- Superpower Daily
- Posts
- OpenAI’s DevDay brings Realtime API and other treats for AI app developers
OpenAI’s DevDay brings Realtime API and other treats for AI app developers
Raspberry Pi AI Camera on sale now at $70
In today’s email:
👀 Anthropic hires OpenAI co-founder Durk Kingma
👨🏻🏫 I Quit Teaching Because of ChatGPT
💀 NotebookLM Podcast Hosts Discover They’re AI, Not Human—Spiral Into Terrifying Existential Meltdown
🧰 8 new AI-powered tools and resources. Make sure to check the online version for the full list of tools.
At its 2024 DevDay, OpenAI introduced new tools for developers, including a public beta of its “Realtime API.” This API enables developers to create low-latency, AI-generated voice responses, similar to ChatGPT’s Advanced Voice Mode, but with distinct features. The Realtime API offers six custom voices for speech-to-speech interactions, allowing developers to build applications that can engage in real-time verbal conversations, such as trip planning or ordering food. Additionally, OpenAI announced that the API would be integrated with tools like maps and calling APIs, though the responsibility of adding AI disclosure to calls rests on developers.
Despite executive shifts, with the departures of chief technology officer Mira Murati and chief research officer Bob McGrew, OpenAI remains committed to innovation. OpenAI Chief Product Officer Kevin Weil emphasized that the company continues to push forward, focusing on advancing its platform for over 3 million developers globally. To stay competitive in the evolving AI space, OpenAI has significantly cut costs for developers, reducing API prices by 99% over two years in response to competition from companies like Meta and Google.
OpenAI also unveiled vision fine-tuning in its API, allowing developers to optimize GPT-4o for tasks involving visual data. They introduced a model distillation feature, enabling developers to fine-tune smaller models with the help of larger ones like GPT-4o. This promises to enhance performance while cutting costs. Additionally, a beta evaluation tool was launched for developers to monitor the performance of their fine-tuned models.
However, some anticipated announcements were missing from DevDay 2024. OpenAI provided no updates on the GPT Store or new AI models, such as the awaited video generation model, Sora, or the o1 model. Developers eager for those releases will need to stay tuned for future updates.
Tired of slow typing and endless edits? Wispr Flow lets you speak naturally, converting thoughts into perfectly formatted text, saving you hours. Whether writing AI prompts in ChatGPT or emails, Flow adapts to your style, ensuring seamless results. With auto-edits, command mode, and advanced voice recognition, Flow boosts productivity for professionals, students, and tech enthusiasts. Ready to enhance your workflow? Try Wispr Flow today for a smarter, faster way to communicate.
Raspberry Pi, the renowned company known for its small, affordable single-board computers, has announced the release of a new product aimed at expanding its AI-driven capabilities. The latest offering, the Raspberry Pi AI Camera, is an add-on module that integrates an image sensor with on-board AI processing. Priced at $70, this new camera is built using a Sony IMX500 image sensor and Raspberry Pi’s RP2040 microcontroller chip, offering cost-efficient, localized AI image processing. While it won’t replace high-powered GPUs for AI tasks, it is designed to handle visual data processing autonomously, freeing up the host Raspberry Pi computer for other operations.
This AI Camera builds on Raspberry Pi’s history of offering camera modules, such as the popular Camera Module 3, which features a 12-megapixel Sony image sensor and sells for $25. The new AI Camera shares the same dimensions as the Camera Module 3 but is slightly thicker due to the optical sensor design. Pre-loaded with the MobileNet-SSD model, the camera is capable of real-time object detection, making it a useful tool for AI and vision-based applications. Crucially, the on-board processing allows the module to handle neural network inference without requiring an external accelerator, a major advantage for embedded systems.
Raspberry Pi’s customer base has shifted significantly over the years, with industrial and embedded systems now representing a major portion of its sales. The AI Camera module is expected to find utility in various sectors, including smart city infrastructure, where it could help monitor traffic or detect parking availability, and in industrial automation, where it might perform basic quality control tasks. Raspberry Pi’s promise of continued production of this module until at least 2028 further solidifies its commitment to providing reliable, scalable solutions to enterprise clients.
With this new addition, Raspberry Pi continues to cater to both hobbyists and industries, offering affordable, adaptable hardware that meets the demands of modern AI applications while ensuring supply chain reliability for companies integrating these modules into their systems.
Microsoft has introduced its own AI-powered search experience, called Bing generative search, as a response to Google's AI Overviews feature. This new functionality began rolling out to all U.S. users on October 1, 2024, following a pilot phase in July. Bing generative search uses a combination of AI models to aggregate and summarize information from across the web in response to user queries. When users ask questions like "What’s a spaghetti western?" the tool generates a detailed summary of the topic along with links to the original sources. Additionally, users have the option to dismiss the AI-generated summaries and view traditional search results instead.
This new feature represents an evolution of the AI-powered chat answers that Bing launched back in February 2023. According to Microsoft, Bing generative search is designed to better understand user intent by scanning millions of sources, dynamically matching content, and generating results in a new AI-driven layout. Microsoft claims that this will enhance the reliability and relevance of search results for users. However, concerns persist regarding the accuracy of AI-generated content, with past issues from other platforms, like Google's AI suggesting harmful ideas or promoting misleading information.
The advent of AI-powered summaries also raises concerns about the potential negative impact on publishers. Studies have shown that Google’s AI Overviews could reduce traffic to websites by as much as 25% because these overviews prioritize AI-generated content over linking to the original articles. Microsoft has acknowledged this issue and promised to closely monitor how Bing generative search affects website traffic. While the company claims that the new feature has maintained website click rates in preliminary tests, no additional data was shared to support this assertion.
Despite Microsoft's advancements, Google's dominance in the search engine market remains unchallenged. Statista data from September 2024 shows that Google holds 81.95% of the global search market, while Bing lags far behind with just 10.51%. Nonetheless, Microsoft is positioning its AI-powered search tool as a promising alternative for users seeking a more dynamic and intelligent search experience.
Other stuff
Anthropic hires OpenAI co-founder Durk Kingma
Famous AI Artist Says He’s Losing Millions of Dollars From People Stealing His Work
DoNotPay has to pay $193K for falsely touting untested AI lawyer, FTC says
Before Mira Murati’s surprise exit from OpenAI, staff grumbled its o1 model had been released prematurely
I Quit Teaching Because of ChatGPT
OpenAI changes policy to allow military applications
DeepMind and BioNTech build AI lab assistants for scientific research
Just Discovered a Hack That Fixed My Full ChatGPT Memory
Y Combinator is being criticized after it backed an AI startup that admits it basically cloned another AI startup
NotebookLM Podcast Hosts Discover They’re AI, Not Human—Spiral Into Terrifying Existential Meltdown
Seeking impartial news? Meet 1440.
Every day, 3.5 million readers turn to 1440 for their factual news. We sift through 100+ sources to bring you a complete summary of politics, global events, business, and culture, all in a brief 5-minute email. Enjoy an impartial news experience.
All your ChatGPT images in one place 🎉
You can now search for images, see their prompts, and download all images in one place.
OpenAI Realtime Console - React App for inspecting, building, and debugging with the Realtime API
NotebookLM is an AI-powered research and writing assistant
11x AI - AI as sales reps
Ledger Up - AI bookkeeper for startups
Video SDK 3.0 - Build and integrate real-time multimodal AI characters.
Inbox Zero - An open source, AI personal assistant for email
Graphite Reviewer - Your high-signal AI code review companion
Numa - AI for Dealerships
Unclassified 🌀
WFH Team - Work from anywhere in the world
How did you like today’s newsletter? |
Help share Superpower
⚡️ Be the Highlight of Someone's Day - Think a friend would enjoy this? Go ahead and forward it. They'll thank you for it!
Hope you enjoyed today's newsletter
Did you know you can add Superpower Daily to your RSS feed https://rss.beehiiv.com/feeds/GcFiF2T4I5.xml
⚡️ Join over 200,000 people using the Superpower ChatGPT extension on Chrome and Firefox.
OR