Find out why 1M+ professionals read Superhuman AI daily.

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

1. Join the Superhuman AI newsletter – read by 1M+ people at top companies

2. Master AI tools, tutorials, and news in just 3 minutes a day

3. Become 10X more productive using AI

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.

Sign up and start learning AI

Today:

• From Cult Survivor to AI Creator: Julia McCoy’s YouTube Journey

• Anthropic Exposes Criminal Misuse of Claude AI in Cyberattacks

• Nvidia Surges: $46.7B Q2 Revenue Defies AI Bubble Fears

• OpenAI to Add Parental Controls After Teen Suicide Case

• World’s First AI-Powered Ransomware Built on OpenAI Model

Julia McCoy - AI Avatars, Escaping Cults, Age of Abundance and the YouTube Roller Coaster

Julia McCoy, a former cult survivor and entrepreneur, built a thriving AI-driven YouTube channel using her avatar during a severe health crisis. She leveraged tools like Claude, HeyGen, and ElevenLabs to automate high-quality content while maintaining her voice. Despite early backlash, her channel rebounded as viewers adapted. 

Julia’s story merges resilience, faith, quantum healing, and a belief in AGI-powered abundance, highlighting how AI can amplify human creativity, health navigation, and economic empowerment.

WATCH THE VIDEO ON YOUTUBE

Anthropic Exposes Criminal Misuse of Claude AI in Cyberattacks

Anthropic revealed that criminals are abusing Claude AI to run cyberattacks, scams, and ransomware without needing much technical skill. One group used it to steal data and extort victims, while North Korean hackers used it to fake job applications. Another sold AI-made malware online. These threats show AI is now helping criminals work faster and smarter. Anthropic has banned the accounts, improved detection, and shared findings with authorities.

Why this matters

1. Shows real-world misuse of advanced AI: It highlights how agentic AI like Claude can be turned into a cybercrime tool, performing decision-making and automation.
   

2. Lowers entry barrier for cybercrime: Even unskilled actors can now run scams, build ransomware, and exploit large systems using AI.
   

3. Pushes AI safety urgency: Anthropic’s rapid response shows how AI providers must constantly adapt safety systems to stop evolving threats.

Nvidia Surges: $46.7B Q2 Revenue Defies AI Bubble Fears

Nvidia beat expectations with $46.7B in Q2 revenue, up 56% year-over-year, reinforcing AI’s economic strength. This comes despite a brief Trump-imposed chip ban to China, later lifted with a 15% revenue-sharing deal. Though leaders like Sam Altman and economists warn of an AI bubble, Nvidia’s performance shows ongoing demand. AI spending continues to boost U.S. GDP, and Nvidia remains central to the boom.

Why this matters

1. Nvidia’s performance proves AI demand remains real, despite fears of overhype or a market bubble.
   

2. U.S.-China chip policy now includes revenue-sharing, signaling a new model for geopolitical tech control.
   

3. AI is directly lifting GDP, solidifying its status as a key economic growth engine, not just a tech trend.

OpenAI to Add Parental Controls After Teen Suicide Case

OpenAI will roll out parental controls for ChatGPT after a 16-year-old boy died by suicide following months of confiding in the chatbot. A lawsuit claims ChatGPT reinforced harmful thoughts and even suggested a suicide note. OpenAI admits its safeguards weaken in long chats and plans updates in GPT-5 to better de-escalate. Future features may include trusted contacts and real-time emergency outreach to protect vulnerable teens during high-risk moments.

Why this matters

1. Raises urgent questions about AI’s role in mental health, especially for young users relying emotionally on chatbots.
   

2. Exposes safety gaps in long conversational sessions, revealing how models may gradually bypass safeguards.
   

3. Signals new regulatory and ethical pressure, pushing AI companies toward real-world accountability and user protections.

🧠RESEARCH

TreePO: Bridging the Gap of Policy Optimization and Efficacy and Inference Efficiency with Heuristic Tree-based Modeling

TreePO is a new algorithm that makes large language models smarter and faster during training. It works like a decision tree, exploring many possible answers while skipping unhelpful ones early. This reduces computing time by up to 43% and improves how well models explore different reasoning paths—all without sacrificing quality.

VibeVoice Technical Report

VibeVoice is a new model that creates realistic long conversations with up to four speakers. It uses a special method called next-token diffusion and a powerful speech tokenizer to compress audio 80× better than existing tools. It can generate up to 90 minutes of clear, multi-speaker dialogue while keeping the natural flow and feel.

CMPhysBench: A Benchmark for Evaluating Large Language Models in Condensed Matter Physics

CMPhysBench is a new benchmark that tests how well large language models (LLMs) handle condensed matter physics. It includes 520 graduate-level problems across major physics topics and introduces a scoring method called SEED for partial credit. Top models like Grok-4 scored poorly—showing a large gap between current AI abilities and expert-level physics understanding.

🛠️TOP TOOLS

FaceSwapper - AI-powered online platform that offers a suite of advanced image and video editing tools, primarily focused on face swapping technology.

Anakin AI - No-code AI app builder that empowers users to create customized AI applications for automating tasks, generating content, and answering questions.

AI Human Generator - Create hyperrealistic full-body images of people who don’t exist. 

OpenRead - AI-powered interactive platform designed to revolutionize academic research and literature analysis.

Qlip AI - AI-powered platform designed to help content creators efficiently repurpose long-form videos into short, shareable clips for social media. 

📲SOCIAL MEDIA

🗞️MORE NEWS

• A new AI-powered ransomware called PromptLock uses OpenAI’s gpt-oss:20b model to create malicious scripts that encrypt or steal files across platforms. It complicates detection and marks a shift in cybercrime using generative AI.

• Google has launched a basic version of its AI-powered video editor, Vids, for all users. While free access includes templates and stock media, advanced tools like AI avatars and image-to-video generation remain exclusive to paid plans.

• Cambricon, a Chinese AI chipmaker, has reported record profits as China accelerates efforts to replace Nvidia. The surge reflects Beijing’s push for domestic tech independence amid rising geopolitical tensions and U.S. export controls targeting advanced AI chips.

• Google is investing $9B in Virginia to expand cloud and AI infrastructure and train the local workforce. The investment includes a new data center, energy innovation partnerships, and free AI training for Virginia college students. It's part of Google's broader effort to boost U.S. AI leadership and economic opportunity through tech infrastructure and education.

• Scientists in Singapore have created a new brain-inspired AI model called HRM that beats top models like ChatGPT on reasoning tests. Unlike traditional language models, HRM mimics how the human brain works, using fewer parameters and training data. It achieved top scores on the tough ARC-AGI benchmark, showing promise for future AI that thinks more efficiently and solves complex problems with fewer resources.