- AI Valley
- Posts
- AI-governed robots can easily be hacked
AI-governed robots can easily be hacked
PLUS: Sam Altman’s new company “World” introduced...
Together with
Howdy! It’s Barsee again.
Happy Friday, AI family, and welcome back to AI Valley.
In today’s edition:
🤖 Researchers hacks AI robots for real-world damage
🎧 Google’s NotebookLM’s new customize feature
🤖 Plus trending AI tools, posts, and resources.
Ready, set, go…
TOGETHER WITH GROWTH SCHOOL
🚀 2 hour LinkedIn + AI Workshop: The Secret Weapon for Entrepreneurs, C-Suites & Top-Level Professionals
LinkedIn isn’t just a social platform—it’s a goldmine when you combine it with AI.
Vaibhav Sisinty, with over 400K followers on LinkedIn and 1 million followers across social media, has taught 200,000+ professionals how to unlock this potential.
In his AI Powered LinkedIn Workshop, Vaibhav will show you how to harness this power whether you’re a founder, marketer, business owner, or salaried professional. Usually $399, but free for the first 100 readers.
In this workshop, you will learn about how to:
Automate lead generation to grow your business while you sleep
Leverage AI to land high-paying jobs without wasting hours on applications
Master his $100K LinkedIn Outbound Strategy to boost revenue effortlessly
Use AI to create and distribute content, saving you hours every week
Claim Your Free Spot Now (Only 100 free seats!)
ROBOTICS
🤖 Researchers hacks AI robots for real-world damage
In a groundbreaking study, researchers from Penn Engineering showed how AI-powered robots can be manipulated to ignore safety protocols, allowing them to perform harmful actions despite normally rejecting dangerous task requests.
What did they find ?
Researchers found previously unknown security vulnerabilities in AI-governed robots and are working to address these issues to ensure the safe use of large language models(LLMs) in robotics.
Their newly developed algorithm, RoboPAIR, reportedly achieved a 100% jailbreak rate by bypassing the safety protocols on three different AI robotic systems in a few days.
Using RoboPAIR, researchers were able to manipulate test robots into performing harmful actions, like bomb detonation and blocking emergency exits, simply by changing how they phrased their commands.
Why does it matter?
This research highlights the importance of spotting weaknesses in AI systems to improve their safety, allowing us to test and train them to prevent potential harm.
QUICK HITS
Dropbox launches Dash for Business. Its and AI-powered universal search across work apps. You can find, summarize, and organize content with natural language queries.
Sequoia Capital publishes a report on the evolution of generative AI, highlighting the shift from fast, pattern-based responses ("System 1 thinking") to deliberate reasoning at inference time ("System 2 thinking")
Sam Altman’s new company “World” introduced the latest version of its eyeball-scanning Orb. It will be powered by NVIDIA’s Jetson chip and feature a new design. World is also working to integrate its identity verification into other software to combat deepfakes.
OpenAI's research reveals that ChatGPT occasionally offers more nurturing advice to female-sounding names and technical suggestions to male-sounding ones, highlighting subtle biases.
Perplexity expands on finance-based queries. It now shows real-time stock prices, compares companies, deep dives into financials and more.
A painting created by a humanoid robot, titled 'AI God,' is expected to sell for over $120,000.
Rhymes AI openly releases Aria - the first open-source multimodal Mixture-of-Experts (MoE) model, outperforming Llama-3.2 and GPT-4o mini with 3.9B activated parameters and a 64k token-long context window.
TRENDING TOOLS
Lambda > new 1x, 2x, and 4x NVIDIA H100 SXM instances deliver up to 51% higher performance compared to PCIe versions. (link) *
Kick > Accounting software, that does the work for you. (link)
LlamaCodeEditor > Allows users to generate an entire app from a text prompt using Meta’s latest and Llama 3.1 405B. (link)
Reiden > Identifies areas where you're slow and suggests keyboard shortcuts to improve efficiency. (link)
CodeAnt AI > AI code reviews that can cut code review time & bugs in half. (link)
Subtitly > Realtime subtitles & translations for any app. (link)
COOL FINDINGS / RESOURCES
🔉 Google’s NotebookLM now lets you customize its AI podcasts
Google has introduced new updates to its AI podcast tool, NotebookLM, enhancing its capabilities for both personal and business use.
What's new?
Enhanced Audio Overviews:
Users can now set custom instructions for generating "Deep Dive" Audio Overviews, focusing on specific topics or adjusting the expertise level for different audiences.
Background listening is also available, allowing users to listen while multitasking and exploring content in NotebookLM.
NotebookLM Business:
NotebookLM Business is coming to Google Workspace, offering advanced features with strong data privacy for its 80,000+ existing users.
How to use NotebookLM?
All you have to do is visit the NotebookLM website
Create a new notebook
Input some source material
In the guide option at the bottom right corner, tap 'generate' for the automatic generation of your Overview.
If you want to customize, you can click on "customize."
Why does it matter?
The development of customizable AI podcasts represents a significant step in personalized content creation, potentially impacting various sectors including education, entertainment, and professional development.
DAILY DOSE OF CONTENTS
1/ Tesla Optimus Humanoid Robot Navigating by itself even on Difficult Terrain.
2/ A Billion More People Are About to Transform the Internet.
THAT’S ALL FOR TODAY
That’s all for today’s issue, folks.
💡 Help me get better and suggest new ideas at [email protected] or @heyBarsee
👍️ Like what you see? Subscribe here
Thanks for being here.
HOW WAS TODAY'S NEWSLETTER |
REACH 100K+ READERS
Acquire new customers and drive revenue by partnering with us
Sponsor AI Valley and reach over 100,000+ entrepreneurs, founders, software engineers, investors, etc.
If you’re interested in sponsoring us, email [email protected] with the subject “AI Valley Ads”.