Meta launches a breakthrough Multilingual AI translation model.

We dive into AI product launches from Meta, Figma, and HuggingFace; Elon Musk's AI firm's ambitions about the Universe; Big funding for AI-based drug discovery.

👋🏽 Hey folks, I’m Nawaz and Welcome to the AI Shortcut. It’s been another fantastic week for AI. In today’s post, we have,

• Speech, Voice, and Text AI translation model breakthroughs from Meta and ElevenLabs

• Elon Musk’s new AI firm’s goal is to learn the Universe’s true nature

• IBM Study calls for reskilling 40% of the world’s workforce in the next three years

• OpenAI’s GPT3.5 Turbo fine-tuning API and ScaleAI partnership

• AI infrastructure and Medical AI startups get funding north of $100M

Without further delay, let’s jump right in!

---

🎙️ Speech, Text, and Voice AI

Photo by Malte Helmhold on Unsplash

Meta: Launches SeamlessM4T; Meta’s multilingual and multitask AI model propels us into a new realm of translation and transcription capabilities with automatic speech recognition for nearly 100 languages, speech-to-text translation across 100 input and output languages, and text-to-speech translation in almost 100 input languages and 35 (+ English) output languages.

ElevenLabs: Launches a new emotionally rich AI audio multilingual voice generation model capable of generating nearly 30 languages.

Figma: Launches Jambot to brainstorm, zoom out of, or zoom into a given topic, get an answer to a quick question, and summarize inputs in your file.

---

🧐 IBM Study: Reskilling Workforce for the AI Uprising

Photo by krakenimages on Unsplash

IBM Study: Reveals nearly 40% of the workforce will need to reskill due to implementing AI and automation over the next three years. Time management and prioritizing are the most essential skills in the AI era.

---

Elon Musk wants to learn the true nature of the Universe

Photo by Maarten van den Heuvel on Unsplash

X.ai: Elon Musk wants to understand the true nature of the Universe through his new AI firm. Elon Musk leads the team and includes people from DeepMind, OpenAI, Google Research, Microsoft Research, Tesla, and the University of Toronto, who worked on AlphaStar, AlphaCode, Inception, Minerva, GPT-3.5, and GPT-4.

---

👩🏽‍💻 Enterprise and Developer AI

HuggingFace: Launches SafeCoder, a secure and privacy-focused coding assistant for enterprise developers to increase productivity. No code leaves the customer’s infrastructure during the AI training or inference.

OpenAI: Releases new API endpoints for developers to customize GPT-3.5 Turbo for their use cases with their own data. And names Scale.ai as its preferred partner for fine-tuning OpenAI models.

---

💰 The big beneficiary of AI

NVIDIA: The Chip manufacturer’s Q2 revenue is up 101% YOY, thanks to the demand for GenAI and the company’s early bet on AI.

---

💸 AI funding soars

Photo by National Cancer Institute on Unsplash

Genesis Therapeutics: A biotech startup for creating breakthrough medicines for patients with severe diseases captures $200M in Series B investment co-led by Andreesen.

Modular: Raises $100M for improving AI infrastructure for developers—AI programmability, usability, scalability, compute efficiency, and hardware utilization.

HuggingFace: Gets $225M in Series D from investors, including Google, Amazon, Nvidia, Intel, AMD, Qualcomm, IBM, Salesforce, and Sound Ventures, valuing Hugging Face at $4.5 billion. HuggingFace provides a GitHub-like hub for AI code repositories, models, datasets, and web apps to demo AI-powered applications.

---

If you liked reading this newsletter, consider sharing it with friends or subscribing if you haven’t already. And give us a follow on Twitter 🙂

Sincerely,

Nawaz 👋🏽

Note: We updated the newsletter format to experiment and see what our readers like more. I’d be grateful if you could take a minute to give your feedback here. ❤️