{{first_name | Hi}}, your strategic AI update is here.
Google found AI in 68% of jobs but automating only 21% of tasks, AMD tied its chip business to Anthropic's model roadmap, and Anduril doubled its valuation in two months. Buying the tools is the easy part. 86% of executives are raising AI budgets this year. Only 21% are redesigning the processes that make it show up in the numbers.
"An organization's ability to learn, and translate that learning into action rapidly, is the ultimate competitive advantage."
— Jack Welch, former Chairman & CEO, General Electric
In today's lineup
Robots
The operating model is what your AI budget can't outspend
Real ROI from Gen AI
The execution intelligence layer your ERP has never had
People
Chad Edwards takes AI into the semiconductor supply chain
Love
Find the constraint before you automate
Power and Prediction by Ajay Agrawal, Joshua Gans and Avi Goldfarb
Reading time: 3 min.
ROBOTS 🤖
How are robotics and AI changing industries? We break down the latest news, tools, and innovations for you.
Top Insights
Google's largest AI study found AI in 68% of jobs but automating only 21% of tasks, meaning most budgets are funding adoption, not output.
AMD's $5B investment in Anthropic ties chips to models, so your AI vendor's roadmap now depends on who controls the hardware beneath it.
Anduril jumped from $61B to $100B in two months, as investors price AI-native operators higher than AI tool sellers.
21 nations backed open-source AI at APEC, making closed-weight costs harder to justify when open models now have government support.
The AI Kill Switch Act would let the US government shut down the models your enterprise runs, making AI governance a legal requirement.
The operating model is what your AI budget can't outspend
A new Accenture survey of 3,650 executives found that 86% plan to increase AI investment this year. Only 21% are redesigning end-to-end processes with AI at the core.
The CFO question that always follows: where is it showing up in the numbers? Seventy percent of tech budgets still fund legacy systems. Fewer than one in ten have redesigned roles around AI. The longer the operating model stays fixed, the harder that answer gets.
The 21% closing the distance share three disciplines:
Codify the process before automating it: decision rules cannot live in email threads.
Tie each deployment to a Finance-validated baseline before launch.
Run 12-week cycles and reinvest savings before momentum dissipates.
If the distance between your AI investment and your P&L outcomes is growing, let's talk.
What is the biggest cause of "AI spend with no P&L return" in your organization?
Real ROI from Gen AI
How human-centric AI delivers business results
Clinomic deployed Mona, an ICU AI assistant running on NVIDIA Jetson AGX, across 40+ hospitals in 8 EU countries, cutting documentation errors by 68% and clinical workload by 33%.
Procter & Gamble built a Molecular Discovery Suite that compresses R&D timelines from more than five years to under six months, using digital twin modeling before any physical prototype is built.
Verizon deployed AI across its network operations, resolving 85% of all network faults before customers are affected, eliminating manual intervention for the majority of issues, per CFO Tony Skiadas.
Tool Spotlight
Your weekly briefing on tools that create competitive leverage
The execution intelligence layer your ERP has never had
Dave Clark raised $50M this month to deploy AI agents making autonomous replenishment, routing, and allocation decisions on top of your existing ERP, WMS, and TMS without replacing any of them.
Choose if you:
Run inventory across disconnected systems with no unified execution layer.
Need AI agents making real-time decisions without touching your current ERP architecture.
project44 split into two businesses this month: a Decision Intelligence Platform for enterprise shippers, and LSP44, standalone AI agent infrastructure running at nine of the ten largest global logistics providers.
Choose if you:
Need real-time AI visibility across 100+ carriers to catch exceptions before they reach your customer.
Operate a 3PL, freight forwarder, or broker that needs embedded AI agents on your own stack.
Also worth watching: Altana for supplier risk mapping, Blue Yonder for multi-party coordination, and FourKites for predictive ETA.
PEOPLE 👥
Meet the innovators turning bold ideas into real-world impact.
Transformation Champion
Chad Edwards takes AI into the semiconductor supply chain
Dr. Chad Edwards built CuspAI on one decision: deploy AI agents to run materials discovery at machine speed, reducing research cycles from years to months, without replacing the chemists running them. The result: $450M raised earlier this month at a $2.6B valuation, with chip and clean energy manufacturers eliminating rare metal dependencies at scale.
LOVE ❤️
Practical wisdom, growth tactics, and a must-read book that will challenge the way you think.
Find the constraint before you automate
Goldratt proved it in manufacturing: fix the bottleneck or nothing else moves.
If your process is broken, an AI agent makes it faster and more wrong. The 21% who redesign first find the constraint. The 79% who don't are spending money to accelerate the problem.
Where is the actual bottleneck in your AI program right now?
Transformative Reads
One book, handpicked from my conversations with friends, industry leaders, and tech innovators:
In Power and Prediction, Ajay Agrawal, Joshua Gans and Avi Goldfarb show why AI doesn't just automate work. It changes who owns the decision. Org charts lag AI adoption by years. The leaders who read this early will know what to settle before authority reorganizes around the model.
In Culture
THANK YOU!
Lighthouse takes us many hours to create, and we don’t run ads or charge anything. We are dedicated to making it the best AI transformation newsletter possible.
We just ask for one thing: tell one friend about Lighthouse today!
Want daily AI & Innovation insights? Connect with me on LinkedIn or X.
Got suggestions? Just reply to this email!
Much Love,
Matt and the Future Works team
