OpenAI’s DevDay Reveals a Future Shaped by Advanced AI: Unveiling GPT-4 Turbo and More

By - November 7, 2023

OpenAI, the innovative force behind ChatGPT, celebrated a significant milestone on November 6 by hosting its inaugural DevDay, headlined by OpenAI CEO Sam Altman. The developer-centric event saw a series of groundbreaking announcements. Here’s the rundown of major updates from the keynote address.

Supercharged Capabilities: Introducing GPT-4 Turbo

The highlight of the day was the unveiling of GPT-4 Turbo, a formidable upgrade to the already impressive GPT-4.0. This new iteration boasts a staggering context window of 128K tokens, translating to about 96,000 words or a 300-page discussion, dwarfing its predecessor’s 50-page capacity. Imagine summarizing ‘The Great Gatsby’ in one go—that’s the prowess of GPT-4 Turbo. This new context window also leapfrogs Anthropic’s 100K limit in its Claude Chatbot.

What’s more, GPT-4 Turbo’s knowledge is as current as it gets, with a cutoff in April 2023. This means it’s equipped with the most up-to-date information, a leap from the September 2021 cutoff of the previous model. The implications for app developers are profound, with token costs significantly reduced by 3x for inputs and 2x for outputs, ensuring a more cost-effective deployment of AI models.

A Unifying Update: ChatGPT All Tools Feature

OpenAI has also announced an “All Tools” update to the ChatGPT chatbot, enabling users to leverage all the latest features concurrently. This includes integration with Bing for browsing, Advanced Data Analysis, Plugins, and the cutting-edge DALL·E 3, paving the way for a seamless multimodal user experience that outpaces rivals like Bing Chat and Google Bard.

Tailoring AI: The Launch of GPTs

In a bold move, OpenAI announced “GPTs,” a feature that empowers users to craft bespoke versions of the AI assistant for diverse roles. The forthcoming “GPT Store” promises to be a hub where users can publish their custom GPTs, with a revenue-sharing model to boot.

Reinforcing Creativity: Copyright Shield

A major concern in the AI domain has been addressed with the introduction of “Copyright Shield.” This initiative promises legal cover for copyright to enterprise customers, safeguarding their use of OpenAI’s software for specialized tasks. Additionally, the coverage extends to OpenAI’s API, granting software developers the freedom to innovate with AI models like GPT-4, worry-free.

Exclusive Offerings: Custom Models Program

For those in need of specialized solutions, OpenAI has introduced a program that collaborates with companies holding proprietary datasets to develop custom models. At its inception, this will be a limited (and expensive) proposition to work directly with OpenAI engineers to create custom models.

OpenAI’s DevDay has laid out a vision for an AI-driven future that is more intelligent, customizable, and secure.

Get more insights from RSM
Generative AI is rapidly changing how companies of all sizes do business. Get insights to capitalize on opportunities while addressing risks.

Diego Rosenfeld is a principal in RSM’s Boston office, serving as the national go-to-market leader for managed IT services (MITS) and a member of the RSM managed technology services leadership team. As go-to-market leader, Diego oversees MITS product strategy and regional and market-based client engagement teams. He works hand in hand with RSM industry teams to develop managed services that integrate our rich capabilities into scalable, industry-relevant offerings.

Receive Posts by Email

Subscribe and receive notifications of new posts by email.