AI News Roundup: ChatGPT-4, Midjourney V5, Microsoft's Copilot, and More!

Every other day, there is something new in the world of AI technology, so let's jump right into it.

OpenAI launched ChatGPT-4, and this version is mindblowing because it's multi-modal now, meaning it can process images and text.


During OpenAI's live stream, the presenter drew a sketch of a website on a napkin, and then ChatGPT-4 turned it into a functioning website!

ChatGPT-4 can pass a bar exam with flying colors with scores around the top 10% of test takers. You can process 25,000 words of text, it's more accurate than GPT-3.5, and you can upload a picture of a meme, and it'll tell you why it's funny.

That's ChatGPT-4 in a nutshell, and it's available to ChatGPT Plus users.

If you're not a Plus user, we have news for you. Microsoft's Bing browser was using GPT-4 the day they integrated it. And Microsoft Office will soon have GPT-4 functionalities in it too.

Microsoft is calling it "Copilot," and it may change how we use Microsoft Office forever.

Here's what Microsoft's Copilot can do. It can summarize meetings live, helps you write emails in Outlook, and can create a rough draft for you in Word.

Powerpoint and Excel will have the Copilot AI feature too. For Powerpoint, it can turn ideas into a well-designed presentation. In Excel, Copilot can analyze insights and trends or create professional-looking visualizations.

Microsoft is slowly rolling out the new Copilot feature to each of its apps, so keep an eye out for it.

And also, keep an eye out for Google bringing AI-powered features to its Workspace products.

Like Microsoft's Copilot, Google plans to integrate generative AI models into all their productivity apps like Docs, Slides, Sheets, Gmail, Meet, and more.

In Gmail or Docs, you can get writing/rewriting help on your first drafts. For Sheets, it could help you generate spreadsheet formulas.

Google says you'll be able to create text, images, audio, and videos in Slides. And capture meeting notes in Google Meets.

Google Workspace is currently being tested with trusted users, and they plan to roll out these new features in the coming weeks.

Microsoft and Google are going head-to-head with their plans of launching similar generative AI features into their lineup of apps.

We'll just have to see how both compare when they're fully available.

Last but not least is Midjourney's newest V5 model.

Midjourney V5 is much better at generating hands with five fingers instead of seven or four when compared to previous models.

It can create photorealistic images based on text more accurately than before and with intricate details, lighting, textures, facial features, and more.

Midjourney V5 isn't picture-perfect and will still make mistakes. You might get a person with floating hands or people's faces are misshapen.

But when Midjourney V5 does get the prompt right, the results are stunning.

That was our recap of all the AI-related news making headlines for March of 2023. We're sure there'll be more, so we'll catch you up in the next 3MDD.

Want to read this in Spanish? Spanish Version >>