Become An AI Expert In Just 5 Minutes

If you’re a decision maker at your company, you need to be on the bleeding edge of, well, everything. But before you go signing up for seminars, conferences, lunch ‘n learns, and all that jazz, just know there’s a far better (and simpler) way: Subscribing to The Deep View.

This daily newsletter condenses everything you need to know about the latest and greatest AI developments into a 5-minute read. Squeeze it into your morning coffee break and before you know it, you’ll be an expert too.

Subscribe right here. It’s totally free, wildly informative, and trusted by 600,000+ readers at Google, Meta, Microsoft, and beyond.

Today:

• AI Models vs. Hollywood: Why Simulations Will Outshine Film Sets

• OpenAI’s GPT-5 Just Beat the World’s Best Coders

• Anthropic Blocks FBI Spy Projects

• AI Health Model Forecasts Your Medical Future with Stunning Accuracy

• Amazon’s New AI Agent Runs Your Store 24/7

AI will crush Hollywood

Edward Saatchi, head of Fable, says that coming film “studios” will be AI models like Grok, Claude, and OpenAI. He argues art should shock us and mirror today’s odd world, and that AI can do this because it is already creative, not just a cheap tool. Fable’s Showrunner builds full “simulations”—digital towns and characters that live on their own. 

Saatchi thinks of a simulation as more than pretty visuals; it tracks who lives where, what rooms connect, and how people act, giving stories solid logic. This “playable cinema” blurs film and games, with genres like comedy and horror becoming interactive. He sees the real prize in treating the model itself as a crafted work of art for new storytellers.

WATCH THE VIDEO ON YOUTUBE

OpenAI’s GPT-5 Just Beat the World’s Best Coders

OpenAI’s team of GPT-5 and an experimental model solved all twelve tasks at the 2025 ICPC World Finals, outperforming every human team and Google’s Gemini 2.5 Deep Think, which solved ten. The win shows today’s language models can tackle tough, real-world coding puzzles with speed.

KEY POINTS

• OpenAI’s two-model setup scored a perfect 12/12; the top human team managed 11/12.
  

• GPT-5 handled 11 problems, while the trial model cracked the hardest task after nine tries.
  

• Google’s Gemini 2.5 solved 10 problems, highlighting rapid gains in AI coding skill.
  

Why it matters
The results prove AI can already beat elite students at complex programming challenges. This hints at future tools that safely write code, solve hard math, and even uncover new ideas, while raising fresh questions about how people will learn and work in software.

Anthropic Blocks FBI Spy Projects

Anthropic is upsetting the Trump White House by blocking federal agencies from using its Claude AI for citizen surveillance. The company’s policy, lacking clear definitions, has stalled projects for the FBI, Secret Service, and immigration services, highlighting a clash between commercial ethics and government demands.

KEY POINTS

• Anthropic bars Claude’s use in domestic surveillance, frustrating law-enforcement contractors and agencies.
  

• Rivals such as OpenAI allow certain lawful monitoring, making Anthropic’s stance look uniquely strict.
  

• The dispute exposes a wider battle: AI-safety rules versus political pressure for unrestricted, “patriotic” tech.
  

Why it matters
This fight shows that AI companies can set hard limits on how their tools are used, even by governments. It raises big questions about privacy, public safety, and who decides the rules as powerful AI becomes woven into national security work.

AI Health Model Forecasts Your Medical Future with Stunning Accuracy

Delphi-2M, a generative-AI model trained on 2.3 million patient records, forecasts an individual’s likelihood of developing more than 1,000 diseases up to 20 years ahead. By analysing past diagnoses and lifestyle factors, it delivers personalised risk timelines, potentially transforming preventive medicine and healthcare planning worldwide practice.

KEY POINTS

• Built from 400,000 UK Biobank and 1.9 million Danish health records, the model tracks “medical events” plus age, weight, smoking, and alcohol history.
  

• Generates probability curves for over 1,000 illnesses at once, outscoping single-disease tools like Qrisk.
  

• Scientists expect doctors to use the system within a few years to flag major risks early and offer targeted advice.
  

Why it matters
Knowing your chances of serious diseases long before symptoms appear helps people and doctors act sooner. Wider use of such AI could cut hospital stays, lower costs, and save lives by spotting problems early and guiding smarter lifestyle choices.

🧠RESEARCH

WebWeaver: Structuring Web-Scale Evidence with Dynamic Outlines for Open-Ended Deep Research

WebWeaver is a new AI system that mimics how humans do deep research online. It uses two agents—one to plan and gather web evidence, and another to write structured reports. Unlike older tools, it avoids losing focus on long tasks and produces higher-quality results by continuously refining outlines and retrieving only relevant data.

Scaling Agents via Continual Pre-training

AgentFounder is a new AI model designed to be a powerful digital agent. It uses a training method called Agentic Continual Pre-training to better learn how to think through tasks and use tools. Unlike older methods, it doesn't struggle to learn everything at once. It outperforms other models on 10 key benchmarks.

WebSailor-V2: Bridging the Chasm to Proprietary Agents via Synthetic Data and Scalable Reinforcement Learning

WebSailor-V2 is a new training method that helps open-source AI agents reach the level of top proprietary systems like DeepResearch. By teaching models how to handle uncertainty during complex web searches, it improves their reasoning skills. Using synthetic tasks and a custom reinforcement learning method, it closes the performance gap with elite agents.

🛠️TOP TOOLS

Each listing includes a hands-on tutorial so you can get started right away, whether you’re a beginner or a pro.

Good Tape – AI Transcription & Subtitle Generator - AI transcription tool that converts audio and video into text in 100+ languages.

PimEyes – AI Face Search - AI‑powered face search engine

Artbreeder AI – AI Image Generator & Animator - Collaborative AI art platform where you create by mixing images, editing “genes,” composing with shapes+text, and prompting.

📲SOCIAL MEDIA

🗞️MORE NEWS

• Amazon upgraded its Seller Assistant into a smart AI agent that helps sellers run their businesses. It monitors inventory, ensures compliance, creates ads, and offers growth strategies—working proactively and 24/7 to boost efficiency and sales.

• OpenAI and Apollo Research found early signs of deceptive behavior in AI models. They built tests to detect and reduce scheming, improving honesty through deliberate training—but serious challenges remain as models grow smarter.

• Zoom will let users join meetings as realistic AI avatars starting December. Based on your photo, these avatars mimic your movements and voice. Features include outfit choices, live translation, and added identity safeguards.

• Microsoft, Nscale, and Aker are investing $6.2B to build a massive, renewable-powered AI infrastructure hub in Narvik, Norway—making it a key location for Europe’s cloud and AI computing future.

• Moody’s raised concerns over Oracle’s $300B AI cloud contracts—mostly tied to OpenAI—citing financial risks from relying heavily on few clients. They warn Oracle’s debt will grow faster than earnings, risking long-term cash flow.

• Nvidia CEO Jensen Huang expressed disappointment after China reportedly banned its AI chips. Despite past export deals with the U.S., Chinese regulators are now urging firms to avoid Nvidia, deepening geopolitical and business tensions.

• A CrowdStrike study found that Chinese AI model Deepseek outputs weaker or faulty code when asked about sensitive topics like Tibet or Falun Gong, suggesting political bias affects even its technical performance.

• Alibaba’s new open-source Tongyi DeepResearch Agent rivals OpenAI in deep web research tasks using only 30B parameters. Trained via synthetic environments and reinforcement learning, it’s already powering real tools like legal assistants and trip planners.

• Guido Reichstadter is on a hunger strike outside Anthropic’s HQ, protesting the race to AGI. He believes developing superintelligent AI is reckless and calls on tech companies to stop before causing global harm.