Big news in the AI world! OpenAI, a leader in AI technology, organized its first developer conference, DevDay 2023, in San Francisco, and let me tell you, it was more than your average tech meet-up.

The food was top-notch (not your typical conference fare), and everyone there got a surprise Oprah-style giveaway - $500 in OpenAI credits

‍

Before diving into the latest launches, let's talk numbers – because the stats shared at the start of DevDay were incredible:

• There are now over 2 million developers actively using the platform.
• 92% of Fortune 500 companies are on board, using OpenAI's tech across diverse sectors like finance, legal, and education. That's up from 80% just in August.
• And get this: ChatGPT, which has broken records as the fastest-growing consumer app ever, now has approximately 100 million weekly active users.

But let's get to the real deal. This wasn't just a party; it was a showcase of some seriously next-level AI. We're talking about new tools and tech that will change the game. Here’s a summary of everything they announced:

1. GPT-4 Turbo: the next level of AI

First up, was the unveiling of GPT-4 Turbo. Think of it as GPT-4, but on steroids. This latest AI model from OpenAI is not just an incremental update; it's a major leap forward. Here's why:

Updated context understanding: GPT-4 Turbo brings its knowledge base right up to April 2023. Remember how the earlier version, GPT-4, couldn't tell you who won the Super Bowl in February 2022? Well, GPT-4 Turbo has got that covered.

Bigger input, better answers: Ever felt limited by the 3,000-word cap when asking questions to ChatGPT? GPT-4 Turbo smashes that barrier, accepting inputs up to the length of a 300-page document. Imagine asking it to summarize an entire book for you – yep, it can handle that now.

Smarter and cheaper: Not only is GPT-4 Turbo smarter than its predecessor, but it's also more wallet-friendly. We're talking about a 2-3x cost reduction compared to GPT-4. That's a huge deal for developers and businesses looking to integrate advanced AI into their operations.

Coming soon to your screens: The best part? GPT-4 Turbo is already available in preview for developers and will be rolling out for everyone in the coming weeks. So, if you've been bummed about ChatGPT struggling with your lengthy transcripts, those days are about to end.

2. Make your own custom chatbot with GPTs:

OpenAI's second big reveal was all about letting you create your own chatbots, right within ChatGPT. This is the first step in OpenAI's vision of creating versatile AI Agents. This is like crafting your personal AI buddy with just a few clicks. (In case you forget here is a refresher on AI Agents.)

Here’s how it works:

Simple set-up: Imagine making a chatbot as easily as sending a text. That's what the new GPT builder tool does. You tell it what you want, and it not only creates your chatbot but also gives it a name and a look (with some cool art from DALL-E 3). It even comes up with questions to get the conversation going.

Make it do what you want: This isn’t just about creating any chatbot. You can make it really smart:

• Want it to read and use info from documents? You got it. No need to be a tech wizard.
• Fancy having it browse the web, write code, or create pictures? You can add those skills too.

Your personal AI assistant: What we're seeing here is the beginning of something big – making AI assistants that can do various tasks, just like a helper. And you’re the one in charge, teaching it what to do.

Customize to your heart's content: It's not just about making a chatbot; it's about making it yours. Choose its name, give it a personality, decide what it should talk about, and even what special skills it should have. It's like creating a character in a video game, but this one can actually help you out in real life.

For those who've always wanted a personalized AI helper but thought it was too complicated, this is your chance to dive in!

Here’s the official video from OpenAI.

3. Surprise appearance by Microsoft CEO Nadella

Sam Altman called Microsoft’s CEO, Satya Nadella, to the stage to discuss the partnership between their companies. Nadella emphasized that their primary goal is to build the best systems so that OpenAI can develop the best models and make them available to developers.

4. GPT Store and revenue sharing

OpenAI is rolling out something that will revolutionize how we interact with AI: the GPT Store. Think of it as an app store but for AI chatbots.

A platform for your creations: Created a cool, custom GPT? Now, you can share it with the world. The GPT Store is a place where people can download different GPTs crafted by users and developers. Imagine browsing a collection of unique AI chatbots, each with its own flavor and purpose.

Earning potential: Here's where it gets even more interesting. In the coming months, OpenAI plans to let creators earn money based on how much their GPTs are used. That's right, your AI creation could not only be popular but also profitable.

Visibility and recognition: Once your GPT hits the store, it becomes searchable. It might even climb up the leaderboards or get featured in categories like productivity, education, or just for fun. This kind of exposure can be a big deal for budding AI developers and creators.

Revenue sharing model: Sam Altman spilled some beans on the financial aspect too. Initially, creators will earn a share of the subscription revenue. The focus will be on active user engagement and potential category bonuses. There's also talk of introducing specific subscriptions for individual GPTs down the line.

5. More affordable models and an all-in-one AI experience

Major price drop: If you've been holding back on using GPT-4 or GPT-3.5 because of the cost, here's some good news. OpenAI has slashed the prices of these models by 2-3 times their previous rates. Plus, they hinted at upcoming speed improvements, which means faster, more efficient AI tools at a fraction of the cost.

Streamlined AI tools: But that's not all. OpenAI is making it easier than ever to use their AI capabilities. Until now, using ChatGPT meant juggling between different apps and websites for various AI tools. Not anymore. OpenAI announced a significant integration that brings everything into one place.

Using ChatGPT now gives you access to a suite of powerful tools:

• Image generation with DALL-E
• Browsing capabilities
• Data analysis
• Document uploads
• PDF search (a feature previously unique to Anthropic’s Claude)

This integration not only simplifies the user experience but also lowers the learning curve for newcomers.

PS. Here’s some pictures from the OpenAI DevDay watch party we hosted at the office 😁