How 2 AI

How-to guides on AI. We explore and write about the latest AI tools daily. • DM if you want us to review your product.

[How 2 AI @How2use_AI]

Claude becomes 10x more powerful when you use the right prompts. Here are 10 prompts worth bookmarking.👇

[Bishal Nandi @LearnWithBishal]

I thought Dreamina's $1.50 Basic offer was gone, but it's somehow still live. 👀

The monthly allowance can cover up to 9 Seedance videos or 393 GPT Image 2.5 images, so this feels less like a trial and more like actual room to create.

https://t.co/4pJo5zjpiL

#Dreamina #DreaminaPartner

[Shane Hegde @shanehegde]

https://t.co/rwtBcUvnIY

[Bishal Nandi @LearnWithBishal]

The shift from software built for humans to software built for AI agents is going to be huge.

What stands out to me about https://t.co/nZiq1MAtss and Proto is the idea of describing what you need, letting AI build the app, and then having an agent actually do the work.

That’s a very different way of thinking about business software. Excited to see where this goes.

[How 2 AI @How2use_AI]

@alexdanilowicz Going from a PRD to a working prototype this quickly is pretty impressive.

[Alex Danilowicz @alexdanilowicz]

Ok, launch went WAY BETTER than expected.

Here's a walkthrough on how to use MagicOS:

If I were on a product team at YouTube shipping a new feature (say 'creator plus') here's how I'd design it in 5 mins:

1. Drop in a PRD, a screenshot, and existing product states
2. Tell MagicOS to build a modal referencing the PRD
3. Fire off 2 more threads in parallel
4. Ask it to audit the modal against the PRD
5. Publish the final version as a live interactive url
6. Check different states and share with stakeholders

This is how most teams use @magicpatterns today.
https://t.co/rjSujam0l1

[How 2 AI @How2use_AI]

This is where AI design tools get really useful.

Not just generating screens, but helping product teams build, test, and iterate on ideas much faster.

[Alex Danilowicz @alexdanilowicz]

Ok, launch went WAY BETTER than expected.

Here's a walkthrough on how to use MagicOS:

If I were on a product team at YouTube shipping a new feature (say 'creator plus') here's how I'd design it in 5 mins:

1. Drop in a PRD, a screenshot, and existing product states
2. Tell MagicOS to build a modal referencing the PRD
3. Fire off 2 more threads in parallel
4. Ask it to audit the modal against the PRD
5. Publish the final version as a live interactive url
6. Check different states and share with stakeholders

This is how most teams use @magicpatterns today.
https://t.co/rjSujam0l1

[How 2 AI @How2use_AI]

@alexdanilowicz Keeping the entire design process in one place makes so much sense.

[How 2 AI @How2use_AI]

The future of AI design isn’t about replacing designers.

It’s about giving them better tools to think, experiment, and bring ideas to life.

Keeping the entire process on one canvas makes a lot of sense.

[Alex Danilowicz @alexdanilowicz]

Today we're launching MagicOS. And we've come to the conclusion that designers are NOT getting replaced.

most of the industry moved to the same answer this year: give the agent its own canvas, let it run, hand you the output.

we made a different bet: the canvas should be yours and the agent is your collaborator. that’s MagicOS.

most AI design tools are a linear conversation: prompt, generate, iterate, repeat. a hundred prompts in, you might have a great final design, but the reasoning, rejected directions, and inspiration that got you there are buried in a chat log nobody will ever scroll back through.

MagicOS is a persistent, spatial canvas. Your PRD, inspiration images, design system, variations, rejected directions, final picks - all of it lives together, all in one place, instead of stacked on top of itself and lost.

it’s voice-first. talk to it in real time and watch work land directly on the canvas as you speak.

it’s also multiplayer. your whole team works on the same canvas together.

and it still runs on your design system and components because none of this matters if the output doesn’t look like your actual product.

three months ago, this was internal only at Magic Patterns, my co-founder's weekend project. it’s now live for everyone.

Design everything, everywhere, all at once.

[Alex Danilowicz @alexdanilowicz]

Today we're launching MagicOS. And we've come to the conclusion that designers are NOT getting replaced.

most of the industry moved to the same answer this year: give the agent its own canvas, let it run, hand you the output.

we made a different bet: the canvas should be yours and the agent is your collaborator. that’s MagicOS.

most AI design tools are a linear conversation: prompt, generate, iterate, repeat. a hundred prompts in, you might have a great final design, but the reasoning, rejected directions, and inspiration that got you there are buried in a chat log nobody will ever scroll back through.

MagicOS is a persistent, spatial canvas. Your PRD, inspiration images, design system, variations, rejected directions, final picks - all of it lives together, all in one place, instead of stacked on top of itself and lost.

it’s voice-first. talk to it in real time and watch work land directly on the canvas as you speak.

it’s also multiplayer. your whole team works on the same canvas together.

and it still runs on your design system and components because none of this matters if the output doesn’t look like your actual product.

three months ago, this was internal only at Magic Patterns, my co-founder's weekend project. it’s now live for everyone.

Design everything, everywhere, all at once.

[How 2 AI @How2use_AI]

@alexdanilowicz @magicpatterns Now I'm curious to see the final video.

[Alex Danilowicz @alexdanilowicz]

I'm so tired of all these AI-generated launch videos. So I called up a friend in LA and went to hollywood to shoot a video for our big launch tomorrow.

we're announcing something that will be a major step forward for @magicpatterns and I'm excited that the launch video is 100% human.

don't go by the BTS (our marketing team was keen to pull footage of me having some fun). the final output is... well, you should see it for yourself tomorrow!

big shoutout to our team.

[How 2 AI @How2use_AI]

@LearnWithBishal This opens up so many creative possibilities!

[Bishal Nandi @LearnWithBishal]

Imagine using the same AI character across a product launch, a funny promotional Reel, and a short brand story.

That's where character consistency becomes useful beyond the initial wow factor.

With GPT Image 2.5, now available through CapCut's Design Studio and AI Image, I can generate four views of a character before moving into AI Video.

From there, I can explore different commercial concepts: a product showcase, a creator-style recommendation, or a recreation of a trending marketing video format.

And the work doesn't end with generation. CapCut PC's multi-track timeline lets me arrange footage, refine transitions, adjust pacing, and add the finishing touches that make each version fit its purpose.

One character can be the starting point for multiple creative directions.

For brands and creators who need fresh social content, that's an interesting workflow to explore.

#GPTImage25 #CapCutPC #AIVideo #Seedance25 #capcutmarketing

[Gunhild Johanne Reumert | AI Tools & News @shedoesai]

I didn't think AI could onboard a person. It took 4 minutes to prove me wrong.

12 tools, 100+ Slack threads, dozens of docs. Mapped before her first day.

I'm Gunhild. I run ShedoesAI, a creator business with a small team.

Onboarding used to mean long meetings and endless questions. We tried AI as an assistant first. Useful, but still one question at a time.

So we gave Viktor the whole job: map our company, people, tools, workflows and priorities, then build an onboarding briefing and a 30-day plan.

He lives in our Slack and works across 3,200+ tools. Four minutes later the plan was ready. My assistant reviewed it and added what only the team knows.

The result: 7 hours of meetings saved in the first week and her first task shipped.

Chatbots answer. Copilots assist. Employees own outcomes.

Hire @viktor_com for your team. $100 in credits included, no card. Full link in first comment.

#AIemployee

[Bishal Nandi @LearnWithBishal]

Finding jobs is easy. Finding the right ones at the right time is the hard part. This is interesting. 👀

[Bishal Nandi @LearnWithBishal]

I’ve spent years making fake commercials for products that do not exist. Thrilled this has finally become an activity with a $100k prize pool.

#TheWorldsCatchiestAd #ElevenCreativeCompetitionVol1

[How 2 AI @How2use_AI]

The internet built trust layers for humans.

The agent economy will need trust layers for AI.

Identity, permissions, and verification are about to become core infrastructure.

[Isha Bhatnagar @Isha]

An AI agent can buy something in seconds.

Can the business receiving that order tell who sent it, what it can do, or whether to let it through?

We built @beltichq to answer that. We're out of stealth with $8.8M raised to make agent traffic verifiable.

[Bishal Nandi @LearnWithBishal]

🚨 Web Scraping is USELESS without this !!

You can write PERFECT web-scraping code...

But one CAPTCHA (or) rate-limiting... can BREAK your entire workflow 💀

I just found Decodo's Web Scraping API...

And this is what it can do 👇

[Isha Bhatnagar @Isha]

An AI agent can buy something in seconds.

Can the business receiving that order tell who sent it, what it can do, or whether to let it through?

We built @beltichq to answer that. We're out of stealth with $8.8M raised to make agent traffic verifiable.

[Bishal Nandi @LearnWithBishal]

Every viral clip has a motion recipe. Dreamina's Extract Motion function lets you pull it back out.

Prompt tools hand you a clip you can't control. A canvas hands you a project you can.

Dreamina 2.0 is canvas-first - the new way AI video is made.

#Dreamina #DreaminaPartner