AI Advancements Unveiled: Latest News and Cutting-Edge Tools of the Week

Your Weekly Shortcut to Deeptech Investing—Exclusive Trends & Startup Reports for VCs & Angels in Just 5 Minutes

By Lirone Samoun, Deeptech Expert

🔎 Latest news on the AI Space

• Voicebox: The Most Versatile AI for Speech Generation by Meta. Voicebox was announced recently and it promises to recreate realistic voices from only 2-seconds of input audio. So you can give it a 2-second clip of your voice and it will create a realistic version of it for text-to-speech, rivaling tools like ElevenLabs. It can also be used to edit out background noise and translate your audio to other languages while maintaining the characteristics of your voice.

  • Synthesizes speech across 6 languages

  • Can perform tasks it wasn't trained on

  • Has noise removal, content editing, style conversion, and more

  • Supports text-to-speech synthesis and cross-lingual style transfer

  • 20x faster than current models and outperforms single-purpose models through in-context learning (yes, 20x)

• Google has opened its Generative AI Platform to everyone
  Developers can now utilize the text model powered by PaLM2, as well as the Embeddings API for text and several other foundational models available in the model garden.
  

  

• Introducing Apple Vision Pro: Apple’s first spatial computer - This was the news that consumed the tech world last week. Apple announced their new mixed-reality headset, the Apple Vision Pro. They pitched it as a replacement for home entertainment centers and desktop computers. However, it won't be available until next year and the price starts at $3,499.
  

  

• Neuralangelo: A high-fidelity neural surface reconstruction method that efficiently recovers detailed 3D surfaces from multi-view images.

• Google Lens Can Now Search for Skin Conditions: Google has expanded the capabilities of Google Lens, its computer vision-powered application that provides information about identified objects, with a new feature: Users can now utilize Lens to identify and search for skin conditions such as moles and rashes by uploading pictures or photos.

  

• Meta released its own AI-powered music generator called MusicGen, and has open-sourced the tool. MusicGen is trained on over 20,000 hours of music. It can turn text descriptions into short audio clips, and can even be guided by reference audio. It produces high-quality music while being conditioned on text description or melodic features.

• Framer releases a tool that allows you to generate and publish a website in seconds with AI.

• Adobe launches Generative Recolor for Adobe Illustrator, powered by Firefly

• AMD revealed their new super-chip aimed at generative AI, a big move considering the landscape (Nvidia currently takes 80% of the market share for AI chips)

• Salesforce unveils AI Cloud, a suite of GPT-powered applications for CRM. The platform blends AI, data, analytics, and automation Creating a new level of customer experience and business productivity

• The RedPajama project releases RedPajama-INCITE-7B-Instruct
  This model represents the top-performing open-source entry on the HELM benchmarks, surpassing other cutting-edge open models like LLaMA-7B, Falcon-7B, and MPT-7B. The instruct-tuned model, designed for versatility, shines when tasked with few-shot performance.

• Runway's Gen-2 text-to-video tool is available to everyone for free
  The tool creates 4-second MP4 videos based on the input prompt. Moreover, it can also generate short video sequences from an image or from the combination of an image and a text description.

•  Hugging Face released a free QR code art generator - This tool allows users to create customized, artistic QR codes. All you need to do is provide an image and a corresponding prompt, then sit back and watch as the generator transforms it into an intricately designed QR code. And the best part is that these QR codes are fully workable.

  

• Lit-Parrot: is an open-source language model repository powered by Lightning Fabric and built upon the Lit-LLaMA and nanoGPT. It features implementations of state-of-the-art open-source large language models (LLMs) including:

  • StabilityAI StableLM

  • EleutherAI Pythia

  • TII UAE Falcon

  • Together RedPajama-INCITE

🔍 AI Radar

• Neuralink Begins Human Trials. The brain-computer interface is one of Musk’s most ambitious bets in a business empire that spans electric cars to rockets propelling humans to space and that has grown most recently to encompass generative artificial intelligence and social media. You can read the latest article about it written by one of our deep tech expert.

• French tech company Mistral AI, comprised of 3 ex-Meta and Google researchers, has raised €105 million - just 4 weeks after being founded. This makes Mistral AI Europe’s largest-ever seed round.

• According to a report from McKinsey Global Institute, generative AI has the potential to add up to $4.4 trillion of value to the global economy annually. The report also suggests that by 2030 to 2060, half of all work could be automated.

• Using AI To Talk To Animals: Scientists are using machine learning to decode animal communication, with the Earth Sciences Project leading this revolutionary approach to understanding the vocal cues of beluga whales, potentially transforming ethology and animal welfare initiatives.

• MIT Enhancing Autopilot for Flying: Israeli scientists at MIT have developed an innovative AI autopilot algorithm that enhances the stability of aircraft during potentially fatal near-crash situations, according to a non-peer-reviewed study.

• Hugging Face and AMD partner on accelerating state-of-the-art models for CPU and GPU platforms

• Synthesia, a company specializing in AI-generated videos using synthetic avatars and voice, just announced a $90 million Series C funding round led by Accel, with participation from Nvidia and other investors.

• Google has announced their new virtual try-on feature, allowing users to see how clothes look on real models with different body shapes and sizes.

• Mercedes-Benz is teaming up with Microsoft's Azure OpenAI Service to integrate ChatGPT into its in-car voice assistant.

• Google is urging caution among employees when using chatbots, including their own Bard. Concerns about data leaks have led to the advice of not entering confidential information into AI chatbots, amidst concerns of chatbots using internal data in it’s training.

• Intel Doubles Down on Chip Manufacturing: Intel announced its plan to invest $4.6 billion in a new chip plant in Poland, as part of its broader strategy to expand chip capacity in Europe and regain its competitive position in the semiconductor industry against rivals like AMD, Nvidia, and Samsung.

• Salesforce pledges to invest $500M in generative AI startups.

• Instagram may be getting its own AI chatbot soon.

• OpenAI, Google DeepMind and Anthropic just granted the UK government priority access to their AI models for research and safety purposes.

• An internal Amazon document was just leaked, showcasing how the company is encouraging workers to embrace ChatGPT and similar AI technologies.

• Japan Goes All In on AI: The Japanese government recently surprisingly stated that it will not enforce copyrights on data used for training AI.

💻 CooI AI Tools / Startup

• Uncrop by Stability AI: Utilizes outpainting to alter the aspect ratio of images by creating an expanded background, free to use!

  

• Snipd: Get personalized notes for podcasts by tapping your headphones, eliminating manual note-taking, and receive takeaways via email.

• AI Voice Cloning (With 99% Accuracy) : Used in television shows, broadcast news, advertisements, and video games, PlayHT is bringing human-like AI voice generation to everyone.

• Klap: for content creators. Plug-in a URL to a YouTube video and it will analyze the video and find short clips. It will cut out those clips and even add captions to make them more engaging.

• You can now transform 3D/4D images from an ultrasound into photorealistic images. Try the tool here. In the future, parents will may be able to see how the baby looks before the baby is born.

• Google announces Dreamix: a model that generates videos when given a video + prompt.

• SingSong, a system that can generate instrumental accompaniments to pair with input vocals.

• Falcon, a new family of very high-quality LLMs just released. It outperforms competitors like Meta's LLaMA and Stability AI's StableLM. Fully open-sourced with over 40 billion parameter.

• Audio Pen - An app that converts your voice notes into concisely summarized text.

• Dummie AI - Turn your videos into engaging shorts that outperform, no editing needed.

• Composer – Build, backtest and execute trading algorithms with AI.

• Paragraphica: This is a very unique tool. You give it a location, and it will craft a prompt based on that location, the time of day, the weather, and nearby points of interest. It will then generate an image based on the prompt. This gives you the ability to take a picture of a location and get a general idea of what it looks like at a given moment.

• Client Zen : Consolidate your customer reviews 6 times faster to better analyze them and identify areas for improvement.

• Gliglish : A language teacher who gives you confidence and improves your speaking skills.

• LoopGenius: builds websites in 30 seconds

  This is one of the easiest and simplest tools to start a business or side hustle.

  Just give your startup/business idea a name, write a short description, and get a marketing site built by AI in 30 seconds .. with what feels like human-written content.

• FashionAI: Take a picture of a person, then modify clothing or explore fashion using AI.

• Graphy AI: Simply explain the metric you want to visualize or the question you would like to answer. It will provide a custom template with a recommendation for the best chart type.

• Pic Finder: A infinite image generation tool that's free forever.

• QRCraft- Turn boring QR codes into captivating works of art.

• WebBotify- ChatGPT trained for your website in 2 min.

• MoveMe: Emotion-led AI recommendations across all your streaming services.

---

✨ That’s all for today. Thanks for reading !

💖Like, follow and subscribe to our Community ! Stay tuned for our next article coming up end of the week with our Deeptech Insights Newsletter.

Much love Deeptechers!👋

---