TL;DR
- Not a wave but a tide change. Decide now whether you are building harbors or clinging to docks.
- Current models are not general intelligence. They are savants in specific high-value tasks, which is where immediate tangible value sits.
- The friction is not capability, it is integration. Slotting AI into existing workflows is genuinely hard, and being good at it is the competitive differentiator.
- Profit pools are reconfiguring across the value chain, away from labor-intensive activity and toward whatever is heavily augmented.
Just when we thought of it as yet another tech trend, AI has already started rewriting the rules of the game: For how value is created and where the profits land. A complete shake-up of almost every industry's value chain is happening, and you need to be ready.
What You Need to Know Now
Dario Amodei gives us some solid ground to stand on, in understanding AI's true impact.
- AI's Got "Savant Spikes," Not General Genius (Yet): Forget the sci-fi movies for a sec. Current AI models aren't human-level general intelligence. Think of them more like an intern who happens to be a "savant" in very specific, high-value tasks. They can ace standardized tests or handle specific writing assignments with superhuman ability, but they can still stumble on broader, open-ended problems. What does this mean for you? Immediate, tangible value comes from using AI for targeted, repetitive, or data-intensive tasks where its specialized strengths shine. It's about augmenting your team, not replacing them wholesale.
- The Future is Fast: Scaling Laws Are Real: Amodei is clear: AI capabilities are going to keep rocketing forward. We're talking about "scaling laws" driving "across-the-board improvements," meaning even current weak spots will get stronger. The idea of a "human-level" AI model being "quite close"—potentially within two or three years—isn't far-fetched. This isn't just about faster research; it could mean AI becomes a "main contributor to scientific progress." For your business, this means a rapidly evolving landscape. Your five-year plans? They just got a lot shorter. Agility isn't a buzzword; it's survival.
- "Mysterious Frictions" Are Your New Battleground: Here's the critical bit: even with super-smart AI, there are "mysterious frictions" that stop it from just instantly transforming your bottom line. It's tough to just "slot AI into existing company workflows." Why? Because truly transformative AI needs to "substantially accelerate the labor of thousands of top experts," not just outperform the average human. Overcoming these integration challenges – the workflow redesign, the organizational readiness, the change management – is going to be your biggest competitive differentiator. It’s not just about having AI; it’s about being damn good at using it.
Where the Profits Will Shift – A Value Chain Deep Dive
AI isn't just tweaking your operations; it's fundamentally altering where value is created and captured in every part of your business. Let's break down the "profit pools":
- Research & Development (R&D): AI can become the "main contributor to scientific progress." Think accelerated discovery, faster hypothesis generation, and predicting complex outcomes. This means new products and services hitting the market faster, boosting your revenue and profit margins from differentiated offerings. R&D moves from cost center to primary profit driver.
- Supply Chain & Operations: Imagine AI handling "extended tasks" and processing "trillions" of data points. This translates to incredibly accurate demand forecasting, real-time logistics optimization, and predictive maintenance. The result? Significant cost reductions, improved agility, and a much more resilient supply chain. Operational efficiency directly pumps up profit margins.
- Manufacturing & Production: AI can deliver "superhuman performance" in quality control and process optimization. This means real-time adjustments on the factory floor, minimizing downtime, and maximizing output. You're looking at improved product quality, increased throughput, and lower production costs, all leading to fatter gross margins and better capital efficiency.
- Marketing & Sales: AI's ability to process "vast data" and do "constrained writing tasks" means hyper-personalized campaigns, smarter sales assistants, and dynamic pricing. This drives higher conversion rates, increases customer lifetime value, and optimizes your marketing spend, leading directly to revenue growth and better sales margins.
- Post-Sales Service & Support: AI chatbots with "savant spikes" can handle routine queries instantly, while sentiment analysis predicts customer issues before they even arise. The payoff? Enhanced customer satisfaction, drastically reduced support costs, and improved retention, leading to lower churn and new upsell opportunities.
- General & Administrative (G&A) Functions: From HR to finance, AI can automate tons of routine tasks – invoice processing, resume screening, fraud detection. This boosts efficiency and accuracy, slashes G&A expenses, and gives you better data for decision-making across the board, directly contributing to higher net profit margins.
Strategic Implications & Your Next Moves
Here’s the blunt truth:
- Value Capture is Reconfiguring: Profits are shifting away from traditional labor-intensive activities toward those heavily augmented or automated by AI, or toward the AI providers themselves. Control over massive, high-quality datasets and compute infrastructure is becoming a new source of competitive advantage.
- New Profit Pools are Emerging: You'll see entirely new profit pools around AI development, infrastructure, specialized models, and critical AI integration services. Traditional activities will face margin compression. Decide now if you want to compete in these new high-margin pools or leverage AI defensively.
- Disruption is Inevitable: The speed of AI advancement means your current competitive advantages are fleeting. If you're slow to adapt, agile, AI-native startups or existing competitors will eat your lunch. Static strategies will lead to value erosion.
Your Action Plan – Direct and Decisive
- Get Your AI Roadmap in Order: Start with quick wins using AI's "savant spikes," but build out your foundational capabilities (data infrastructure, AI talent) for bigger plays down the line.
- Tackle Those Frictions Head-On: Redesign your workflows, implement robust change management, and focus on seamless human-AI collaboration. This is where the rubber meets the road.
- Build Internal AI Muscle: Don't just outsource. Develop your own expertise in data science, prompt engineering, and AI governance. This "AI integration competence" is your superpower.
- Embrace Continuous Learning & Agility: The AI landscape is a moving target. Set up systems for constant monitoring, rapid prototyping, and agile deployment. Experiment, learn, and adapt fast.
- Reskill Your Workforce: Prepare your team to work with AI, not against it. Focus on human skills like critical thinking, creativity, and complex problem-solving.
- Stay Ahead on Policy & Ethics: AI regulation is coming. Engage proactively to shape the environment and build trust.
The bottom line? AI is creating unparalleled opportunities. The companies that navigate these "mysterious frictions" effectively and integrate AI deeply into their operations will not just survive; they will thrive, capturing the lion's share of value in this dynamic new economy. It's time to act.
Enjoy your Sunday!
Vinod Venkatesan, Head of AI Staff & Engagements, herds the smartest cats at Moative, the quant-grade AI consulting studio.
Where Moative fits
Moative works on the friction rather than the capability, because capability is increasingly something everyone can buy and friction is specific to your workflows and your people. That is why engagements tend to start with a redesign of how the work flows rather than with a model selection. In healthcare revenue cycle management the margin moves when the handoffs between clinical documentation, coding and submission stop losing information, which is an integration problem before it is an AI one.
Frequently asked questions
Where does AI create the most value in a business?
In targeted, repetitive or data-intensive work where current models genuinely outperform, rather than in broad open-ended judgment. Across the value chain that means forecasting and maintenance in operations, quality control in production, personalization in sales, and routine handling in support and administration.
What are the "mysterious frictions" in AI adoption?
The gap between a capable model and a changed profit and loss. It consists of workflow redesign, organizational readiness and change management, and it is difficult enough that being good at it is a durable differentiator.
Will AI replace jobs or augment them?
In the near term, augment. Current systems are savants at specific tasks and unreliable across broad ones, so value comes from pairing them with people rather than substituting wholesale. That balance will shift, and it has not shifted yet.
How are AI profit pools changing?
Value is moving away from labor-intensive activity toward augmented activity and toward the providers of models, infrastructure and integration. Traditional activities face margin compression, and new pools open around specialized capability.
Should companies build internal AI capability or outsource it?
Both, with the judgment kept inside. Integration competence, meaning knowing how your own work should be redesigned, cannot be bought from outside. The commodity layers can and should be.
What should be on an AI roadmap?
Quick wins where today's capability clearly applies, foundational data infrastructure underneath them, an explicit plan for workflow redesign and change management, and a reskilling track for the people whose jobs change. A roadmap that is only a list of use cases is missing most of the work.
