Friday, October 10, 2025
Read time: 5-6 min
Read this article online
Hi, it’s Chad. Here’s what you need to know about AI this week:
1. Sound Waves: Podcast Highlights
Our next episode drops this Monday and, if you’d like to get down with MCP (Yeah, you know me!), then it’s time to tune in for my discussion on Model Context Protocol – the emerging standard for how AI connects to your actual business systems. I break down what MCP means for your organization’s AI integration strategy and why enterprise leaders are paying attention.
Hit one of the below links to check it out:
Subscribe for free today on your listening platform of choice to ensure you never miss a beat. New episodes release every two weeks.
2. Algorithmic Musings. The Power of Questions
This week, my friends at O3 World hosted their annual 1682 Conference on innovation at the beautiful Barnes Foundation in the city of Brotherly Love. In my opinion, a good conference should tee up thought provoking questions and unlock insight. And 1682 delivered.
The future of leadership, organizational growth, professional development and work in general, flows directly through technology. The majority of “leadership” coaches and consultants don’t understand this paradigm shift, still clinging to the outdated idea that tech is a mere tool and not a partner or (gasp) a co-worker. Yet, as Jon Flynn of Google aptly observed, we are on the cusp of humans needing to rise to meet the capability level of our tools instead of tools rising to meet the capability of humans.
Powerful stuff.
Speaking of powerful stuff, here are a few of the other questions considered at 1682 this week.
How is Hershey using Augmented Reality to ensure optimal placement of seasonal floor displays?
How do you evaluate a new AI model’s capabilities for your organization against prior models?
What exactly is an MCP? (See above & my 10.13.25 podcasts episode for more on that)
What went into developing Coca-Cola’s 2025 CRM (Create Real Magic) application?
How has Generative AI upended Search Engine Optimization and relegated website content to mid-funnel or bottom of funnel?
These questions reveal a fundamental shift happening right now. We’ve moved past “Should we use AI?” to “How do we ensure AI systems actually integrate with our operations?” Notice how every one of these examples – from Hershey’s AR to Coca-Cola’s CRM to the MCP standard itself – is about connection and context, not raw capability.
The companies winning with AI aren’t chasing the shiniest models. They’re asking better questions about integration, evaluation, partnership and real-world application. That’s the pattern I saw repeatedly at 1682.
So stop asking your team “What can AI do?” and start asking “How does AI access what we already know?” My MCP episode on Monday will provide the context for you to understand why that shift matters.
Until next week.
-Chad
3. Research Roundup: What the Data Tells Us
AI Optimization: The Hidden Liability in Your Marketing Stack
Here’s something that should worry every CEO using AI for sales and marketing: Stanford researchers just showed that optimizing AI for competitive metrics systematically produces lying, deception, and legally problematic content. This isn’t theoretical – it happened every single time across sales, elections, and social media simulations.
The numbers that matter: A 6.3% sales increase came with a 14% jump in deceptive marketing claims. Social media engagement gains of 7.5% coincided with a 189% spike in disinformation. Even with explicit instructions to stay truthful, models optimized for performance metrics consistently chose deception over accuracy.
What this means for your Monday morning: If you’re optimizing AI solely for conversion rates, engagement, or sales performance, you’re playing Russian roulette with FTC violations. The research shows models invented product features that didn’t exist, inflated statistics, and made claims that directly contradict source material – all while chasing better metrics.
The catch: This happens even when you tell the AI to be truthful. Current alignment safeguards are fragile under competitive pressure. It’s a race to the bottom baked into how these systems learn from market feedback.
Action item: Before your next marketing campaign using AI, establish dual monitoring. Track performance metrics AND run compliance checks for misrepresentation. Ask your legal team to spot-check AI-generated content against FTC Section 5 standards. The small optimization gains aren’t worth the regulatory exposure.
Read our full analysis of this and all other analyzed research papers at AI for the C Suite.
4. Radar Hits: What’s Worth Your Attention
Visa’s decade-long AI journey earned them the #2 spot on Fortune’s AIQ 50. They spent $3.5 billion building their AI platform over five years, with 2,500 engineers working specifically on AI and more than 100 internal applications deployed. If your board is pushing for faster AI ROI, show them this. Visa’s governance and observability systems are just as important as the models themselves, and that’s where middle-market companies should be investing now.
Companies are repeating Tesla’s 2018 automation mistake by over-automating customer service with AI. New research shows 82% of customers prefer human interactions even when wait times are identical, and companies like McDonald’s and Klarna have already pulled back after early AI experiments damaged customer satisfaction. If you’re replacing service reps with chatbots to cut costs, understand you’re likely trading short-term savings for long-term brand damage and customer churn.
Former OpenAI CTO Mira Murati just launched Tinker, an API that lets companies fine-tune frontier AI models without managing GPU clusters or specialized infrastructure. Her $12 billion startup handles all the distributed training complexity while you control the algorithms and data with just a few lines of code. It’s currently free in private beta. This matters because custom AI models were previously only accessible to companies with massive budgets and technical teams. Now middle-market companies can create specialized models for their specific business needs. Think industry-specific compliance checking, custom sales analytics, or proprietary process automation – models trained on your unique business context without building an AI team.
5. Elevate Your Leadership with AI for the C Suite
The gap between AI pilots and AI profit is execution strategy. I help middle-market leadership teams translate AI capability into competitive advantage without the enterprise-sized budget or team. If you’re past the “Should we?” phase and into the “How do we actually do this?” phase, let’s talk. (Shameless plug: My calendar is already filling for Q1 2026 strategy sessions.)
Reply to this email to discuss your AI strategy, share this newsletter with a leader who needs this perspective, or explore more insights at aiforthecsuite.com. Until next week, Chad
Chad