TikTok Owner’s New AI Tool Creates Lifelike Videos From Just One Photo

Want to advertise? Click Here | Subscribe to our newsletter

Hey everyone,

Just a quick heads-up: Long gone is the story that TikTok is getting banned. ByteDance, the parent company behind TikTok, released a new tool that takes deepfake to the next level

What’s inside today’s newsletter:

• 🌐 Tech Pulse: Researchers created an open rival to OpenAI’s o1 ‘reasoning’ model for under $50

• 📁 Unfold AI Tricks Automate Summarizing Text Using AI in Google Docs

• 5 New Tools on the Block

• 🔎 Spotlight: TikTok Owner’s New AI Tool Creates Lifelike Videos From Just One Photo

• 🤝 Who is getting funded?

• ✅ To-dos

🌐 Tech Pulse: What’s happening now?

• Researchers created an open rival to OpenAI’s o1 ‘reasoning’ model for under $50

• Will AI revolutionize drug development? Researchers say it depends on how it’s used

• Hugging Face researchers aim to build an ‘open’ version of OpenAI’s deep research tool

• AMD pulls up the release of its next-gen data center GPUs

• Alphabet praises DeepSeek, but it’s massively ramping up its AI spending

📁 Unfold AI Tricks: Automate Summarizing Text Using AI in Google Docs

Overview: Need to quickly summarize text in Google Docs without copying it into another AI tool? Using Google Apps Script, you can automatically generate text summaries directly inside your documents using Google’s built-in AI-powered natural language processing.

Duration: 15-30 minutes

Skill Level: Beginner to Intermediate

Set Up Your Google Doc: Create a Google Doc with a section of text you’d like summarized.

Write the Script: Go to Extensions > Apps Script and enter this code to create an automatic text summarizer:

function summarizeText() {
  var doc = DocumentApp.getActiveDocument();
  var body = doc.getBody();
  var text = body.getText();

  // Make a request to Google's NLP API
  var apiKey = "YOUR_API_KEY"; // Replace with your Google Cloud API key
  var url = "https://language.googleapis.com/v1/documents:analyzeEntities?key=" + apiKey;
  
  var payload = {
    "document": {
      "type": "PLAIN_TEXT",
      "content": text
    },
    "encodingType": "UTF8"
  };
  
  var options = {
    "method": "post",
    "contentType": "application/json",
    "payload": JSON.stringify(payload)
  };
  
  var response = UrlFetchApp.fetch(url, options);
  var result = JSON.parse(response.getContentText());
  
  var summary = result.entities.map(e => e.name).join(", ");
  
  body.appendParagraph("\n\n**Summary:** " + summary);
}

Automate: Set a trigger to summarize new text automatically when your document is updated.

Customize: You can modify the script to highlight specific keywords or even perform sentiment analysis on the text.

Pro Tip: Combine this with Google Sheets by creating a batch summarizer that processes multiple texts and exports the summaries into Google Sheets for easy reference!

💼 From our Partners

🔧 New Tools on the Block

1. Airtop Auth: Automate web tasks on authenticated sites

2. a0.dev: Generate full mobile apps with AI

3. Screen Studio 3.0: Beautiful screen recordings with instant shareable links

4. Concierge AI: Talk to your apps with natural language.

5. Lyzr Agent Studio: Design, build, and deploy AI agents effortlessly

Want to sponsor your tool in our newsletter? CLICK HERE

🔎 Spotlight: TikTok Owner’s New AI Tool Creates Lifelike Videos From Just One Photo

OmniHuman-1 by ByteDance, the company behind TikTok, found a cool new way to create human videos from just a single image and some motion signals like audio or video.

It's an end-to-end framework that uses a mixed training strategy to make the most of available data.

And I’m blown away by the results I saw on their GitHub pages promises a lot

Here are some of the exciting things OmniHuman-1 can do:

• Generates realistic human videos using weak signal inputs, especially audio.

• Supports various visual and audio styles.

• Handles different aspect ratios and body proportions (portrait, half-body, full-body).

• Greatly improves gesture handling.

• Supports diverse inputs like cartoons, artificial objects, animals, and challenging poses.

• Accommodates various music styles and body poses for singing.

• Mimics video actions and controls specific body parts by video driving or combined audio and video driving.

Keep in mind that the images and audio used in the demos are from public sources or generated by models and are solely for demonstrating the capabilities of this research. If you're interested in citing OmniHuman-1 for your research, you can use the provided BibTeX entry.

Important note: Currently, there are no services or downloads available, so be careful of any fraudulent information.

This puts the situation for the future that Deepfakes will be much more easily accessible, and who knows at what length consequences can go?

🤝 Who is getting funded?

1. Krafton leads Cashfree’s $53M funding at $700M valuation (TechCrunch)

2. Hitachi Ventures raises $400M fund to invest in everything from fusion to AI (TechCrunch)

3. Okta competitor SailPoint races toward $11.5B IPO (TechCrunch)

4. Deel unloads $300M in secondary sale, brings General Catalyst on as an investor (TechCrunch)

 😂 Tech Memes

✅ To-dos

Do you have a topic in mind? We’re always open to suggestions—let us know what you want to learn next!

If you like today’s issue, consider subscribing to us.

That’s a wrap! Catch you in tomorrow’s edition. 👋

—Harman