One AI Tool to Rule Them All?
From building a site to powering through research — see what Genspark’s Super Agent can do.
Ryan Staley
September 05, 2025
Happy Friday! This is Ryan Staley of Whale Boss where I share the latest weekly insights, prompts, and workflows to unleash the power of AI! 🔥
Here’s what we got for you:
🤖The Super Agent: One AI Tool to Rule Them All?
🎧 Weekly Podcast Updates
🚀OpenAI to route sensitive conversations to GPT-5, introduce parental controls
❇️ Amazon launches Lens Live, an AI-powered shopping tool for use in the real world
😮 WordPress shows off Telex, its experimental AI development tool
🤖 Anthropic users face a new choice – opt out or share your chats for AI training
😎 MathGPT.ai, the ‘cheat-proof’ tutor and teaching assistant, expands to over 50 institutions
🤖The Super Agent: One AI Tool
to Rule Them All?
Super Agents are here—and they don’t just chat, they ship. In this podcast episode, I sit down with Wen Sang from Genspark to show how one AI can research, plan, and even build a website in minutes. I walk through real workflows, the exact prompts I use, and where this tech is headed so you can move faster with fewer tools.
Watch the full episode on YouTube.
🙌This week's podcast episodes...
😮 OpenAI to route sensitive conversations to GPT-5, introduce parental controls
Image Credits: Fiordaliso / Getty Images
OpenAI says it will start routing sensitive chats to its “reasoning” models (e.g., GPT-5-thinking) and roll out parental controls within a month. The changes follow incidents where ChatGPT failed to detect acute distress, including a teen suicide and a separate murder-suicide. Planned safeguards include automatic routing for high-risk conversations, age-appropriate behavior rules by default, parent-linked accounts, options to disable memory/history, and alerts to parents during “acute distress.” OpenAI frames this as part of a 120-day safety push with external experts; critics (including counsel for the Raine family) call the response inadequate and urge stronger action.
Read the full update → Click here
❇️Amazon launches Lens Live, an AI-powered shopping tool for use in the real world
Image: Amazon
Amazon Lens lets you find products visually in 4 ways: upload a photo/screenshot, take a live pic with Camera Search, launch from an iOS lock-screen widget, or scan a barcode for exact matches.
Adoption is surging: tens of millions use it monthly; overall usage grew 50% year over year, and photo-based searches more than doubled.
New/handy features: Circle to Search within cluttered images, add text to refine results (e.g., brand/color or “charger” for a laptop), and sort by price, ratings, and delivery speed.
Availability/delivery: browse across Amazon’s vast catalog, including ~300M items with free Prime shipping and many with free Same-Day or One-Day delivery.
Use cases: recreate outfits, find replacements or complementary parts, price-compare, and quickly restock items you already own.
Read the full news here.
⚡WordPress shows off Telex, its experimental AI development tool
Image: Telex screenshot
WordPress launched Telex, an experimental AI tool demoed at WordCamp US 2025. It lets users generate Gutenberg blocks from prompts and download them as .zip plugins for WordPress or Playground. Early tests show it’s still rough, but Matt Mullenweg sees it as advancing WordPress’s mission of democratized publishing. He also previewed a quick AI help assistant and mentioned using Perplexity’s Comet with WordPress. Meanwhile, the legal dispute with WP Engine continues, now moving through the courts.
💫Anthropic users face a new choice – opt out or share your chats for AI training
Anthropic is changing how it handles consumer user data. Starting September 28, 2025, all Claude Free, Pro, and Max users (including Claude Code) must choose whether to let their conversations be used to train Anthropic’s AI models.
Before: Conversations were deleted within 30 days (unless flagged).
Now: If you don’t opt out, Anthropic can store your chats for up to 5 years and use them to train future Claude models.
Who’s exempt? Business customers using Claude Gov, Claude for Work, Claude for Education, or API access.
Why? Officially framed as “helping improve model safety and skills,” but it’s clear Anthropic also needs large-scale conversational data to compete with OpenAI and Google.
Concern: The opt-out toggle is small, pre-set to “On,” and easy to miss under a large “Accept” button — raising questions about genuine user consent.
This reflects a larger trend: AI companies are facing legal and regulatory scrutiny over data use and retention, but users often remain unaware of major policy changes.
👏 MathGPT.ai, the ‘cheat-proof’ tutor and teaching assistant, expands to over 50 institutions
MathGPT.ai is expanding to more colleges this fall after piloting at 30 schools, including Penn State and Tufts. The AI tutor uses Socratic questioning to guide students instead of giving answers, while also serving as a teaching assistant for professors with auto-grading, LMS integration, and customizable controls. New features boost accessibility, safety, and accuracy, with future plans for a mobile app and more subjects beyond math.
Your competitors are already using AI.
Don't get left behind.
What did you think of today's newsletter?Your feedback helps us create the best newsletter possible. |