From Custom GPTs to the anticipated launch of the OpenAI GPT Store, these developments are not just technological advancements; they are milestones in democratising AI and opening up a world of possibilities for developers and users alike.
In this post, we explore the recent AI changes empowering creators but also enhancing the AI experience for everyone, especially exciting for businesses and developers.
1. The arrival of Custom GPTs
OpenAI has extended the ability to build Custom GPTs to ChatGPT+ and enterprise users. This enables users to create their own Chat-GPT specific to their use cases. The best part? With no requirement of coding,
They've been able to do this by avoiding the need for a 'Vector Database' or 'Retrieval Augmented Generation' aka RAG. - solving the problem of : information/data retrieval being integrated and legible by LLMs .
In short, this makes data and information 'easy to read' by the AI and so you do not need a data scientist / developer to get started.
This is great for the following use cases:
- building Chatbots where the data you use is not private or doesn't need to be secured
- building POCs - proof of concepts before you engage with developing your own, secure proprietary technology
Examples can include:
- Creating an online 'chef' that spills generic recipes
- Creating an online 'librarian' that suggests the best books for your interest
But it is not ideal for :
- Creating and internal employee knowledge base and training centre (unless you do not mind others accessing this / the LLM learning from the information you feed it)
- Creating a way to research and inform employees about the businesses you work with
- Creating marketing strategies solely based on your own first party data
- Any information you would not want your competitors knowing about!
2. OpenAI GPT Store
OpenAI is planning to launch a GPT store, similar to an app store, where developers can share and monetise their Custom GPTs. This will create an ecosystem of AI solutions, further accelerating innovation in the field.
You can expect to see a lot more readily available tools to explore in the near future.
3. ChatGPT Turbo
LLMs (Large Language Models) like Google’s bard has been praised for providing more up-to-date information compared to ChatGPT. Google does this by continuously scraping the internet for the good, bad and ugly.
At OpenAI’s DevDay, they announced the updated version “ChatGPT Turbo” OpenAI's It incorporates information up until April 2023. This bridges the latency gap previously experience by ChatGPT 3.5.
4. SoulChat
In other LLM news, SoulChat, a new large language model designed specifically for more human-like interactions is being explored by developers in the US, Trained on data from patients and psychologists, SoulChat aims to foster deeper and more empathetic conversations with users - watch this space as it is not released yet!
5. DALL-E-3 and Copywriter Sheild
More updates from OpenAI, includes their steps to make DALL-E-3, its powerful image generation tool, more accessible to developers by reducing its computational costs. This will encourage more developers and businesses to explore the creative possibilities of DALL-E-3.
OpenAI is also offering Copywriter Shield, a service that provides financial protection to users in case they encounter legal or reputational issues arising from the use of their tools. This demonstrates OpenAI's commitment to responsible AI development.
The developments that continue to emerge are so exciting, though they need clarification before you engage with them and share information you thought would stay private.
