OpenAI Unveils Game-Changing AI Features at DEVDAY
Leading artificial intelligence company OpenAI held its first DevDay conference on 6 November 2023 and launched innovative capabilities that mark a new era in artificial intelligence. As Zeo, we believe that this event will profoundly shape the future of AI assistants and developers. In this article, we will explore the enormous announcements and their impact.
The star of the event was GPT-4 Turbo, which represents the next evolution of OpenAI's main language model GPT. GPT-4 Turbo is an improved version of the original GPT-4 in several important ways:
- Massive Content-Length: GPT-4 Turbo will be able to process content lengths of up to 128,000 tokens, 16 times longer than GPT-4, which has a capacity of 8,192 tokens. This increased capacity allows it to more accurately and reliably process conversations over longer periods of time.
- Improved Functionality: Developers will now be able to run multiple functions at a single prompt and receive responses in JSON format. GPT-4 Turbo will also deliver more reliable results when executing complex logic operations.
- Repeatable Outputs: For the first time, developers will be able to achieve consistent model outputs using Data Seeding.
- More Current Information: GPT-4 Turbo has been updated with information about events up to April 2023 and will refresh its information every 6 months, avoiding outdated responses.
- Multi-Mode Models: GPT-4 Turbo will now support DALL-E rendering and state-of-the-art text-to-speech synthesis, enabling the development of new applications.
- Cost Reductions: GPT-4 Turbo provides a price reduction of more than 2.75 times compared to GPT-4. Prompt tokens will now be priced at $0.01 per 1000 tokens and completion tokens at $0.03 per 1000.
These enhancements make GPT-4 Turbo the most advanced publicly available AI system. It expands the areas where software developers can build using agents with deeper judgment and knowledge, and the greatly reduced new pricing paves the way to make this technology economically viable for other applications.
Perhaps the most important announcement was OpenAI's release of GPTs - customized variants of ChatGPT tailored to specific domains. GPTs allow anyone to create customized AI assistants for their own needs without any coding.
Users will be able to create GPTs in GPT Builder:
- Providing Clear Guidelines - Assistants will be trained on how to respond in various scenarios.
- Including Relevant Data - Documents, spreadsheets, and APIs can be added to facilitate integration.
- Enabling Actions - Through API calls and integrations, the assistant can be allowed to perform different tasks.
You can look at Zeo’s "Jarvis The SEO Geek," powered by ChatGPT here! Jarvis not only provides Zeo's extensive #SEO knowledge through technical documents and case studies but also has enhanced capabilities. It integrates APIs such as Google PageSpeed Insights and Ahrefs to offer a more comprehensive data analysis for your websites. Jarvis is now alive!
GPTs combine the three key elements of an AI agent - instructions, knowledge, and skills - into a single package. This is likely to make it possible for everyday users to turn the ChatGPT into a unique avatar tailored to their specific needs.
OpenAI also introduced a GPT marketplace where content creators can bring their own custom-made assistants to the stage. We can call it a kind of App Store. The best GPTs will be rewarded with revenue sharing, creating a new and exciting ecosystem. This important change will make personalized AI assistants available to everyone.
To empower developers, OpenAI announced new APIs that make it much easier to build advanced AI assistants:
- Persistent Topics: Automatically maintain a conversation state between messages.
- Information Storage/Recall: Add your own data sources and recall them contextually.
- Code Interpreter: Writing code with prompts and instantly establishing logic.
- Streamlined Functions: Call multiple functions in a single prompt with structured output.
These new developer kits mark an important milestone in providing wider access to the latest AI technology. They can now focus on creating delightful user experiences powered by AI, rather than spending time on the underlying infrastructure.
OpenAI emphasized during the keynote that they are following a careful strategy to develop AI systems. Initially, they are focussing on limited applications such as GPTs.
As AI capabilities continue to grow, OpenAI is focusing on systematic testing and collecting user feedback to responsibly address potential risks. Their goal is to proactively assess and mitigate hazards while giving users the power of AI. By prioritizing safety, they can provide better assistance.
OpenAI and Microsoft have strengthened their partnership to provide the infrastructure necessary to power and scale AI systems worldwide.
Microsoft is committed to providing Azure capabilities optimized for AI workloads. It is also integrating OpenAI APIs into its developer products, enabling software engineers to leverage advanced AI.
The two companies are collaborating to make AI more accessible to businesses around the world. They are doing this by industrializing AI and providing enterprise APIs.
OpenAI's DevDay conference marked a historic milestone in making AI assistants more capable, affordable, and pervasive. With new ideas like GPTs and developer APIs, customized AI will now be accessible to everyone.
Both enterprises and developers will be able to leverage OpenAI's cutting-edge AI to create the next generation of intelligent agents and experiences. I expect OpenAI's innovations to lead to new beneficial applications across industries, and their prudent iterative approach provides a model for the safe development of AI.
Overall, OpenAI has realized an exciting breakthrough in AI systems that understand, reason, and assist. The announcements made at DevDay lay the foundations for an intelligent AI tool in every home and workplace.