😤 Our honest opinion on OpenAI Sora + alternatives you can use today

Hi AInauts,

Welcome to the latest issue of your favorite newsletter. Today with very hands-on practical tools.

Free up your weekend so you can build and experiment with AI!

That's what we've got for you today:

• 😤 Our honest opinion on OpenAI SORA

• 🎥 Alternative video models that you can use now (without subscription)

• 🤯 How to build your own voice assistant in 3 minutes

Let's get started!

😤 Our honest opinion on OpenAI Sora + alternatives

Of course, we have to start with this topic. There's no other way...

The supposedly best, most hyped AI video model has finally been released this week: OpenAI Sora

Before we look at what it can do, the question is: can we really call it a release?

It is clear that it is not possible to use it in the EU. That's not OpenAI's fault. We're used to that.

Why so strict?

Video is simply tricky. Especially for OpenAI. Sora is therefore heavily restricted: Age verification, metadata for authenticity tracking, blocks for copyrighted content and public figures, and the waiving of styles by living artists. You also need the rights to everything you upload.

That makes a lot of sense. Nevertheless, it's disappointing to see all the cool examples and then not really be able to use it, even though there's talk of a release...

Enough complaining, what can it do?

If you look strictly at the results shown so far, Sora is really impressive.

Here you can see a feed to scroll through.

Really brilliant, except perhaps for the topic of sport 🙂 …

What's also pretty cool are the editing options:

• Storyboard: Organize and edit video sequences on a timeline.

• Re-cut: Find and isolate the best frames to extend scenes (before/after).

• Remix: Modifies elements in existing videos (replace, remove, reinterpret).

• Blend: Combines two videos into one seamless clip.

• Loop: Creates seamlessly repeating videos.

We don't want to write too much. Instead, watch the following video by our favorite YouTuber MKBHD. He got Early Access, of course:

If you are one of the lucky ones and get to try it out yourself, here is the official tutorial video from OpenAI.

Quick note about the price: Currently, you can make about 50 videos of 5 seconds each with a ChatGPT Plus subscription. As a Pro subscriber (200 USD per month 🤯) you will then get almost 500 videos with higher resolution and up to 20 seconds in length.

🎥 Alternative video models you can use now (without a subscription)

We are a practice-oriented newsletter, so let's talk about video models that you can actually use today.

If you've been reading us for a while, then you probably already know about Runway or KlingAI.

Both are brilliant video AIs. The catch, as is so often the case, is that you need a subscription to really make good use of the tools.

We hate subscriptions, which brings us back to our favorite platform, Replicate.

On Replicate, you can use some really powerful video models and only pay for what you use. From as little as 30 cents per video.

You should try the new Hunyuan Video model from Tencent as well.

And our current favorite model, Minimax Video-1 (also known as Hailuo).

Simply enter the prompt and off you go. We stole the prompt from the following Sora video and entered it into Minimax.

The Sora video is slightly nicer, but Minimax sticks to the prompt much better.

🤯 Your own voice chatbot on your website, in 3 minutes

To finish off today, we've come up with what we think is a really great topic. It's a really useful AI use case that is very easy to implement.

Wouldn't it be cool to have a website with your own voice chatbot?

A chatbot that you can talk to naturally in over 29 languages, that you can feed with knowledge and that really helps you.

A chatbot that has skills like booking appointments. A chatbot that offers perfect customer service day and night? A chatbot that you can integrate anywhere online?

That's exactly what we're looking at today. The best thing is: you don't need any technical skills.

We do this with the new Conversational AI feature from ElevenLabs.

The setup is super simple:

• Select language

• Set first message

• Select language model (LLM)

You can easily give your bot knowledge in the form of documents, texts or simply URLs.

Things get really exciting when you add tools. You can easily create these using a webhook.

For example, we created a small tool in two minutes with which the bot accepts bookings for coaching calls and saves them in calendars or Google Sheets.

Completely natural, not mechanical.

Last step: Simply embed the code on your website and you're live. And all in under 10 minutes!

We made it! But no need to be sad. The AInauts will be back soon, with new stuff for you.

Reto & Fabian from the AInauts

P.S.: Follow us on social media - that motivates us to keep going 😁!
Twitter, LinkedIn, Facebook, Insta, YouTube, TikTok

Your feedback is essential for us. We read EVERY comment and feedback, just respond to this email. Tell us what was (not) good and what is interesting for YOU.