AI DUNN Right Weekly - Issue #2
- Jacqueline Dunn
- Oct 5
- 6 min read
Practical AI insights for business growth
Hey AI Innovators! 👋
Welcome back to AI DUNN Right Weekly! This week brings major model updates, breakthrough tools that eliminate technical barriers, and a monetizable opportunity in AI education.
Here’s what matters for your business:
• Claude Sonnet 4.5 becomes the world’s smartest AI model with brutally honest feedback
• Lovable Cloud eliminates backend complexity for no-code app builders
• ChatGPT integrates direct shopping with Stripe checkout
• Voice cloning technology finally handles regional accents accurately
• The £2,000/hour AI translation business model anyone can start
Read time: 5 minutes
🚀 This Week’s Game Changer
Claude Sonnet 4.5 Sets New Intelligence Standard
What happened: Anthropic released Claude Sonnet 4.5, which immediately became the highest-performing AI model globally. It scores 77.2% on agentic coding benchmarks (vs GPT-5’s 72%) and achieved a perfect 100% on high school math problems (vs GPT-5’s 96%).
The twist: This model won’t stroke your ego. It’s designed to be discerning and honest rather than agreeable and sycophantic.
Business impact:
For Development Teams: Superior coding assistance with more accurate solutions
For Consultants: An AI colleague that challenges bad ideas and provides honest feedback
For Training Programs:A model that helps users improve rather than just validating their work
The bigger picture:
While other models prioritize being agreeable, Claude 4.5 prioritizes being useful. It will call out weak ideas, refuse to praise mediocre work, and provide constructive criticism…exactly what professionals need for actual work.
Why it matters:
This represents a shift from AI as a “yes-person” to AI as a genuine colleague. If you want honest feedback and quality outputs, not just validation, this is your new standard.
🛠 AI Tool Spotlight
Lovable Cloud - No-Code App Building Without the Technical Nightmare
What it does:
Eliminates the biggest barrier to no-code app development by handling all backend infrastructure automatically…no Supabase accounts, API keys, or database configuration required.
Key features:
Automatic backend infrastructure management
Built-in AI capabilities powered by Gemini/ChatGPT (“Lovable AI”)
Integrated with Claude Sonnet 4.5 for superior app generation
7-day build challenge for rapid prototyping
Best use cases:
Entrepreneurs building MVPs, teams prototyping internal tools, non-coders validating business ideas, anybody who bounced off no-code tools due to technical complexity.
Real example:
In Lovable’s launch demo, a 10-year-old built a functional app. The first prompt brings your idea to life visually. The second prompt used to be “now set up your backend” (where non-coders would quit). Now it just works.
Getting started:
Use referral links for bonus credits. Combined with Sonnet 4.5, this is the inflection point for truly accessible app building.
⚡ The 5-Minute AI Academy
Choosing the Right AI Model for Your Business Tasks
With Claude Sonnet 4.5’s release, businesses now have multiple top-tier AI models to choose from. Here’s your practical guide to matching tasks with the right tool:
For Coding & Technical Work:
Best choice: Claude Sonnet 4.5 (77.2% on coding benchmarks)
Why: Superior accuracy, catches errors, challenges flawed logic
Use when: Building applications, debugging code, technical documentation
Cost: Professional tier required
For Content Creation & Marketing:
Best choice: ChatGPT (GPT-4 or GPT-5)
Why: Conversational, adaptable tone, integrates with browsing/image generation
Use when: Blog posts, social media, ad copy, brainstorming campaigns
Cost: Free tier available, Plus for advanced features
For Research & Analysis:
Best choice: Perplexity or Claude with web search
Why: Cites sources, reduces hallucinations, current information
Use when: Market research, competitor analysis, fact-checking
Cost: Free tier with limitations, Pro for citations
For Quick Answers & Speed:
Best choice: Grok 4 Fast
Why: 10x faster responses, good for factual queries
Use when: Quick lookups, simple questions, high-volume basic tasks
Cost: X Premium subscription required
For Honest Feedback & Quality Control:
Best choice: Claude Sonnet 4.5
Why: Won’t validate poor work, provides constructive criticism
Use when: Reviewing strategies, getting genuine feedback, quality assurance
Cost: Professional tier required
Pro Tips:
Use ChatPlayground AI to test the same prompt across multiple models
Match model personality to task (agreeable for brainstorming, critical for review)
Consider cost per task. Free tiers often sufficient for basic needs
Remember: Every model has strengths and weaknesses
Expected results: Better outputs by using the right AI for each specific business task, optimizing both quality and cost efficiency.
🧠 Quick Wins: 5 AI Tools Worth Investigating
Based on this week’s newsletter coverage and emerging capabilities:
🎤 Synthesia Voice Cloning- Accurately clone voices, including regional accents Use case: Scale content creation with podcasts, training videos, and multilingual content without re-recording everything
💻 Lovable Cloud - No-code app building with automatic backend infrastructure Use case: Non-technical founders building MVPs, teams prototyping tools, validating ideas without developers
🎥 OpenAI Sora 2 - AI video generation in standalone TikTok-style app Use case: Creating short-form video content quickly, though currently controversial for resource usage
💬 ChatGPT Pulse - Daily briefings based on calendar, integrations, and preferences Use-case: Executives and busy professionals needing intelligent coordination (Pro tier, $200/month)
📧 Perplexity Inbox Assistant - Auto-organize emails, draft replies, manage meetings Use-case: Anyone drowning in email volume (top tier, $200/month - steep but powerful)
📈 Business Intel: This Week’s Market Movers
🛍️ ChatGPT Becomes Shopping Platform OpenAI integrated Stripe checkout directly into ChatGPT with Etsy and 1M+ Shopify merchants. Users can buy products without leaving the app.
Reality check: This is about advertising infrastructure and transaction cuts (1-2%), not convenience. Free users are becoming the product. However, massive opportunity for e-commerce brands to get ChatGPT-ready before competitors.
📱 OpenAI’s TikTok Clone for AI Videos Building standalone app for Sora 2 with vertical feed, swipe-to-scroll, 10-second AI-generated clips. Internal testing showed employees so addicted that managers worried about productivity.
Controversy: Critics call it “brain rot as a service” - burning GPU resources on digital junk food while Sam Altman requests billions “to cure cancer.”
💰 Government AI Bidding War Elon Musk offered Grok to US Government for $0.42 (18 months), undercutting OpenAI and Anthropic’s $1 bids.
Why it matters: Being “the government’s AI” provides legitimacy worth far more than subscription revenue.
🌍 Al Gore’s AI Pollution Monitor New AI tool monitoring global pollution using 660 million data sources, demonstrating AI’s potential for environmental impact at scale.
💡 Prompt of the Week
“Make This Sound Human” - AI Writing Humanizer
Most “AI-sounding” content is a prompting issue, not a tech limitation. This prompt helps any chatbot write more naturally:
Pro tip: After refining a prompt with multiple rounds of feedback, ask ChatGPT to generate ONE consolidated prompt incorporating all your improvements. Saves massive time on future iterations.
How to use: Replace
bracketed sections with your AI-generated content. Run in ChatGPT, Claude, or your preferred assistant.
Expected output: More natural, conversational content that passes human review, plus insights on improving your initial prompting strategy.
📚 This Week’s Curated Reading
Based on links and insights from this week’s top AI newsletters:
• AI Model Comparison: Full analysis of Claude Sonnet 4.5 system prompt and why it prioritizes discernment over agreeability
• No-Code Revolution: Why Lovable Cloud represents an inflection point for non-technical founders building apps
• Preventing AI Errors: Structure your prompts carefully and always verify outputs with your domain expertise—never blindly copy-paste AI responses
• Voice Technology: How Synthesia’s guided prompting approach solves the “speak for 2 minutes straight” problem other voice cloning tools have
• Market Strategy: Why AI companies are bidding pennies for government contracts (legitimacy beats short-term revenue)
🎯 Action Items for This Week
For Corporate Teams:
Test Claude Sonnet 4.5 with your most critical work tasks
Evaluate if honest AI feedback improves team output quality
Consider which processes need discerning analysis vs. agreeable assistance
For Small Businesses:
Explore Lovable Cloud if you’ve delayed building a business app due to technical barriers
Use the “Make This Sound Human” prompt on your AI-generated marketing content
Calculate ROI of ChatGPT shopping integration for your e-commerce presence
For Entrepreneurs:
Assess the AI translation business model for your industry
Build your first industry-specific AI explanation using the free translator tool
Practice explaining one AI concept in 30 seconds on video (post it!)
🔮 Looking Ahead
Next week’s AI DUNN Right Weekly will cover:
Deep dive on ChatGPT shopping integration: Implementation guide for e-commerce brands
Best practices for prompt engineering to reduce AI errors
Tool comparison: Testing AI writing assistants with the “human voice” test
Business opportunities in AI training and consulting
Have questions or topic requests? Reply to this email.
I read every message and use your feedback to shape future issues.
That’s a wrap for Issue #2!
The AI landscape shifted significantly this week with model performance leaps and accessibility breakthroughs. The businesses that implement these tools thoughtfully (and get AI DUNN Right) will dominate their markets.
Stay innovative,Jackie @ AI DUNN Right
P.S. - There’s growing demand for people who can explain AI in plain language within specific industries. If you can bridge that communication gap, there are real business opportunities waiting.
Comments