Customizable ChatGPTs – Everyone Can Easily Build ChatGPT
OpenAI DevDay event was anything but boring. Lots of exciting updates and capabilities that you need to know about. So let’s get started!
Let’s begin with one of the significant things – ‘GPTs’, a new feature that allows creating personalized versions of ChatGPT that incorporate instructions, additional knowledge base, and various skill combinations.
GPTs enable everyone to create personalized versions of ChatGPT tailored to their needs and work style. How to create one? Easily and without a single line of code. These GPTs can be created with text instructions – prompts – for personal or professional use.
Some examples of personalized GPTs include:
- Writing coach
- Startup advisor
- Canva GPT for designing
- Zapier AI Actions GPT for automating tasks
And to see how it works in action, here’s a video showing OpenAI CEO, Sam Altman, building a custom GPT in just 4 minutes:
https://twitter.com/tmiyatake1/status/1721661275011506179
Got an interesting GPT idea? You can share it soon and earn money through the GPT store coming later this month.
Think of it as the App Store for AI, but now everyone can build and profit from an application without needing coding skills. This is a true game-changer that can completely open up a new GPT economy with millions of people building, sharing, and generating profits from their creations. A bit like how social networks opened the door for everyone to be a content creator.
Organizations can also allow employees within the company to design internal GPTs without code and securely share them across the organization to improve efficiency.
The GPT feature will be launched for Plus and Enterprise users in the coming days.
While personalized GPTs was the biggest announcement from yesterday’s event, OpenAI also announced a long list of features and new upgrades for ChatGPT and GPT-4. Here’s a brief summary:
GPT-4 Turbo – The Next Generation of GPT-4
The Turbo brings with it new and essential capabilities:
- 128k meaning prompts of book length can be uploaded
- Updated on global events until April 2023
- Enhanced ability to follow improved instructions
Assistants API
It will allow developers to build agent-like experiences within their applications such as: a language-based data analysis app, a coding assistant, a vacation planner operated by AI, and more.
Vision API, Dall-E 3, and Text-to-Speech API
GPT-4 Turbo can receive images as part of the guidance. For example: detailed image analysis and reading documents with integrated images.
The Image API will enable developers to create images and designs in their products programmatically.
The new Text-to-Speech API will allow developers to create human-quality speech from text.
Lower Prices and Higher Usage Limits
Usage fees for GPT-4 Turbo will now be 3 times cheaper than GPT-4 at $0.01.
GPT-3.5 Turbo will also receive a similarly lower pricing.
OpenAI is also doubling the minute limit for paying GPT-4 customers.
Copyright Protection
OpenAI updated that it will protect customers and cover the costs if they face copyright infringement lawsuits.
This will apply to features generally available in ChatGPT Enterprise and the developer platform.
For the full list of new editions, including Whisper v3 and Consistency Decoder, and more information, visit the post on OpenAI’s official blog.
