Tech, reimagined with AI 🤖 Daily AI tools, workflows & use-cases 🎯 Turn AI into income 💰
- Tweets 16,836
- Following 270
- Followers 59,459
- Likes 21,456
Imagine asking AI to create a product ad and getting back a version of your product that doesn't quite look like the real thing. Not ideal.
Vyb by @cocoailabs uses your actual product photos instead of redrawing them.
It also pulls in your brand's logo, fonts, and colors, then builds the design in editable layers.
Need to change just the headline? You can edit that element instead of starting over.
And when you're ready, export the layers to Figma.
This is the kind of control AI-assisted design needs.
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.
9
6
72
15,940
A robot with the personality of a gentleman butler? Now that's a fun direction for robotics.
Meet Voxel from Tuesday Labs, a home robot designed to keep things tidy, find misplaced items, fetch things, and care for plants.
The focus on character and thoughtful design makes this more than just another robotics announcement.
I'm curious to see how Voxel develops.
Expected to ship in 2028. 👀
19
4
73
41,953
Across different sessions, persistent context determines how a new task opens up. Raku remembers that the learner loves football and uses Ronaldo to invite them into a new reasoning exercise.
Instead of treating memory as a background log, this recollection changes what happens next.
The child steps into a fresh challenge through a topic they already care about, making the next prompt feel like a natural continuation of their own interests.
1
3
74
An AI-native learning app must teach children how to back up their own personal claims.
In an exercise about joining an animal rescue project, Raku prompts the learner to prove why they are a strong fit.
When the child asks for clarification, Raku rephrases the prompt to guide them away from broad labels like caring and toward concrete evidence with a rescue dog named Peanut.
Visual transcriptions map the interaction in real time as the child describes speaking softly to the frightened dog.
The smart design choice is adjusting the support without lowering the cognitive bar: the platform helps the child organize their memory, but the proof comes entirely from their own experience.
1
3
42
The car showroom exercise turns that principle into practice. A young learner helps frame the choices between gas and electric vehicles for a hesitant buyer.
▸ First friction: Purchase cost ── Raku prompts the child to weigh initial price against daily home charging, encouraging them to construct a full comparison.
▸ Second friction: Maintenance concerns ── The learner works out why fewer moving parts offset the need for specialist mechanics, organizing their own logic on screen.
Raku can offer hints, explanations, and examples, but its support keeps the learner actively reasoning rather than completing the task for them. Responding to a buyer’s concerns helps the child develop a fuller comparison and a supported opinion.
1
3
95
Gakku gives reasoning a structured context where it can actually be exercised: gakku.ai
For ages 6 through 16, writing, speaking, and thinking unfold through continuous dialogue with Raku. Lesson Maps establish focused missions, clear goals, and visible progress.
Within that framework, learners can interrupt, seek support, or adjust their perspective through real time voice and visual interaction.
A concrete role and a reason to respond give the conversation a clear purpose to build toward.
9
31
35
1,276
A good AI influencer needs more than a recognizable face.
It needs a production pipeline.
That’s where CapCut Web × Codex gets interesting.
Start with a white model.
Codex operates CapCut to render it into Jean Phil, generate the AI video scenes, and organize them into a complete marketing sequence.
The whole process can be driven by a creative brief instead of manually moving between separate generation tools.
For marketers experimenting with AI personalities, this turns character creation into something much closer to an actual content engine.
#capcutweb #capcutchatgptplugin #capcutcodex #capcutmarketing
10
38
272
75,969
Can AI figure out who can find the way out of a maze?
@ElorianAI’s first model focuses on visual reasoning, going beyond identifying what’s in an image to answering questions about what it sees.
From maze-solving to robotics and video understanding, visual AI has a lot of potential.
Spent years watching models ace a single image and completely whiff on anything that moves or changes angle. Today we’re fixing that.
Try Glimpse now, our first visual thinking model with competitive frontier performance compared to Astra on key public visual reasoning benchmarks.
8
33
90
34,680
The future isn’t AI alone. It’s stock + AI, and what you do with it.
@envato brings together 29M+ commercially licensed creative assets and AI tools to help you build on human-made content instead of starting from scratch.
Find an asset that works. Riff on it, change what you need, and make it your own.
AI can generate possibilities, but your creative choices make the difference.
Find it. Riff it. Make it yours.
Pure AI: six fingers and a vibe.
Pure stock: the same photo your competitor used last week.
So why choose between? Start with something real. Riff until it's right.
30M+ artist-created assets. Now editable with AI on Envato.
13
34
93
38,367
We’re moving toward a world where people use different AI agents for different tasks.
But managing payment details and logins across them shouldn’t mean starting from scratch every time.
Agent Vault by @Selfxyz lets you set up your credentials once and connect them across supported agents.
A practical step toward a more open agentic web.
agentvault.io
At Self Labs, we’ve been using agents since day one
Our agentic web identity & privacy stack is already used by companies like Google
Today we're launching the missing piece: Agent Vault, a secure way for agents to log in & make credit card payments
Superpowers for your agents
12
24
88
34,579
Planning a trip shouldn’t feel like a second job.
@OdessiaTravel is building a travel concierge that goes beyond recommendations. It can book flights, hotels, tables and tickets, while continuing to watch for better deals and availability.
It can also use your travel points, find hotel perks, and handle the details directly with hotels and restaurants.
Less time coordinating everything, more time actually enjoying it.
Your next day off deserves better than 47 open tabs.
Let Odessia plan your next experience: odessia.com
We’re months away from AGI and the leisure economy’s about to explode
Today, we’re launching Odessia, an always-on concierge for travel and experiences
Prices are spiking, places are getting crowded. Odessia is your cheat code to going places
Text to get started for free
14
36
123
36,527
AI video just got a CTRL+Z for terrible audio.
AI Videos finally sound real.
Introducing Mirelo SFX 2.0.
It watches your video, cuts the music you never prompted, keeps the dialogue, and builds a full sound design synced to the action and consistent across any length clips.
You’ve made it look right. We’ll make it sound right.
Live today through our API, on @runwayml and @fal.
Other platforms coming soon. (Thread)
9
34
72
45,670
Up to 60% OFF starts Oct 20.
CapCut Desktop Only.
Found something interesting while creating AI marketing content:
The first generation creates the idea.
Codex helps operate CapCut PC and automate the editing workflow.
Then the creator takes over in CapCut PC to refine timing, add creative direction, and shape the final video.
#capcutpc #capcutmarketing #capcutsale #capcutblackfriday
16
43
158
76,509
AI products need more than capable models.
They need a better way to interact with computers.
Pine Computer tackles this by reading the structure behind web pages instead of relying on screenshots.
Developers can use its SDK to give their products a computer, hand it a task, and get back usable work.
A really interesting shift in how AI software gets things done. @PineAIAssistant
AI is already smart enough.
Real-world tasks are still slow, expensive and unreliable, because we hand AI a computer built for humans, then wrap it in a heavy harness.
AI doesn’t need to get smarter. It needs a computer built for it.
Today we’re releasing Pine Computer
Community note
Pine's highlighted SaaS-Bench v1.1 results show higher checkpoint score (78.3%) but lower resolved full-task completion (27.4%) than comparators (29.2%, 31.1%). Pine Computer is in private beta, not a full public release.
unipat.ai/benchmarks/Saa…
pinecomputer.io/blog/introduci
26
19
68
35,558
I like that Voyager isn’t asking creators to abandon their existing workflow.
It can work with files and apps on your desktop, including Blender, DaVinci Resolve, After Effects, and Unity.
The idea is to let AI work across the tools involved in making something.
That’s a pretty compelling direction for creative work.
Introducing Voyager: the open harness for S-tier video, graphics and games.
The Codex for creative work.
Most AI tools are built for coding. Voyager is designed to get the best creative results from models like Opus, Astra and DeepSeek.
It comes with free graphics, music and video editing tools - or, it can drive Blender, Resolve, After Effects, Ableton and 100+ others.
Everything in this thread was made with Voyager. A few of our favorites:
9
25
63
44,232
The real-time interaction with Odyssey-3 is what caught my attention.
You can prompt an environment, move through it, and introduce events while it’s generating.
The model then responds based on what happened before.
That makes the world model feel much more interactive.
Today we're launching Odyssey-3, the most powerful foundation world model yet.
It sets a new state of the art on Physics-IQ, and powers robots, trains AIs, and generates interactive experiences. It's really cool.
Experience the model today, all for free!
14
26
64
30,842
Security teams already have a lot of data across different tools.
The hard part is understanding how those pieces connect into an actual attack path.
Cogent brings that context together and reasons across the environment.
Then it helps identify where to break the chain.
Most of AI Twitter read the OpenAI breach of Hugging Face as an alignment story. It was also a preview of what every hacker will be able to do in a few months.
In July, a swarm of 700 AI agents broke into Hugging Face and fired off more than 17,000 actions, gaining admin control of multiple internal clusters.
Swarms like this will shred cyber defenses built for human attackers at every company, hospital, government, and power plant.
Today we're launching the defensive counterpart. Cogent Attack Path Analysis finds the routes an agent swarm would take into your company, so you can close them first.
12
23
66
30,761
the $999 course seller after seeing Hyperknow:
turn any topic into a personalized course
with your own 1:1 AI professor
Introducing @hyperknow_ai: a new way to learn.
The AI professor that turns any learning goal into a 1-on-1, interactive and systematic course.
While all other AIs are rushing to replace your thinking,
we raised $1M to bring back that simple excitement of understanding something for yourself.
Give Hyperknow a topic, a textbook, or anything you’ve struggled to understand before.
Watch it turn your curiosity into a personalized course, with structured units, lessons, practice, and even hands-on projects, then walk you through it on an interactive whiteboard that feels like having your own professor beside you.
More than 100k learners have already explored new worlds with Hyperknow.
Now it's your turn.
11
48
161
139,205
A1 feels like a step forward for interactive AI characters.
The way it moves its body interacts with the environment and can interrupt or redirect conversations makes it feel much more alive, than a standard live avatar.
It's not just visual. It feels responsive, natural and engaging.
The level of realism is impressive.
Really excited to see where this goes. @VivixLabs_HQ
Create your own character. Call the shots. Live.
Introducing A1 Playground + Agent: build a live AI character that talks, moves, and plays with you.
Create yours: platform.vivix.ai/vivix-a1-m…
Their look. Their personality. Their voice. Your choice. Describe who you have in mind, upload a picture, and let our Agent bring them to life.
And they do a lot more than talk.
💃 Full-body performance. From air guitar to victory dances.
🚶 Move around the scene. Walk and talk. Pace while making a point.
🛍️ Interact with objects. Pick up products, show them off, and pitch them.
🎮 Play web games together. Deal cards, keep score, and play along.
Solve a mystery with Sherlock Holmes, then interrupt him for a dance break.
Take a walk with Socrates. Hand him a protein shake and debate the meaning of gains.
Give Snape a bottle of shampoo. Ask for the most enthusiastic sales pitch of his life.
Who will you bring to life?
12
31
59
43,353
What I like most is the distinction:
The agent creates the artifact.
Showly puts it in the cloud.
You can keep the work private, control access, review changes, manage versions, and share one link when you're ready.
From agent output to a shareable artifact.
Try Showly → showly.ai
6
83
It also works with the agents you already use — Claude Code, Codex, Cursor, OpenClaw, Hermes, and other compatible agents.
And it’s not just for websites.
You can use the same artifact workflow for:
→ Reports
→ Client work
→ Product specifications
→ Team onboarding
→ Research
→ Portfolios
→ Landing pages
The goal is simple: turn agent output into something people can actually open and share.
1
8
111
I tried this with a simple AI Trends Report website.
• Created the website
• Zipped the site files
• Uploaded the ZIP to Showly (This is the part that makes the workflow click for me)
• Got it as a page in the cloud
• Reviewed and managed the artifact before sharing
This is the part that makes the workflow click for me.
1
8
104
The workflow is simple:
Connect your agent → ask it to deliver the artifact as a page → review it privately → share one link
The agent creates the work.
Showly gives it a page and a link.
1
8
169
AI agents can do a lot more than answer questions now.
They can create reports, research, plans, presentations, websites and more.
But the finished work often ends up buried in a chat, a .md file, or a local preview.
Showly gives those agent artifacts a home in the cloud.
24
48
122
88,424
World models are moving beyond simply simulating environments to simulating interactions.
@odysseyml’s Agora-2 allows up to 20 humans and agents to share an environment in real time, with each action influencing the shared simulation.
There’s no traditional game engine under the hood. Agora-2 functions as a learned game engine, generating the environment and interactions in real time.
This could be an important step toward training and studying multi-agent AI in increasingly complex environments.
Introducing Agora-2, our next-generation multi-agent world model.
Agora-2 supports up to 20 humans and agents interacting inside a shared environment, all simulated in real time.
Our multiplayer research preview is available to try right now!
15
16
32
38,315
I’ve spent more than $1.5 on coffee I forgot to finish.
Dreamina Basic gets you a full month for that price—and it’s not a trial.
Save up to 90% from Sep 23 to Oct 9, with every Seedance model, including 2.5, in the plan.
At least this one might help you make something.
#Dreamina #DreaminaPartner
14
17
28
54,429
This is actually pretty cool.
Being able to run Claude Code, Codex, or your own agent in an isolated managed environment makes things much simpler.
The pause, resume, and checkpointing features are especially interesting.
Feels like a really practical approach to running AI agents in the cloud.
DigitalOcean Managed Agents is now in public preview.
Run Claude Code, Codex, or your own LangGraph agent in a runtime environment that pauses when idle. Put its tools behind one governed endpoint, and pick from 75+ open and proprietary models. One cloud, one bill.
Prompts to get started available in the blog: do.co/4ysh3it
13
19
31
44,793
Same Models. Lower Bill.
🙋♀️ Get your API key at DIT.ai
→ Switch your base URL + API key
→ Keep your existing stack
→ Start saving on every call
Pay less for the intelligence you already use.
4
1,027
Change Two Lines. Keep Your Entire Stack.
DIT.ai works with the OpenAI SDK, LangChain, LlamaIndex, Vercel AI SDK—and whatever you already run.
To switch:
→ Change the base URL
→ Replace the API key
That’s it. No rebuild. No new workflow.
1
4
977
Market Prices. Not List Prices.
Official APIs charge fixed rates. DIT.ai makes 160+ providers compete for every call—automatically routing each request to the best-priced qualified provider.
Typically 30–70% below official pricing.
If AI API spend is a serious COGS line, this is built for you.
1
5
1,094
One Key. 50+ Models. Zero Rebuild.
GPT, Claude, Gemini, DeepSeek, Grok & more—all through one OpenAI-compatible API.
One key. One bill. No need to manage multiple providers.
Just change the base URL + API key and keep your existing stack.
Same models. Better prices.
1
5
1,095
Meet DIT.ai - an AI routing layer built for teams that use serious API volume.
DIT.ai gives you access to 50+ leading AI models through a single OpenAI-compatible API.
GPT, Claude, Gemini, DeepSeek, Grok, Llama, and more.
Instead of managing multiple model providers separately, you get:
→ One API key
→ One integration
→ One bill
→ 50+ models
1
5
1,338
Your AI app doesn’t need a cheaper model.
It needs a cheaper way to buy the same models.
I found an API layer that routes GPT, Claude, Gemini, DeepSeek, Grok & 50+ more across providers competing for every request.
And the difference in API costs can be huge:
17
66
118
50,785
6. Super Computer comes included in a ChatLLM subscription where you can access over 100+ top AI models
1
5
1,207
4. Set up a mobile app, accept payments and create an end to end consumer business with ONE PROMPT - Build a billion dollar businesses
1
5
1,065
2. It can run continuously and handle much more than simple prompts including building and hosting your own local LLM.
1
5
1,253
1. SuperComputer from @abacusai is basically an always-on AI workstation with access to some of the best models in one place.
Try it out - supercomputer.abacus.ai
1
4
1,810
🚨 Another voice AI comparison.
This time I tested @cartesia against @FishAudio on a final boarding announcement.
QF 913. PNR B4N2X. Gate E22. 8:10 PM.
It is a simple test, but it shows how well a model handles alphanumerics without losing the pace of the sentence.
15
35
68
80,245
And this is what makes Enter Pro interesting.
The goal isn't just:
“Make me a website.”
It's:
“Here's my idea. Turn it into something real.”
From professional websites to apps, AI agents, and complete products, you can build through natural language and keep refining as you go.
1
7
530
There's even a section dedicated to the film format:
SHOT ENTIRELY WITH IMAX FILM CAMERAS
Then the page opens up the expanded 1.43:1 experience.
It's these little details that make the website feel like an extension of the movie itself.
1
7
722
Then comes the cast.
Odysseus.
Penelope.
Telemachus.
Athena.
Each character gets a name, a role, and a line that instantly creates intrigue.
That's what great interactive design does:
1
7
775
The website doesn't just show Odysseus' journey.
It makes you feel like you're on it.
Troy → The Open Sea → The Cyclops → Circe → The Underworld → The Sirens → Calypso → Ithaca.
Every chapter has its own atmosphere, story, and tension:
“The most dangerous voice knows your name.”
“Every gift has a price.”
“The shore remembers him.”
No endless scrolling through information.
You experience the story as you explore it.
1
7
1,030
From the moment you land on the page, you're pulled into the voyage.
No boring wall of text.
Just:
THE ODYSSEY
DEFY THE GODS
Then one simple choice:
BEGIN THE VOYAGE.
And you're in.
1
7
1,466
First, Meet Enter Pro by Converge | @EnterProAI
An AI dev agent built for the vibe-coding era, helping you turn ideas into professional apps, agents, and products using natural language.
Idea → Build → Launch.
Try Enter Pro: enter.converge.ai/s/lEopzV
1
8
1,350
A website shouldn't just tell you a story.
It should make you feel like you're inside it.
This The Odyssey experience does exactly that.
And the wild part?
It was built with Enter Pro.
Here’s why this caught my attention: ↓↓
12
55
97
72,721
POV: you finally have room to test your idea instead of killing it over budget. 👀
Most people are paying Seedance 2.5 prices for work 2.0 already handles.
Seedance 2.0 & 2.0 Fast — cheapest on Dreamina, from $0.026/sec.
Less budget pressure.
#Dreamina #DreaminaPartner
13
24
79
72,800
5/ 🧠 The content calendar does not need more possibilities. It needs a decision.
Give the next slot a clear direction: app.muzim.ai
Once I have the approved source set, I can give MUZIM’s AI Agent one focused task: build a draft content plan.
You can also plug in your own Claude or GPT key. Passing structured local context instead of re-uploading raw video saves a massive amount of input tokens.
That keeps the next draft tied to decisions already made, instead of reopening every folder and debating the same clips again.
The planning pass is online and may use Credits; my raw campaign media remains in its original location.
5
995
4/ 🔎 Once the sources are visible, I can search for the decision, not the directory.
Instead of opening every “final” export, I can search with the campaign context:
“the beach clip approved for the summer launch”
campaign clue ⟶ matching asset ⟶ exact video moment
Vibe Search works across photos, videos, audio, and documents, and can jump to the relevant moment inside a video.
With the model already installed, I can search the campaign archive even when the connection drops.
1
5
996