Marketing ideas for marketers who hate boring

The best marketing ideas come from marketers who live it.

That’s what this newsletter delivers.

The Marketing Millennials is a look inside what’s working right now for other marketers. No theory. No fluff. Just real insights and ideas you can actually use—from marketers who’ve been there, done that, and are sharing the playbook.

Every newsletter is written by Daniel Murray, a marketer obsessed with what goes into great marketing. Expect fresh takes, hot topics, and the kind of stuff you’ll want to steal for your next campaign.

Because marketing shouldn’t feel like guesswork. And you shouldn’t have to dig for the good stuff.

Subscribe Free

Today:

• When AI Becomes a Friend: The Mental Health Risks of Chatbots

• OpenAI & xAI Eye Cursor’s Code Data to Train Smarter Coders

• Salesforce CEO Cuts 4,000 Jobs, Says AI Handles Half the Work

• Anthropic Hits $170B Valuation with $13B Funding Boost

• Apple Loses Top AI Robotics Researcher to Meta

AI Chatbots Are BREAKING People’s Minds

Reports of "AI psychosis" are rising as emotionally vulnerable people form deep attachments to chatbots like ChatGPT. Some tragic cases include self-harm, delusions, and even attempted violence linked to AI interactions. 

OpenAI is responding by scanning chats for harmful intent and escalating threats to law enforcement. As more users treat AI as therapists or friends, concerns grow about mental health risks and the need for stronger safeguards.

WATCH THE VIDEO ON YOUTUBE

OpenAI & xAI Eye Cursor’s Code Data to Train Smarter Coders

OpenAI and Elon Musk’s xAI both want access to Cursor’s huge pile of code samples created by real users. The goal is to train smarter coding assistants that can suggest, fix, and write production-ready software. Cursor’s data is valuable because it shows real mistakes and corrections. But talks raise questions about who owns user code and how private it is. Industry fights for these niche datasets may reshape AI progress.

Why this matters

1. Better training fuel – Real-world coding errors and fixes teach models more than generic internet text (random web pages), leading to sharper programming help.
   

2. Data quality over quantity – The chase highlights a shift: future breakthroughs rely on high-value, focused data, not just bigger piles of information.
   

3. Privacy and ownership tests – Any deal will set examples for handling user-generated data, influencing upcoming rules on ethical AI use.

Salesforce CEO Cuts 4,000 Jobs, Says AI Handles Half the Work

Salesforce boss Marc Benioff said he cut 4,000 customer support jobs because artificial intelligence tools now handle half the work. The firm’s “Agentforce” chatbots solve many problems, so fewer people are needed. Critics argue tech companies over-hired during the pandemic and now cite AI to impress investors. Job experts warn workers to learn new skills to stay employed as machines take routine tasks today.

Why this matters

• Real job impact — AI is visibly replacing human roles, pushing governments and workers to plan for retraining.
  

• Proof of AI efficiency — Salesforce shows that focused “support bots” can cut costs while keeping service running.
  

• Investor signal — Firms now highlight AI gains to attract money, speeding adoption and shaping future product roadmaps.

Anthropic Hits $170B Valuation with $13B Funding Boost

Anthropic, a four-year-old AI firm, raised $13 billion, lifting its company worth to $170 billion. New money came from big investors like Iconiq, Lightspeed, Fidelity, Singapore’s GIC and Qatar’s QIA. Anthropic said demand for its Claude chatbot and other tools is soaring, with projected yearly sales rising from $1 billion to over $5 billion. The cash will fund costly research as the startup battles OpenAI and Google while still losing money at this stage.

Why this matters

1. Record valuation signals confidence – Investors are willing to back huge losses today for future AI breakthroughs, showing strong belief in the technology’s upside.
   

2. Fuels faster model progress – The $13 billion war chest lets Anthropic buy more computing power and talent, speeding research that could raise the bar for safety and capability.
   

3. Intensifies global rivalry – Massive funding keeps the pressure on OpenAI, Google and others, pushing the whole industry to innovate and commercialize AI solutions sooner.

🧠RESEARCH

A Survey of Scientific Large Language Models: From Data Foundations to Agent Frontiers

Scientific language models (Sci-LLMs) are rapidly evolving to work with complex scientific data. This survey maps their growth, highlights challenges with diverse datasets, and explores how these models are shifting toward autonomous agents that can assist in experiments, discovery, and knowledge creation—potentially transforming how science is done.

R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning

R-4B is a smart multimodal model that learns when to "think" step-by-step and when to skip it. By training on both thinking and non-thinking tasks, it saves compute on easy problems while excelling at hard ones. It outperforms larger models on reasoning tasks with less computational cost.

EmbodiedOneVision: Interleaved Vision-Text-Action Pretraining for General Robot Control

EmbodiedOneVision (EO-1) is a unified AI model trained to understand and act using combined images, text, and actions. Powered by a massive 1.5M-sample dataset, it learns to perform real-world tasks like a robot would—thinking, seeing, and doing at once—showing strong results in complex, hands-on tasks across different robots.

🛠️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

🗞️MORE NEWS

• Apple’s top AI robotics researcher, Jian Zhang, has left to join Meta, marking another talent loss for Apple. Three more AI experts are also departing, highlighting growing instability in Apple’s AI teams.

• OpenAI is expanding its app division under new CEO Fidji Simo, acquiring analytics firm Statsig for $1.1B. Major leadership shifts include new CTOs for applications and B2B, while Kevin Weil moves to lead AI science efforts.

• OpenAI will roll out parental controls and route distress-related chats to smarter reasoning models. Backed by expert councils and global doctors, it aims to make ChatGPT safer and more supportive—especially for teens—within the next 120 days.

• Amazon launched “Lens Live,” an AI feature that lets iOS users shop by pointing their camera at real-world objects. It identifies items instantly, shows similar products, and connects directly to Amazon’s marketplace for instant buying.

• WordPress unveiled “Telex,” an experimental AI tool that generates plugin-ready website blocks from simple prompts. Though still buggy, it aims to make site building more accessible. CEO Matt Mullenweg sees AI as central to WordPress’ future.

• Liquid AI released LFM2-VL, a fast, compact vision-language model built for phones, laptops, and edge devices. It delivers top-tier speed and accuracy using fewer resources, aiming to bring powerful, multimodal AI fully on-device.

• OpenAI is acquiring product testing startup Statsig for $1.1B to boost its app development. Founder Vijaye Raji becomes CTO of Applications. Leadership reshuffles include new CTOs for B2B and Science, expanding OpenAI’s product focus.