- NATURAL 20
- Posts
- Vibe Coding Gains Popularity Fast
Vibe Coding Gains Popularity Fast
PLUS: Perplexity Eyes TikTok Acquisition, AI Detects Cancer With 99% Accuracy and more.

Cold Email Setup Offer
We started sending 10,000 cold emails per day, and scaled a brand new B2B offer to $108k MRR in 90 days. Now, you can have the same system set up (completely done-for-you) inside your own business - WITHOUT going to spam, spending thousands of dollars, or any manual input. Close your next 20 clients easily. We’ll set up the tech, write your scripts, give you the leads, give you the inboxes, and the sending tool - all starting at $500/mo.
Today:
Vibe Coding Gains Popularity Fast
OpenAI, Meta Partner With Reliance
Tencent Unveils T1 AI Model
Manus Gains China’s AI Backing
Perplexity Eyes TikTok Acquisition
AI Detects Cancer With 99% Accuracy
VIBE CODING is Eating the World…
Vibe coding is a new trend where AI generates most of the code, popularized by Andrej Karpathy. Some embrace it, like Peter Levels, who built a successful game using AI, while critics argue it lacks depth.
A 2025 Vibe Coding Game Jam is planned, judged by Karpathy and John Carmack. The debate mirrors past shifts, like digital cameras disrupting film photography. While AI coding isn’t perfect, it's gaining traction.
OpenAI and Meta are in talks with India’s Reliance to expand AI in India. OpenAI wants Reliance’s telecom giant Jio to distribute ChatGPT, possibly at lower prices. Meta, already invested in Jio, is also pursuing AI collaborations. Reliance may sell AI products and offer data center capacity to both firms. A deal could help OpenAI and Meta strengthen ties with Indian businesses and officials while competing in a rapidly growing AI market.

Why It Matters
India’s AI Growth – India is OpenAI’s second-largest market, making it crucial for AI adoption and expansion.
Local AI Infrastructure – Reliance’s data centers could help AI firms run models efficiently in India.
Regulatory Influence – Partnering with Reliance may help foreign AI firms navigate India’s regulatory landscape.
Tencent has launched the official version of its T1 reasoning model, enhancing AI competition in China. T1 offers faster responses, improved long-text processing, and low errors. It runs on Tencent’s Turbo S model, which outperforms DeepSeek’s R1 in certain benchmarks. This comes as Tencent ramps up AI investments and increases spending for 2025. The move reflects China's push for homegrown AI models that rival Western technology at lower costs.

Why It Matters
Stronger AI Competition – Tencent’s T1 challenges global AI leaders and increases innovation in China.
Faster, Smarter AI – T1’s improved reasoning and efficiency could push AI capabilities forward.
Big AI Investments – Tencent’s funding signals long-term growth in AI infrastructure and development.
China is backing AI startup Manus, which gained global attention for launching an autonomous AI agent requiring minimal prompts. Following DeepSeek’s success, investors seek the next major AI disruptor. Manus' AI assistant Monica has cleared China’s regulatory hurdles, and state media is now promoting the company, signaling government support. A partnership with Alibaba’s Qwen AI team further strengthens Manus’ domestic rollout, amid rising competition in China's AI sector.
Why It Matters
China’s AI Push – Beijing is accelerating support for homegrown AI firms.
Autonomous AI Evolution – Manus' agent model suggests a shift beyond chatbots.
Tighter Regulations – AI firms must comply with strict Chinese content policies.
🧠RESEARCH
A new method, RSD, improves image super-resolution by speeding up diffusion models while keeping details realistic. It enhances the ResShift model through a smart training process, outperforming existing methods. RSD delivers high-quality images with fewer resources, making it a strong alternative to text-to-image-based super-resolution techniques.
This survey examines how AI agents using large language models (LLMs) are evaluated. It reviews benchmarks for key skills like planning and memory, application-specific tests, and general assessment frameworks. The paper highlights trends toward realistic evaluations, identifies gaps in safety and efficiency, and suggests improvements for future research.
This survey explores how to make large language models reason more efficiently. It examines ways to reduce unnecessary steps in reasoning, optimize model training, and adjust inputs for better performance. The paper highlights methods to cut computational costs while maintaining accuracy, offering insights into improving AI reasoning speed and efficiency.
🛠️TOP TOOLS
GPTKit - AI-powered text detection tool designed to distinguish between human-written and AI-generated content with high accuracy.
ImageToCartoon - AI-powered online tool that transforms photos into cartoon-style images quickly and easily.
Wonder Dynamics - AI-powered visual effects company that revolutionizes the film and entertainment industry.
Watermark Remover IO - AI-powered online tool designed to efficiently remove watermarks, logos, text, and other unwanted elements from images while preserving the original quality.
Typeframes - AI-powered text-to-video creation tool designed to simplify the process of producing engaging video content for platforms like YouTube, Instagram, and TikTok.
📲SOCIAL MEDIA
Seems like China might be using open source as a way to damage US tech.
(See @balajis recent post)
Unitree contributing to open source robotics.
Manus is planning to open source their AI agent tech.
DeepSeek, obviously.
— Wes Roth (@WesRothMoney)
8:36 AM • Mar 23, 2025
🗞️MORE NEWS
Perplexity, an AI search startup, wants to buy TikTok and rebuild its algorithm in U.S. data centers with government oversight. It plans to open-source the recommendation system, integrate search capabilities, and enhance personalization and multilingual support.
Scientists developed an AI model that detects cancer with 99.26% accuracy, far surpassing doctors. The tool quickly identifies early-stage cancer, improving treatment chances. It’s not a replacement for doctors but enhances diagnosis and monitoring efficiency.
Builder.ai, a UK tech start-up backed by Microsoft, used an auditor linked to its founder, raising conflict of interest concerns. The company also frequently changed auditors for overseas subsidiaries, prompting questions about its financial oversight.
MIT researchers developed HART, a new AI tool that merges two methods to generate high-quality images nine times faster. It uses less energy, runs on laptops or phones, and could improve robotics, gaming, and vision-language models.
Cloudflare's AI Labyrinth tricks web-scraping bots by leading them through AI-generated decoy pages. This wastes their resources while helping Cloudflare identify bad actors. The tool aims to curb unauthorized AI training data collection.
What'd you think of today's edition? |
Reply