. He also flagged what he couldn't verify.
Seeing another brand do something doesn't mean it works. It gives you a hypothesis.
Now I have three test plans to evaluate and the research behind them, in one place.
Not bad for a morning's work.
If you have a research backlog of your own, hire @viktor_com for your team. Free to start, no credit card.
#AIemployee #Retention
Paid Partnership
UMER AI retweeted
🎃 YOUR AI. YOUR NIGHTMARE. YOUR CHANCE TO WIN!
💰 $1,000 in cash prizes
🎁 24,000+ AI credits
🏆 20+ winners
📅 Oct 26 – Nov 3 UTC
Join the EaseMate AI Halloween Challenge!👻(easemate.ai/halloween-ai-cre…)
#HalloweenChallenge #AIArt #EaseMateAI
CapCut PC Black Friday is on! Get 60% off Pro and 40% off Ultra through Oct 30, Desktop only. 🕸️
My brain at 2 AM: “What if the pumpkin turns into a portal?” My brain at 2:01 AM: “You have to edit that before morning.” 💀
tinyurl.com/2v23rn22
#capcutpc #capcutmarketing #capcutsale #capcutblackfriday
AI video’s next milestone may be cultural as much as technical: when does a filmmaker see a reason to use it? The next evidence to watch is how it performs in actual workflows.
Go from a prompt to a promo with a combo starting at $6.29.
Dreamina helps generate your visuals. CapCut PC helps shape them into a finished marketing video.
Try the combo while it's live.
First-month combo · Oct 20–30 · CapCut Desktop only.
#CapCutDidThat #Dreamina #Smallbusinessmarketing #CapCutSale
UMER AI retweeted
I launched the first image and video generation models at Google, and kept asking why all AI designs look the same.
AI has brought us the WORST era of design.
It made content creation effortless, but it also made 90% of it look painfully generic.
My co-founder and I spent 2 years building in stealth to change that.
Meet Vyb, the first AI design model that understands your taste like a designer would.
RT + comment your website URL and we’ll make your first design.
AI tools are much more useful when people can adapt them to their own needs. open weights and fine-tuning options make this worth exploring.
Aloha!🌺 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
✅SWE-bench Verified: 0.845
✅TB2.1: 0.813
✅CyberGym: 0.800
✅NL2Repo: 0.512
Agentic
✅HLE: 0.446
✅BrowseComp: 0.785
✅MCP-Atlas: 0.746
✅Toolathlon: 0.708
🤗Hugging Face: huggingface.co/ornith-ai/Orn…
We can’t wait to see what you build with Ornith-1.5-122B.
From AI images to video generation, Flova's current savings cover multiple creative workflows.
Five days remain before the October 20 deadline.
@Flovaai
flova.ai/?refCode=UPYRYPR3
UMER AI retweeted
Inworld is your new favorite inference provider.
We already process tens of trillions of tokens a month because we understand consumer AI different, and we've built our inference from the ground up for it.
Today, the inference top consumer AI apps run on is opening up.
The bigger idea here is that inference is becoming part of product design, not just backend infrastructure.
When response time, context reuse, reliability, and price shape the user experience, the serving layer deserves as much attention as the model itself.
