{{first_name | Hi}}, your strategic AI update is here.
Uber burned through its 2026 AI coding budget by April, AirTrunk bet $30 billion on AI infrastructure, and Shell's agents closed the maintenance loop on 30,000 assets. Buying the tools is the easy part. Companies that anchor every deployment to available data and a CFO-trackable metric are pulling away from the 44% funding wave two on returns that never arrived.
"Knowledge has to be improved, challenged, and increased constantly, or it vanishes."
— Peter Drucker, Management Theorist and Author
In today's lineup:
Robots
The AI savings gap leaving 44% of companies in Pilot Purgatory
Real ROI from Gen AI
When AI starts predicting your revenue before your reps do
People
Jonathan Godwin uses AI to design materials faster than any lab on Earth
Love
The knowledge gap blocking your AI savings
The Lean Startup by Eric Ries
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
Uber's AI budget runs dry by April, as per-developer token costs surge 18.6x in nine months across the industry.
AirTrunk pledges $30B for India AI data centers, committing to 5GW of capacity as global AI infrastructure investment accelerates.
Shell's C3 AI agents close the full maintenance loop, automating 30,000-asset repair cycles for measurable wave-two returns.
NSA deploys Anthropic's Mythos for offensive cyber operations, despite an active DoD supply chain ban on the company.
Great American AI Act would preempt state AI rules, giving enterprises three years of federal compliance clarity.
The AI savings gap leaving 44% of companies in Pilot Purgatory
Nearly 40% of companies landed below 10% in AI cost savings, well below the 11–20% they targeted.
Companies that can't prove last year's AI savings are committing to next year's wave. A Bain survey of 951 companies found 40% saw cost reductions below 10%, well short of the 11–20% projected, and 44% are financing the next round on those unconfirmed returns.
Data access is the bottleneck. Most organizations can't query their own systems. Bain calls this the primary separator between top performers and everyone else. Only 4% achieved savings above 30%.
The top 4% tie every deployment to a CFO-trackable metric before scaling.
Three moves to close the gap:
Audit data access across your AI deployments before the next budget cycle, not after it starts.
Assign one CFO-approved savings metric to every active AI initiative and measure it quarterly.
Refuse to sign off on wave two until wave one's savings are confirmed, not projected.
What is the biggest cause of the "AI savings gap" in your organization?
- We set AI savings targets against projections, not against confirmed baseline data
- Teams report AI activity as progress, without tying it to a measurable financial outcome
- The data our AI tools need to prove results isn't structured or accessible yet
- Leadership approved the next AI investment before the last one was fully evaluated
Real ROI from Gen AI
How human-centric AI delivers business results
Airbus partnered with Mistral AI to deploy AI across four aerospace workflows, including technical documentation automation and edge AI models onboard aircraft for situational awareness.
Kaiser Permanente deployed AI ambient scribes across its physician network, saving the equivalent of 1,794 working days of documentation time in a single year, freeing doctors to spend more time with patients.
Citi freed 100,000 hours of weekly developer capacity through AI-powered code reviews, and is now piloting agentic AI with 5,000 colleagues for complex multistep task automation.
Tool Spotlight
Your weekly briefing on tools that create competitive leverage
When AI starts predicting your revenue before your reps do
Gong surpassed $500M ARR this past month as enterprises doubled down on revenue AI. Teams using Gong report 77% more revenue per rep compared to teams not using AI.
Best for: Revenue and GTM leaders who need a single source of truth across every customer conversation, deal, and forecast signal.
Choose if you...
Need AI to flag at-risk deals from call patterns before reps notice the slip.
Want coaching that fires automatically after every call without manager overhead.
Run a complex enterprise sales cycle and need a forecast the CFO will trust.
Best for: Operations and finance leaders who want AI-generated revenue forecasts tied directly to CRM pipeline data, not gut feel.
Choose if you...
Prioritize forecast accuracy over conversation intelligence and already have a call-recording tool.
Need a revenue operations view that spans sales, renewals, and expansion in one model.
Want to connect AI forecasting directly into your financial planning workflow.
3 other tools to explore
PEOPLE 👥
Meet the innovators turning bold ideas into real-world impact.
Transformation Champion
Jonathan Godwin uses AI to design materials faster than any lab on Earth
The materials science industry takes a decade and $100 million to design a single new industrial compound. Godwin, a former Google DeepMind researcher, built Orbital Industries on the thesis that AI can collapse that timeline to months. His team's first product, a PFAS-free GPU cooling fluid, is on track to become the first AI-designed molecule to hit commercial markets by next year. Earlier this month, Orbital closed a $50M Series B with Nvidia's venture arm as a backer.
LOVE ❤️
Practical wisdom, growth tactics, and a must-read book that will challenge the way you think.
The knowledge gap blocking your AI savings
Most AI programs generate activity without leaving a trail. When the CFO asks what six months of investment actually produced, nobody has the answer. That's the behavioral root of the savings gap. The fix: one synchronized log, every active AI initiative, one measurable outcome per sprint.
Log every active AI initiative this week with a single CFO-trackable metric.
Assign one person per sprint to capture the result, win or miss.
Present two logged quick wins to leadership before each budget cycle as proof.
Build the knowledge infrastructure behind every AI breakthrough in my book, The LEAP Guide.
Transformative Reads
One book, handpicked from my conversations with friends, industry leaders, and tech innovators:
In The Lean Startup, Eric Ries argues that most organizations build too much before proving anything. His Build-Measure-Learn loop forces teams to validate what actually works before scaling investment. That's the discipline separating the 4% of companies that hit AI savings above 30% from everyone else funding the next wave on projections.
Perfect for: AI program leads, COOs, and CFOs deciding whether to scale an existing pilot or shut it down.
In Culture
The CFO who asked whether wave one actually paid off. Every AI kickoff, without fail.
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
