CONVO OVER COFFEE ☕

Big topics to discuss with friends and colleagues

OpenAI's DevDay and whole bunch of news 🚩
OpenAI's recent developer conference, their first ever, marked a significant moment in the tech industry, showcasing the burgeoning impact of AI. Sam Altman, CEO of OpenAI, delivered a keynote that reflected the shift in tech's epicenter from smartphones to AI.
1️⃣ Here comes GPT-4 turbo:
- 128k context length!
- multi-modal with DallE-3, Vision, TTS all in one place in the API
- improved function calling and better instruction-following
- can do RAG on external knowledge bases within the platform (wow!)
- knowledge cutoff: Apr 2023
- able to build customized model with fine-tuning for GPT-4
- higher rate limits (2x tokens per minute)
- 3x cheaper for prompt tokens
- 2x cheaper for completion tokens

2️⃣ Not good enough? Let us build our own GPT then!
- do it using natural language (no-code GPTs!)
- publish make it available for others to use
- share revenue with OpenAI and developers
- GPT app store coming soon

3️⃣ It only talks? Nah, nah, my friends. OpenAI has introduced the "Assistants APIs," simplifying the creation of AI agents. These APIs include features like persistent threads for retaining conversation history, built-in retrieval, code interpretation, function calling, and integration with voice and Text-to-Speech (TTS) capabilities. They are also connected to the internet.

My take: I can imagine many, many founders now having super hard time. We all have been there. My first product I was to release was killed by OpenAI 2 months into work. Just iterate if you can. If not - this is life. Sorry - AI life. Ruthless innovation.

Google Amplifies AI Partnership with Anthropic for Enhanced Safety and Efficiency 🚩
Google has announced the expansion of its partnership with Anthropic, focusing on AI safety and security. This venture involves the use of Google's latest Cloud TPU v5e chips for AI inference, marking a significant step in AI technology and safety. The collaboration, deep-rooted since Anthropic's founding, leverages Google's cutting-edge computing and database solutions, underscoring their commitment to advancing AI responsibly.
My take: This strategic move by Google and Anthropic reflects a growing trend in the tech industry towards partnerships of giants for AI development. By prioritizing robust computational infrastructure resources, they're setting a standard for the industry, balancing innovation computing considerations.

Microsoft's GitHub Copilot Surpasses 1 Million Paying Users 🟠
Microsoft CEO Satya Nadella announced significant growth in the generative AI sector, with over a million paid users for GitHub Copilot, marking a 40% increase in the September quarter. This achievement aligns with Nadella's strategy to integrate AI across Microsoft's product lines, including Bing's integration with OpenAI's ChatGPT, resulting in over 1.9 billion user chats. Additionally, Microsoft's Azure cloud computing revenue, bolstered by AI services, exceeded expectations, highlighting the financial success of Microsoft's partnership with OpenAI.
My take: This milestone for GitHub Copilot is not just a win for Microsoft but a clear indicator of the burgeoning demand for AI-assisted development tools in the tech industry. Nadella's vision of infusing AI into Microsoft's diverse product range is proving to be a savvy move, not only enhancing user experience but also driving significant revenue growth. The success of GitHub Copilot, coupled with the widespread adoption of AI in Bing and Azure, goes far with praising their vision, but I am worried about a future creeping onto us with endless code repos being written entirely by AI.

In other mainstream news:

Xbox and Inworld AI Join Forces to Revolutionize Game Development with AI 🟠

NVidia pilots Gen AI across their engineering teams 🟠

YouTube Tests New Generative AI Features for Enhanced User Interaction 🟠

UK's Major Investment in AI Supercomputing: Aiming for Global Leadership 💚