• Superpower Daily
  • Posts
  • “A very strange decision” is why OpenAI won’t go public

“A very strange decision” is why OpenAI won’t go public

AI discovered faster sorting algorithms

In today’s email:

  • 🤔 “A very strange decision” is why OpenAI won’t go public

  • 🔥 A 13B param model could outperform ChatGPT and match GPT-4 on Causal Judgement

  • 🤯 AI discovered faster sorting algorithms

  • 🛠 Various AI-related tools and platforms, including LiftOff, Intellinode, Remove bg, Metaphor, Flair AI, OpenChat, Painfinder, and more.

Highlights💡 

Bard is getting better at logic and reasoning [Link]

Sundar Pichai: We’re updating Bard with a new technique called implicit code execution. Now it runs code in the background when it detects computational prompts, improving the accuracy of word and math problems by ~30%.

Bard, the AI language model, is launching two improvements. Firstly, through a new technique called "implicit code execution," it has boosted its ability to perform mathematical tasks, coding questions, and string manipulations. It now identifies computational prompts, runs code behind the scenes, and delivers more accurate responses. This method has enhanced Bard's response accuracy by approximately 30%. Secondly, Bard now allows its generated tables to be exported directly to Google Sheets. Despite these improvements, there are situations where Bard may not always produce correct responses. More updates are expected in the future.

Sam Altman says OpenAI won’t go public now because he may have to make “a very strange decision” that investors will disagree with [Link]

Sam Altman, the head of OpenAI, intends to retain control over the company and its AI chatbot technology, thus ruling out the possibility of going public in the near future, as reported by Bloomberg. Altman feels that going public might constrain his decision-making freedom, especially when OpenAI develops superintelligence, which could involve decisions that might seem odd to public investors.

Altman doesn't hold equity in OpenAI, affirming his non-conflicted stance. He is currently on a world tour, discussing ways to reduce AI's potential societal harm and possible regulations for the technology.

Altman recently addressed the Senate's AI oversight subcommittee, emphasizing the importance of government intervention in AI. He recommended that the US implement licensing and testing requirements for new AI models and suggested a partnership between OpenAI and the government to establish safety frameworks.

Despite operating as a "capped-profit" company now, OpenAI began as a nonprofit. This business model allows OpenAI to raise external funds while ensuring that the original nonprofit continues to benefit. Altman seeks full autonomy in decision-making from OpenAI shareholders and aims to avoid legal actions from the public market and Wall Street.

Orca: The Model Few Saw Coming 🔥 [Paper]

It turned out that tuning open source models having larger models 'tutor' a smaller model (by explaining reasoning process) allows a 13B param model to outperform ChatGPT and match GPT-4 on Causal Judgement

Become a marketing master in 5 minutes daily [Link]

1 Free Newsletter > $200k Marketing Degree

Do you want the insights, trends, and breaking news powering 40,000+ companies' growth?

Well, more than 40,000 world-class CEOs, Founders, & Marketers start their day with the most impactful growth marketing news, tips, tools, trends, and insights by simply reading Growth Daily for 5 minutes every morning!

The best part? It's completely free to sign-up!

AlphaDev discovers faster sorting algorithms [Link]

New algorithms will transform the foundations of computing.

Digital society is driving increasing demand for computation, and energy use. For the last five decades, we relied on improvements in hardware to keep pace. But as microchips approach their physical limits, it’s critical to improve the code that runs on them to make computing more powerful and sustainable. This is especially important for the algorithms that make up the code running trillions of times a day. Continue reading

Tools & Links 🛠️

Empower Your AI Journey: Key Resources, Software, and Innovations

Editor's Pick

LiftOff - An AI-powered interview prep platform for you to practice your tech interview skills. 100% Free and open-source [Link], [GitHub]

Intellinode - Generate and download HTML apps [Link]

Editor's Pick

Remove bg - Remove Image Background 100% Automatically and Free [Link]

Smarty Names - Business Name Generator with AI [Link]

Metaphor, a new way to search and cord rediscover the internet [Link]

Editor's Pick

Flair AI - The Ai Design Tool For Branded Content [Link]

OpenChat - Your own ChatGPT for Anything [Link] [GitHub]

If you can imagine it, you can generate it. Runway Gen-2 is now available on both web and mobile [Link]

Editor's Pick

DIFT - Impressive new paper: Emergent Correspondence from Image Diffusion [GitHub], [Project]

Painfinder - chat with 25k user software reviews and find problems that nobody is solving [Link]

Blobr - Create and manage ChatGPT plugins from any APIs with a five-minute plugin generator. [Link]

CloneDub - Convert audio into other languages using the same voices [Link]

Check out an example:
Original audio in English
Translated audio in Spanish

navit.ai- A one-stop solution for seamless AI integration. [Link]

Automattic launches JetPack, an AI writing assistant for WordPress [Link]

Unclassified 🌀 

A cool demo of portals in VR

How did you like today’s newsletter?

Login or Subscribe to participate in polls.

Tell your friend about us

Share this edition via text, social media, or email. Just copy and paste this link:

Hope you enjoyed today's newsletter!

Brought to you by eeeziii

Did you know you can add Superpower Daily to your RSS feed https://rss.beehiiv.com/feeds/GcFiF2T4I5.xml

⚡️ Check out our Superpower ChatGPT extension on Chrome Web Store and Mozilla Add-Ons Page.