Strange Daily: Issue #11

Ideogram, a gen AI company started by former Google Brain researchers, seems to have cracked text-rendering. Runway announces Creative Partner program. Deepmind works on watermarking AI images

Welcome to your bi-weekly bulletin of the latest in AI, technology, and creativity.

Latest news

  • Ideogram, started by former Google Brain researchers, takes on Midjourney with their text-generation and rendering engine, and raise $16.5M from a16z and Index Ventures [Link]

  • Runway ML announces an open call for their Creative Partners program [Link]

  • Stability AI introduces open-source chatbot StableChat, a web-based chat interface for their advanced language model Stable Beluga [Link]

  • Google rolls out Duet AI for Workspace with generative AI features across apps, including text summarization, writing emails and creating presentations, taking live notes during video call, creating custom plans, images from text prompts, and more [Link]

  • Snapchat launches AI-selfie generator Dreams that transforms selfies into hyper-realistic images [Link]

🤝 Partnerships

  • DeepMind partners with Google Cloud to launch SynthID, a tool for watermarking and identifying AI-generated images [Link]

  • Call of Duty battles against toxicity in voice chats with Modulate’s AI-based ToxMod voice chat moderation system that identifies toxic speech in real-time [Link]

  • FOX Sports expands partnership with Google Cloud: Leveraging generative AI for content creation and transforming viewer experiences [Link]

  • SoundHound AI and online ordering platform ChowNow collaborate: Handling customer queries via a voice assistant [Link]

🎓 Recent Academic Papers

  • Microsoft introduces Algorithm of Thoughts (AoT): A novel approach to AI to make LLMs learn in a manner similar to humans [Link to abstract/full-text]

  • Researchers from UCL and Google propose AudioSlots: A slot-centric generative model for audio domain blind source separation [Link to abstract/full-text]

  • DenseDiffusion: A training-free AI technique to address dense captions and layout manipulations in text-to-image generation [Link to abstract/full-text]

🔧 Cool Tools and Experiments

  • Meta tests its VR avatars with legs in beta for Quest headsets [Link] 

  • Samsung releases AI-driven culinary platform Food that enables users to explore recipes, craft comprehensive shopping lists, orchestrate weekly meal plans, and engage with creators [Link]

  • Meta tests a feature to turn any flat surface into a keyboard with the Quest VR headsets [Link]

  • Grammarly rolls out AI tools for students with features such as ideation and feedback prompts, auto-citations, and AI guideline reminders [Link] 

  • AI21 Labs rolls out upgrades for its writing assistant tool Wordtune, including longform editing, work related templates, proactive AI suggestions, live outlines and knowledge-based control tools [Link]

đź’° Latest Startup Fundings

  • AI companion creator Intuition Robotics secures $25M to combat loneliness [Link]

  • Gaming tech startup NPCx raises $3M in a seed round led by Kakao Investments: Improving motion capture for non-player characters in games [Link]

  • Allseated, a virtual venue tour and event planning platform, raises $20M in a debt funding round [Link]

  • AI21 Labs raises $155M at a $1.4B valuation, accelerating its text-based generative AI services for enterprises [Link]

Stay Strange!

Share the newsletter: If you've found this email insightful, feel free to forward it to a friend. They can subscribe here.

About us: Want to know more about our mission and team? Check out our about page.

\