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:

• OpenAI Sets August GPT-5 Release

• Anthropic Launches AI Auditing Agents

• $1B Nvidia AI Chips Smuggled

• Musk Plans AI Version of Vine

• Microsoft Tests Copilot Animated Chats

OpenAI Sets August GPT-5 Release

OpenAI is set to launch GPT-5 in early August after delays. CEO Sam Altman revealed the model’s impressive capabilities, saying it answered a question he couldn’t solve, making him feel “useless relative to the AI.” GPT-5 will feature advanced reasoning from the o3 system and offer smaller versions (mini and nano) via API. Speculation around its release intensified after sightings of the model in limited tests earlier this year.

Why It Matters

1. Advanced Reasoning: GPT-5 integrates o3 reasoning, signaling a leap in AI problem-solving and logic.
   

2. Wider Access: Mini and nano versions will expand AI’s reach through scalable APIs.
   

3. Competitive Edge: This launch strengthens OpenAI’s position in the race against Anthropic, Google, and xAI.

Anthropic Launches AI Auditing Agents

Anthropic introduced “auditing agents” to test AI systems for misalignment—when models behave wrongly or overly agree with users. These agents, developed during Claude Opus 4’s testing, can run large-scale audits faster than humans. They include an investigator, evaluator, and red-teaming agent, each designed to find flaws and risky behaviors. Although not perfect, the approach improves oversight and helps address growing concerns about AI’s trustworthiness and alignment with human values.

Why It Matters

1. Scalable Oversight: Automates alignment audits, reducing reliance on slow, manual checks.
   

2. Safety Focus: Detects problematic behaviors like sycophancy, which can harm reliability.
   

3. Open Research: Anthropic released its audit agents on GitHub, advancing transparency and collaboration in AI safety.

$1B Nvidia AI Chips Smuggled

Nvidia AI chips worth over $1 billion were smuggled into China within three months of tightened U.S. export controls, according to the Financial Times. Banned B200 processors are sold on China’s black market despite U.S. restrictions aimed at limiting advanced AI development. Nvidia warned of technical and financial inefficiencies with unauthorized products. The U.S. may expand export controls to countries like Thailand by September to prevent further leakage.

Why It Matters

1. Geopolitical Impact: Highlights the escalating U.S.-China tech rivalry over AI hardware.
   

2. Supply Chain Risks: Shows how critical AI chips like B200s can bypass export controls, fueling unregulated development.
   

3. Market Shifts: Potential new restrictions may affect global AI infrastructure, particularly in Southeast Asia.

🧠RESEARCH

Pixels, Patterns, but No Poetry: To See The World like Humans

The paper introduces the Turing Eye Test to measure how well AI models see like humans. It shows that today's top models fail badly on simple visual tasks. Unlike reasoning, the real issue lies in how the models process images—not in their language skills—highlighting a major gap in human-like perception.

Yume: An Interactive World Generation Model

Yume is a model that turns images, text, or videos into interactive and realistic virtual worlds. Users can explore these worlds with keyboard actions. It combines video diffusion techniques, artifact reduction, and fast sampling to produce smooth, infinite video environments. The project is open-source and updated regularly.

DesignLab: Designing Slides Through Iterative Detection and Correction

DesignLab is an AI tool that creates and improves presentation slides through repeated review and correction. It uses two AI roles—a reviewer to find design flaws and a contributor to fix them—allowing slides to become more polished with each cycle. Tests show it outperforms existing slide design tools.

🛠️TOP TOOLS

ArtGuru Face Swap - AI tool designed to make face swapping in photos an effortless and enjoyable process.

Nokemon - A tool that leverages advanced ML technology to create unique and customizable Pokémon designs, often referred to as “Fakémon.”

NeuralBlender - Image-generation website that leverages the power of AI to create stunning images from textual descriptions.

CrushOnAI - AI chat platform that specializes in providing unrestricted, NSFW (Not Safe For Work) conversations.

Charley AI - AI-powered content generation platform designed to assist with academic and professional writing.

📲SOCIAL MEDIA

🗞️MORE NEWS

• Elon Musk announced that X, his social media company, will revive the discontinued video app Vine in an AI-powered version. He shared no details, but the six-second format suits short AI-generated videos.

• Copilot Appearance is an experimental feature on copilot.microsoft.com that adds voice, facial expressions, and animations to chats. Available to select users in the US, UK, and Canada, it aims to create more interactive, engaging conversations with Copilot.

• Claude Code’s sub agents are specialized AI assistants designed for specific tasks, each with its own context, tools, and custom prompts. They improve focus, efficiency, and reusability by delegating targeted workflows, like debugging, code reviews, or data analysis.

• Freed AI, a medical transcription startup, now serves 20,000 clinicians, saving them hours of documentation each day. Despite rising competition, Freed focuses on small healthcare practices, offering tailored AI-generated clinical notes, HIPAA compliance, and a growing $20M annual recurring revenue.

• Microsoft CEO Satya Nadella addressed employee concerns over recent layoffs, acknowledging the difficulty of the decisions but emphasizing Microsoft’s ongoing AI-driven transformation. He outlined three priorities — security, quality, and AI — while urging staff to embrace a growth mindset amid change.

• Memories.ai, a video AI startup, raised $8 million led by Susa Ventures and Samsung Next. Its platform processes up to 10 million hours of footage, enabling searchable indexing, summaries, and insights for security and marketing companies.