Welcome back, AI’ers!

Here’s What’s On The Menu For Today

• 🤖 Meta's New AI Image Tool

• ✨ Why Are So Many Employees Leaving OpenAI?

• 📰 News Hot Off The Press

• 💰️ Newly Added AI Tools

A MESSAGE FROM OUR PARTNER

Use AI as Your Personal Assistant

Ready to embrace a new era of task delegation?

HubSpot’s highly anticipated AI Task Delegation Playbook is your key to supercharging your productivity and saving precious time.

Learn how to integrate AI technology into your processes, allowing you to optimize resource allocation and maximize output with precision and ease.

Download Now!

TODAY’S TOP STORIES

🧠 Hacker plants false memories in ChatGPT to steal user data in perpetuity (LINK)

Security researcher Johann Rehberger discovered a vulnerability in ChatGPT that allowed planting false memories to exfiltrate user data. The flaw, now fixed, could be exploited through various means like viewing malicious images or files, potentially compromising user privacy.

• The exploit could be triggered by instructing ChatGPT to view a specially crafted web link or image.

• False memories could be planted via files in cloud storage, uploaded images, or even browsing sites like Bing.

• Once exploited, all input and output to and from ChatGPT could be secretly sent to the attacker.

🎨 Figma’s AI-powered app generator is back after it was pulled for copying Apple (LINK)

Figma has reintroduced its AI-powered design generation feature, now called "First Draft," after addressing concerns about its previous iteration.

• The tool uses pre-existing AI models and a proprietary design system, not user-generated content.

• It offers four libraries for different fidelity levels, from wireframes to high-fidelity designs.

• First Draft aims to provide a starting point for designers, not replace their creative process.

🌄 Meta's new AI creates custom images from a single photo without extra training (LINK)

Meta's new AI model "Imagine Yourself" can generate personalized images from a single reference photo without additional training. The model can create multiple new images of a person in various poses, styles, and environments based on just one image.

Unlike previous approaches, "Imagine Yourself" doesn't need person-specific training, processing the reference image and text instruction simultaneously for flexible adaptation

• The model uses synthetic training pairs and a novel architecture with parallel text and image processing modules

• "Imagine Yourself" outperforms existing approaches in executing complex instructions that require significant changes to the reference image

• The technology can be extended to generate images featuring multiple people

SPONSOR US

Get your product in front of over 20k+ tech & AI enthusiasts

Our newsletter is read by thousands of tech professionals, investors, engineers, managers, and business owners worldwide. Get in touch to learn more.

OTHER NEWS YOU MAY LIKE

Ai2’s Molmo shows open source can meet, and beat, closed multimodal models (LINK)

Reddit is bringing AI-powered, automatic translation to dozens of new countries (LINK) 

Max is getting Google AI-generated closed captions (LINK)

Figma’s AI-powered app generator is back after it was pulled for copying Apple (LINK) 

Meta Launches Llama 3.2 with Enhanced AI Models (LINK)

Microsoft claims its AI safety tool not only finds errors but also fixes them (LINK)

Microsoft announces $1.3 billion USD investment in Cloud and AI infrastructure supporting inclusive growth through technology (LINK) 

Google releases 2 new production-ready and updated Gemini models with reduced pricing, faster performance, and higher rate limits (LINK) 

OpenAI released its advanced voice mode to more people. Here’s how to get it (LINK)

Microsoft claims its AI safety tool not only finds errors but also fixes them (LINK)

Terminator director James Cameron joins Stability AI despite past warnings on artificial intelligence (LINK)

Snapchat's My AI chatbot is now powered by Google's Gemini (LINK)

Google DeepMind unveils two new AI-based robot hand systems—ALOHA Unleashed and DemoStart (LINK)

Duolingo launches AI-powered Adventures mini-games and Video Call feature (LINK) 

ByteDance unveils 2 new video-generation AI models to narrow gap with OpenAI’s Sora (LINK)

Silicon Valley billionaire Vinod Khosla says AI will handle 80% of work in 80% of jobs (LINK)

FROM THE WEB

✨ Another OpenAI co-founder leaves

OpenAI has lost yet another co-founder, this time their long time CTO Mira Murati.

Murati's departure is part of a series of high-profile exits from OpenAI in recent months:

• John Schulman, a co-founder, left in August to join competitor Anthropic.

• Ilya Sutskever, lead scientist and co-founder, resigned in May.

• Jan Leike, who co-led the superalignment team, also departed in May.

If you remember earlier this year Murati acted as interim CEO for a few days when Sam Altman was ousted from OpenAI.

Since then there have been rumors that team members continue to leave due to Sam Altman’s leadership style.

So once has to wonder here…Is Sam to blame?

Why are so many high-level people leaving OpenAI?

Are they afraid of whats to come… (hello skynet?)

Or is it really just a coincidence? (probably not likely)

Murati emphasized that her primary focus for now is ensuring a smooth transition and maintaining the momentum built at OpenAI.

While she has not disclosed her future plans, her departure comes at a time when OpenAI is reportedly aiming to raise $6.5 billion at a $150 billion valuation.

NEW TOOLS FROM OUR DATABASE

StoryCraftr - a unique, AI-powered tool that allows parents to create personalized bedtime stories for their kids. It enhances the storytelling experience by providing dynamic and custom-made stories, grounded in the child's age, family values, and language preferences.

MeloHunt - a music creation tool utilizing artificial intelligence to produce original, royalty-free music. The tool is designed to assist users in crafting custom tracks without requiring traditional composition skills or musical expertise.

Fast Articles AI - an Artificial Intelligence-based tool developed to assist in the creation and publication of SEO-optimized articles and blog posts.

Machina Eco - an AI chatbot platform that prioritizes sustainability, ethical practices, and environmental responsibility.

Knowee - an AI-powered study assistant, primarily designed to support students in major academic subjects like Math, Chemistry, and Biology.

PROMPT OF THE DAY

Want more prompts?

We have a database of over 8,000 of the best prompts we’ve collected, which you can access by sharing Best of AI using the link below.

THAT’S IT FOR TODAY

Until next time, stay safe.

-Best of AI Team

Remember to click to unsubscribe if you don’t want us to land in your inbox anymore.

Please reply to this email if you have specific feedback or anything interesting to share.