Release of the year: 4 new use cases of GPT-4

Release of the year: 4 new use cases of GPT-4

What's different, improved, & how the pros are using GPT-4 in the first week.


Written by Mirna, Ljubljana, 2023

In case you missed the release of the year - OpenAI’s GPT-4 is here to change the game.

One tool has it all, or as they call it - a large multimodal model that exhibits human-level performance on various professional and academic benchmarks. Tho, still less capable than humans in many real-world scenarios … 4 now.

Here are 4 significant improvements of Chat’s V4:

  • Image generation: forget about using 4 other apps besides Chat. In GPT-4, you could ask it to generate an image mid-conversation, and it can also understand images. Use case: GPT-4 can create promo image examples when planning a marketing campaign.

  • Understanding images: it can use pictures + text as prompts & understand the connection. Use case: if you provide a picture of tomatoes, cheese & dough on the table and ask, “what to make 4 dinner with those ingredients?” it will suggest a recipe option from your list. Hopefully, a pizza, but we still have to try this one …

  • Enabling audio inputs: provided with an audio clip of people talking, background radio, spoons hitting the cups, and coffee machine rumbling: GPT-4 runs a model asking itself, “Where am I?" and can recognize a coffee shop. Use case: you could ask about a transcript of a YouTube video, which makes a lot of jobs easier, but GPT-4 would also be able to understand the concept behind it, t, giving you information about how it was created, the dialogue, and much, much more.

  • Speed: GPT-4 is twice as faster as GPT-3, with an intermediate time of 33 tokens per second, which is understandably dependent on the complexity of your prompt. GPT tokens represent generated sub-words, which split the sentences. For example, the sentence "I like to eat pizza" would be split into individual tokens, such as "I", "like", "to", "eat", and "pizza".

As we speak about speed, we are still patiently spending our time on the waitlist. And while we wait ...

Screenshot 2023-03-16 at 10.37.32.png

Here are 4 more examples of what other users are already creating with GPT-4:

  1. Building a business: Jackson Fall set GPT-4 a $100 budget and instructed it to generate revenue. He's doing his part as its human intermediary. GPT-4 suggested building an affiliate marketing website focusing on sustainable living and eco-friendly items. So far, they have created GreenGadgetGuru and landed some investments.

  2. Dating: what? Well, yes, the online AI dating service Keeper uses GPT-4 to analyze users' profile information and preferences when someone appears to be a suitable match for you, ensuring that you, ideally, receive the matches you seek before even talking to them.

  3. Helping out: Be My Eyes created a wonderful social initiative powered by GPT-4. Their Virtual Volunteer will answer any question about an image and provide immediate visual assistance to visually impaired people in real-time within the app.

  4. Fighting for justice: DoNotPay - the World’s First Robot Lawyer, is leveraging GPT-4 to create "one-click lawsuits" to sue scammy robot callers for $1,500. Imagine getting a scam call, pressing a button, and having a 1,000-word lawsuit initiated 4 you. GPT-3.5 was insufficient. However, GPT-4 does a fantastic job of handling this task.

Text, audio, video, and other multimodal input will fundamentally alter how we engage with these models. If you thought that the hype around the ChatGPT wasn't worth it, you would be in for a big surprise!

Stay around on fresh OpenAI updates, and use cases on our ongoing projects. Subscribe to the monthly newsletter below.

Stay in the loop


Welcome to our email list

We promise we won't spam you!

Mirna Germšek