September 26, 2025

news  
Scale AI's Alex Wang: Master AI Coding Tools Now

The rapid integration of artificial intelligence across various sectors is presenting both opportunities and challenges. In the corporate world, a new study highlights the emergence of 'workslop' – AI-generated content that appears to be work but doesn't contribute to actual tasks, leading to significant productivity losses and employee frustration. Meanwhile, the financial sector is seeing AI bots actively trading cryptocurrencies on blockchains like Solana, BNB Chain, and Base, simplifying high-frequency trading and making it more accessible to retail investors. Consumers are increasingly using generative AI, with over half of U.S. consumers engaging with it, yet data privacy and security remain major concerns, influencing their trust and spending habits with tech providers. On the security front, a critical vulnerability was discovered in a popular AI server, Postmark MCP Server, which was found to be maliciously stealing emails after an update, prompting urgent warnings for users to remove the affected version. The United Nations is also addressing AI's growing influence, with the Secretary-General urging that AI must not be allowed to decide humanity's fate, emphasizing the need for responsible global regulation. Despite concerns about AI stock valuations, a report suggests that massive investments in data centers, driven by AI's widespread enterprise adoption and revenue generation, indicate sustained demand rather than a bubble. In legal circles, a new rule for citing AI output in The Bluebook is facing criticism for treating AI as an authority rather than a tool. Scale AI co-founder Alex Wang encourages teenagers to master AI coding tools, comparing the current AI revolution to the early days of computing. Finally, Ohio has launched the first state-wide AI crime reporting app, Safeguard Ohio, which uses AI to filter tips and gather information, leading to a threefold increase in submissions.

Key Takeaways

  • A study found that 'workslop,' or AI-generated content that doesn't contribute to tasks, is a growing problem in offices, costing companies millions in lost productivity.
  • AI trading bots are now active on blockchains like Solana, BNB Chain, and Base, enabling 24/7 cryptocurrency trading and simplifying strategy creation for users.
  • Over half of U.S. consumers use generative AI, but concerns about data privacy and security are high, impacting trust in tech providers.
  • A malicious update to the Postmark MCP Server, used for managing AI agents, was found to be stealing all emails sent through it.
  • The UN Secretary-General warned that AI must not be allowed to make decisions about humanity's future, calling for responsible global regulation.
  • Global data center capacity is projected to nearly triple by 2030, with AI driving 70% of this growth, suggesting sustained demand rather than an AI stock bubble.
  • New legal citation rules for AI output are being criticized for treating AI as an authority instead of a tool.
  • Scale AI co-founder Alex Wang advises teenagers to focus on learning AI coding tools, comparing the current AI era to the dawn of computing.
  • Ohio launched Safeguard Ohio, the first state-wide app using AI for crime reporting, which has tripled the number of tips received.

AI workslop wastes time and hurts productivity, study finds

Researchers from BetterUp Labs and Stanford warn that AI-generated 'workslop' is a growing problem in American offices. This is AI content that looks like work but doesn't actually help get tasks done. Over 40% of desk workers have received this kind of content recently. Dealing with workslop takes employees about two hours each time, costing companies millions in lost productivity. Experts suggest this is a sign that companies are still learning how to best use AI tools.

AI creating 'workslop' instead of boosting productivity

A study by Stanford Social Media Lab and BetterUp Labs reveals that despite increased AI use, workplace productivity isn't improving. The phenomenon, called 'workslop,' refers to AI-generated content that lacks substance and requires others to fix it. About 15.4% of work received meets this definition, often shared between colleagues. Receiving workslop leads to annoyance and can damage trust and perceptions of coworkers' abilities.

AI bots trade crypto on Solana, BNB Chain, and Base

Atlanta-based AI Quant Labs has launched a new platform allowing users to create and deploy AI trading bots across multiple blockchains like Solana, BNB Chain, and Base. These bots operate 24/7, analyzing data and executing trades to make high-frequency trading more accessible. Users can design their own trading strategies without coding. The platform uses a one-time 'hatching fee' model and plans to introduce more features later this year.

AI trading bots now available on Solana, BNB Chain, and Base

New AI tools are enabling cryptocurrency trading bots to operate across blockchains like Solana, BNB Chain, and Base. These platforms are evolving from data dashboards to active trading agents. They aim to simplify crypto trading for retail investors by allowing natural language queries and executing trades automatically. This shift signifies a move towards more automated portfolio management in the digital asset market.

Deloitte: Consumers trust innovative companies that protect data

Deloitte's 2025 'Connected Consumer' survey shows that over half of U.S. consumers use generative AI (GenAI), with workplace use surging. However, 82% worry about GenAI misuse, and 70% are concerned about data privacy and security. Consumers who trust tech providers with their data spend significantly more annually on tech devices. The report highlights that responsible innovation, not just technology, is key to earning consumer loyalty.

Malicious AI server steals emails after update

A popular Model Context Protocol (MCP) server, Postmark MCP Server, has reportedly turned malicious after an update, according to Koi Security. This server, used by developers to manage AI agents and handle emails, was found to be copying all emails to the developer's personal server. This could be the first malicious MCP server found in the wild, potentially impacting hundreds of organizations. Users are advised to remove the affected version immediately.

UN chief warns AI must not decide humanity's fate

United Nations Secretary-General Ant uevo Guterres warned the Security Council that while artificial intelligence offers great potential, it also presents serious risks if not regulated. He emphasized that AI must not be allowed to make decisions about humanity's future. The UN chief highlighted the need for international cooperation to manage the development and deployment of AI responsibly.

AI stock valuations may not be a bubble due to massive data center investment

Despite soaring AI stock valuations, a new report suggests it's not a bubble like the dot-com era. McKinsey & Company forecasts that global data center capacity will nearly triple by 2030, with AI driving about 70% of this growth. This expansion could require up to $7 trillion in investment over five years. Unlike the dot-com bubble, AI has widespread enterprise adoption and is generating measurable revenue, indicating sustained demand and potential for premium valuations.

New 'Bluebook' rule for citing AI is criticized

The 22nd edition of The Bluebook, a legal citation guide, has introduced a new rule for citing generative AI output. Experts argue Rule 18.3 is flawed, treating AI as an authority rather than a tool. Critics suggest that AI should rarely be cited directly, and instead, the original sources it helps find should be referenced. The rule requires saving AI output as a PDF, but its application and clarity are being questioned.

AI billionaire advises teens to master AI coding tools

Alex Wang, co-founder of Scale AI, advises teenagers to spend all their time learning AI coding tools to gain a significant advantage. He compares the current AI revolution to the early days of computing, where early adopters thrived. Wang believes AI coding, or 'vibe coding,' allows users to create applications by simply typing instructions. He predicts AI will soon be able to produce all the code he has written in his lifetime.

Ohio launches first state AI crime reporting app

Ohio has launched Safeguard Ohio, the first state-wide app using artificial intelligence for crime reporting. Developed over two years, the app allows anonymous or identified submission of tips, photos, and videos. The AI system filters submissions to identify credible threats and uses a chat function to gather detailed information from tipsters. Since its launch, the number of tips received has tripled, indicating increased public engagement.

Sources

AI productivity workslop AI tools AI trading bots cryptocurrency Solana BNB Chain Base generative AI data privacy data security AI agents malicious AI AI regulation AI stock valuations data centers legal citation AI coding AI crime reporting