This humanoid robot can do your household chores

Hello AI enthusiasts! Here’s what you need to know about AI today:

👉 Google’s Career Dreamer helps you easily find your next job

👉 Helix, a humanoid robot that does household chores

👉 OpenAI launches “Operator“ in more countries

and many more!

📧 Did someone forward you this email? Subscribe here for free to get the latest AI news everyday!

Read time: 5 minutes

What’s going on: Google has introduced an AI-powered tool called Career Dreamer, designed to assist individuals in exploring a broader range of career opportunities. Launched as an experimental feature, this tool uses AI to analyze a user's experiences, educational background, skills, and interests, identifying patterns that connect them to potential career paths. Interested? Check it out here.

What does it mean: Unlike traditional job search platforms, Career Dreamer focuses on career exploration rather than linking users directly to job postings. It aims to simplify the often daunting task of finding a career fit without requiring users to sift through countless search results manually.

More details: 

• This tool allows you to input your current and past roles, skills, educational history, and personal interests, from which it generates a “career identity statement.” This concise summary can be added to a resume or used as talking points during interviews.

• It also presents a visual web of career options tailored to your profile, enabling you to explore diverse paths and dive deeper into specific roles that pique your interest.

• Career Dreamer is available only in the United States as part of Google’s experimental rollout, with no clear timeline for expansion to other regions.

What’s going on: Figure, a Bay Area robotics firm, has introduced Helix, a new Vision-Language-Action (VLA) model designed to enable its humanoid robots to perform household tasks through natural language voice commands. This system combines visual data and language processing to allow robots to assess their environment and execute tasks in real time. For example, Helix can direct robots to hand over a bag of cookies to another robot or place items in a drawer, showcasing its ability to handle novel objects with varying properties without prior training.

What does it mean: The development of Helix addresses the challenges of deploying robots in unstructured home environments, which differ significantly from controlled industrial settings. Homes present variability in layout, lighting, and object placement, making traditional manual programming or repetitive training methods impractical. Figure emphasizes that Helix’s approach reduces reliance on extensive human intervention, such as expert programming or thousands of demonstrations, by leveraging generalized learning capabilities.

More details:

• Vision-language-action (VLA) models represent a new approach in robotics, enabling robots to understand and act upon visual and linguistic instructions. A leading example is Google DeepMind's RT-2, which uses both video and large language models (LLMs) for robot training.

• Helix has achieved several groundbreaking firsts for vision-language-action (VLA) models, including full-upper-body control, multi-robot collaboration on shared tasks, the ability to pick up novel objects based on language prompts, learning all behaviors with a single neural network, and commercial readiness through onboard, low-power GPU operation.

• Helix uses a novel "System 1, System 2" VLA, combining a slower and generalizing VLM with a fast, reactive visuomotor policy to achieve dexterous, whole-upper-body robot control.

• Interested? Visit this official blog post for more details and visual demos.

🐋 DeepSeek plans to open-source 5 code repositories next week to promote transparency and share its progress in AI.

🎮 Microsoft Research has unveiled Muse, a generative AI model designed for gameplay ideation, capable of generating game visuals and controller actions to assist developers in prototyping and potentially preserving classic games.

🙆‍♀️ Nvidia has launched "Signs", an AI-powered platform designed to make learning American Sign Language (ASL) easier, featuring a 3D avatar to demonstrate signs and providing real-time feedback to users. Check it out here.

🚀 OpenAI has launched its AI agent "Operator" for ChatGPT Pro subscribers in multiple countries including Australia, Brazil, Canada, India, Japan, Singapore, South Korea, and the UK.

👩‍🔬👨‍🔬 MIT researchers have developed FragFold, an AI-based computational method that systematically predicts protein fragments capable of binding to or inhibiting target proteins.

💰 Together AI, a startup providing access to AI computing, has raised $305 million in a funding round led by General Catalyst, increasing its valuation to $3.3 billion.

💸 Mercor, an AI recruiting startup founded by three 21-year-old Thiel Fellows, has raised $100 million, achieving a $2 billion valuation. Learn more about Thiel Fellowship.

📈 OpenAI now serves 400 million weekly active users, a 33% increase from 300 million in December 2024, with its paying enterprise users doubling to 2 million since September 2024.

🎧 Spotify has partnered with ElevenLabs to expand its library of AI-narrated audiobooks, allowing authors to use ElevenLabs’ AI voice technology to narrate their works in 29 languages.

AI + Meeting preparation

Faraday Future - AI Research and Development Engineer

Omnicell - ML Engineer

Legend Biotech - Data Science Intern

Staples - Sr AI/ML (GenAI) Engineer I

Thank you for staying with us like always! If you are not subscribed, subscribe here for free to get more of these emails in your inbox! Cheers!