Hey Radarians, MiniMax makes AI nearly 3x faster, while Amazon challenges Nvidia with a powerful AI dev tool.

From practical AI workflows and fascinating research to useful prompts, side hustle ideas, and tools worth trying.

Let's dive in.

⚡ MiniMax makes AI nearly 3x faster

• MiniMax launched M3.1 Flash on October 19, saying responses are about three times faster than M3 while AI agent tasks cost 70% less to run than its prior model
• On company tests, the model scored 73.7% on the DeepSWE coding benchmark and 82% on CyberGym, while improving its ability to operate PC apps and tasks
• Developers can use it through MiniMax's API and coding tools. Input costs $0.15 per million tokens, but its benchmark results still come from MiniMax itself

⚔️ Amazon takes aim at Nvidia with latest AI tool

• Amazon unveiled TorchNeuron, a tool that runs existing PyTorch AI code on its Trainium chips with one configuration change. No full rewrite is needed, AWS says
• Nvidia's CUDA toolkit has kept many AI developers tied to its GPUs. Amazon wants TorchNeuron to make switching hardware easier and give teams more chip choices
• Amazon says Deepgram processes 70% more audio using Trainium. TorchNeuron enters public beta on October 20, with support for Trainium2 and Trainium3

More News on the Radars

• ​Researchers say an AI trained on 40 million face images can estimate cancer survival from one photo, but needs more validation
• ​Inception Labs says Mercury 3 generates 1,068 tokens per second and matches or beats leading speed-optimized AI models
• ​OpenAI is accelerating a $30 billion funding round that could value it above $1.4 trillion before Anthropic's IPO

⚡ Steal This Workflow

Generate Personalized Feedback From Form Responses

Stop sending the same generic reply to everyone who fills out your forms. Use AI to analyze each submission and automatically deliver feedback tailored to their answers.

1. Create your feedback form​
   Open Google Forms and create a form with questions relevant to your goal.​
   Include fields such as:​
   ​→ Name and email address​
   → Main challenge or problem​
   → Goals or expectations​
   → Additional comments​
   Under Settings → Responses, enable email collection so you can send personalized feedback.
2. Connect your form to Make​
   Open Make and create a new scenario.​
   Add Google Forms → Watch Responses, connect your Google account, and select your form.​
   This triggers the workflow whenever Make detects a submission.
3. Generate personalized feedback with AI​
   Add OpenAI → Generate a Response. Connect your OpenAI API account and map the form answers into the prompt.​
   ​Use this prompt:​
   ​Analyze the following form submission and write personalized feedback based on the respondent's answers. Identify their main challenge, provide three practical recommendations, and suggest one next step. Keep the response friendly, concise, and actionable. Do not invent information. Responses: [INSERT MAPPED FORM ANSWERS].
4. Automatically deliver the feedback​
   Add Gmail → Send an Email.​
   Map the respondent's email address to the recipient field and the AI-generated feedback to the message body.​
   Run a test submission, review the generated email, and activate the scenario once everything works correctly.​
   Each submission can receive tailored recommendations without manually reviewing and replying to every response.​

🔎 AI RESEARCH: 91% of Anthropic Staff Expect AI to Make Nobel-Level Discoveries

What researchers studied:

Anthropic surveyed U.S. adults and its own employees in August 2026 to compare expectations about AI's future capabilities, productivity, and impact on jobs.

What they found:

91% of Anthropic staff believe AI will be capable of making Nobel-level scientific discoveries by 2030, compared with just 39% of U.S. adults.

Employees also expect AI to handle 75% of their current work hours by 2030, while U.S. workers estimate just 28% on average.

Why it matters:

People working closely with advanced AI expect much faster progress than the general public. The gap highlights how differently experts and everyday users perceive AI's future, although these predictions do not guarantee the technology will achieve those capabilities.

Source: Anthropic Research, October 16, 2026

🛠️ Tools Worth Watching

• ​AirJelly: Remembers your work and suggests helpful next steps
• ​MyLens: Creates personalized presentations that answer viewers' questions live
• ​Embeddable: Builds interactive quizzes and website widgets using AI

📝 Prompt Radar: Stress-Test a Business Idea Before Investing

What it does: Identifies weak assumptions, potential risks, and affordable ways to test a business idea before investing.

Best for:

• Startup validation
• Testing side hustle ideas
• Reducing investment risks

💼 AI Side Hustle Idea: AI Restaurant Menu Management Service

Restaurants often have menus scattered across websites, Google, and food delivery apps. Keeping everything accurate and updated takes time.

The problem:

Restaurant owners struggle with outdated prices, inconsistent descriptions, missing menu items, and translating menus for different customers.

Your AI service:

Use ChatGPT or Claude to rewrite menu descriptions, translate dish details, organize menu categories, and prepare consistent updates for websites, Google Business Profile, and delivery platforms.

Simple workflow:

Collect existing menus → check prices and ingredients → improve descriptions with AI → translate content → receive owner approval → update platforms → check accuracy

Best clients:

Restaurants, cafes, cloud kitchens, restaurant chains, and food delivery businesses.

What you can charge for:

One-time menu optimization + multilingual menu creation + platform updates + monthly menu management.

Learn AI with This Course

​🤖 Building Effective Human-Agent Teams by Anthropic​

What you’ll learn:

Learn how people and AI agents can work together as a team, assign responsibilities, share information, and build effective workflows with clear goals and human oversight.

Key topics:

• Understanding how AI agents collaborate with people and other team members
• Defining clear roles, shared goals, and information access for AI agents
• Building trust, assigning tasks, and gradually increasing AI agent responsibilities