- The Rundown AI
- Posts
- Physics' new AI prodigy
Sign Up | Advertise | Podcast | AI University
Welcome, AI enthusiasts.
Step aside, Sir Isaac… There's a new Newton in town, and this one is powered by AI.
Archetype's new model is learning physics from scratch — potentially breaking barriers in autonomous learning. Let’s get into it…
In today’s AI rundown:
Newton AI learns physics from scratch
NotebookLM launches business pilot
Automate web data extraction with Zapier
Worldcoin unveils next-gen eye scanner
5 new AI tools & 4 new AI jobs
More AI & tech news
Read time: 4 minutes
LATEST DEVELOPMENTS
ARCHETYPE AI
Image source: Archetype AI
The Rundown: Archetype AI just unveiled 'Newton,' a new foundational AI ‘Large Behavior Model’ that learns complex physics principles directly from raw sensor data, without any human guidance.
The details:
Newton ingests raw sensor measurements to build its understanding of physical phenomena without pre-programmed knowledge.
The model can accurately predict behaviors of systems it wasn't explicitly trained on, like pendulum motion.
It outperformed specialized AI in tasks like forecasting citywide power consumption and discovering systems from data instead of training.
Archetype AI was founded by ex-Google researchers and has secured $13M in funding to date.
Why it matters: Newton is a paradigm shift in AI's interaction with the physical world. A single model could replace highly specialized systems by developing a generalized understanding rather than a narrow focus. The tech also opens the door to truly autonomous AI that can adapt to environments and tasks without human intervention.
TOGETHER WITH GALILEO
The Rundown: Don't miss GenAI Productionize 2.0, the premier conference for GenAI application development — featuring AI experts from leading brands, startups, and research labs.
Register for GenAI Productionize 2.0 to learn:
How to design a GenAI stack for enterprise scale
Strategies and techniques for AI evaluation and observability
Expert takeaways on measuring the ROI of AI
Image source: Google
The Rundown: Google just pushed an update for its viral AI note-taking assistant NotebookLM, adding new features that let users guide AI-generated audio summaries and announcing the upcoming launch of a new business-focused version.
The details:
Users can now customize the AI podcast Audio Overviews feature by providing instructions to focus on specific topics or adjusting the expertise level.
A new Background Listening feature allows users to listen to Audio Interviews while multitasking within NotebookLM.
A pilot program for NotebookLM Business is coming, offering enhanced features for organizations like higher usage limits and team collaboration tools.
Audio Overviews, which turns docs, videos, and other content into podcasts between AI hosts, went viral earlier this month for its realistic audio outputs.
Why it matters: Google is dropping the ‘experimental’ tag on NotebookLM, and the viral feature built in just two months is suddenly being called a ‘ChatGPT’ moment for the company. It’s also an interesting case of users actually enjoying AI-generated content — a quality that is hard to find in most mainstream sentiment for the tech.
AI TRAINING
The Rundown: Zapier’s Chrome extension allows you to extract and process key information from websites, automating data collection and analysis tasks.
Step-by-step:
Install Zapier Central Chrome extension and log in.
Create a new AI assistant within the extension.
Navigate to your target website and craft a clear data extraction prompt.
Review the results and refine your request for more specific information.
Pro tip: You can turn your AI assistant’s contents into a Google Doc or draft and automatically send emails.
PRESENTED BY FIDDLER AI
The Rundown: Fiddler Trust Service scores trust-related dimensions to detect LLM correctness, safety, and privacy issues with proprietary trust models — helping you enhance the performance and safety of your enterprise LLMs.
With Fiddler, you can:
Score LLMs with high accuracy and low latency using proprietary, task-specific trust models
Reduce costs by 6x, improve accuracy by 50%, and increase speed by 1.5x compared to public benchmarks
Detect and prevent issues like hallucinations, PII leakage, and prompt injection attacks
Start the self-guided tour today to score your LLM applications.
WORLD
Image source: World
The Rundown: Worldcoin, the ‘proof of personhood’ startup founded by OpenAI CEO Sam Altman, just announced a rebrand to ‘World’, along with a new version of its iris-scanning ‘Orb’ technology and updated core platforms.
The details:
A new streamlined Orb promises 5x performance to its predecessor, alongside new countries, self-serve, and on-demand Orbs for easier onboarding.
The company introduced World ID 3.0 protocol, featuring new World ID Credentials, Deep Face to combat AI-generated deepfakes, and added privacy infrastructure.
An updated World App 3.0 allows for anonymous integration with third-party apps, and World is also launching the mainnet of its Worldchain blockchain.
The company has previously faced backlash and even bans from certain countries over privacy concerns.
Why it matters: Verifying human identity in the increasing flood of AI-generated content, agents, and systems is clearly going to be massively important — but given Worldcoin’s rocky launch and international struggles, the question is whether the company can overcome the early drama to actually achieve its goals.
NEW TOOLS & JOBS
🤝 Feta - Run seamless standups, sprint sessions, or syncs with AI-powered documentation and automated workflows
💸 Kick - AI-driven accounting software that does the work for you
🐜 CodeAnt - Automatically detect and fix code quality issues, bugs, and security vulnerabilities in real-time
🎹 Reiden AI - A keyboard shortcut copilot that uses computer vision and integrates with popular apps
🗣️ Convo - Qualitative user research powered by AI
✍️ Anyscale - Technical Writer
🛠️ Meta - ASIC Engineer (Infra Silicon Enablement)
🖥️ Shield AI - Perception Software Engineer, Sensor Fusion
🧑💼 Weights & Biases - Senior Product Manager, Identity & Governance
QUICK HITS
The U.S. Treasury Dept. shared that it leveraged AI to recover $1B in check fraud and prevent $4B in overall fraud in the 2024 fiscal year, showcasing the tech’s growing role in combating financial crime.
OpenAI expanded its partnership with consulting firm Bain & Co. to develop and sell industry-specific AI tools to corporate clients, with OpenAI reporting 1M paying business customers.
Meta is partnering with Blumhouse and other select filmmakers to test its Movie Gen AI video generation tools, gathering feedback to refine the tech before its public release in 2025.
Researchers from Alibaba and Skywork showcased Meissonic, a small, open-source text-to-image model that can generate high-quality outputs that outperform larger models.
Salesforce CEO Marc Benioff criticized Microsoft’s AI initiatives for overhyping the sector in an interview with Fast Company, calling its Copilot assistant the ‘next Clippy.’
OpenAI released a preview of its ChatGPT Windows app for paid users, offering file and photo interactions, model improvements, and a companion window mode.
THAT’S A WRAP
That's it for today!Before you go we’d love to know what you thought of today's newsletter to help us improve The Rundown experience for you. |
See you soon,
Reply