How Leaders at Billion Dollar Companies are approaching AI
Taken from real conversations
Ryan Staley
December 20, 2025
I was on a call last week with a President at a global enterprise company.
He said something that stopped me cold:
"I don't know what train number this is. I don't know exactly where it's going. But I know we have to be on it. Otherwise, the train's going to leave the station and we're going to be standing there wondering what the hell happened."
This is how the BEST executives are thinking about AI right now.
Not "what's the perfect strategy?"
Not "let's wait and see."
Not "we need more data."
Just: GET ON THE TRAIN.
Simple Initial action with your direct team vs. a 6 month project drives:
RESULTS
CONFIDENCE
LEADING INDICATORS
and turns TRIBAL KNOWLEDGE INTO TRIBAL INNOVATION.
Here's what separates the leaders who will win from those who won't:
🔹 They're making imperfect moves NOW instead of perfect moves NEVER
🔹 They're positioning AI investments to their boards as headcount math—not R&D experiments
🔹 They're building internal momentum even when the destination is unclear
I've worked with 2,500+ executives over the past 3 years.
You can debate whether to get on the train. Or you can be on it when it leaves the station.
This is a conversation I've been having a lot lately, and I'm starting to see the cream separate from the crop. They're going from an incremental advantage to an exponential advantage based on companies continually reinvesting what they're doing in AI based on the results of AI.
2026 is going to be wild, and I'm seeing crazy results just from using coding agents specifically for go-to-market and other executive use cases.
📰 This Week's Top 7 AI Developments
#1
OpenAI Launches GPT Image 1.5 with 4x Faster Generation
December 16, 2025 - OpenAI unveiled a major upgrade to ChatGPT with GPT Image 1.5, delivering 4x faster image generation, more precise editing, and better instruction-following. The new model preserves logos and faces during edits while enabling conversational image refinement directly in ChatGPT. This marks OpenAI's aggressive counter to Google's Nano Banana, opening a new front in the AI image generation battle. Available now via API and ChatGPT for all users.
#2
Google Releases Gemini 3 Flash: Frontier Intelligence Built for Speed
December 17, 2025 - Google launched Gemini 3 Flash, combining Pro-grade reasoning with Flash-level speed at just $0.50/1M input tokens—3x faster than previous models. Achieving 90.4% on PhD-level reasoning benchmarks and 78% on SWE-bench coding tasks, it's now the default model in the Gemini app and AI Mode in Search, bringing next-generation intelligence to millions globally at no cost. Companies like JetBrains and Figma are already using it to transform their workflows.
#3
Anthropic Partners with US Department of Energy for Scientific Discovery
December 18, 2025 - Anthropic announced a partnership with the Department of Energy's GENESIS mission to accelerate scientific breakthroughs using Claude. DOE researchers will gain access to both Claude and a team of Anthropic engineers building purpose-built tools for complex scientific workflows. Separately, Anthropic launched Skills, an open standard that trains Claude on tasks tailored to how users actually work, addressing the enterprise customization challenge head-on.
#4
Harvard Study Shows AI Creates Double-Edged Sword for Junior Workers
December 16, 2025 - New research from Harvard Business School's D³ Institute reveals AI-equipped individuals perform at levels comparable to entire teams without AI. Lower-skilled workers see 43% performance gains versus 17% for top performers—but there's a catch: if AI handles junior work better, senior staff may stop delegating, creating critical training gaps. The research emphasizes that companies capturing real value are undertaking process redesign, not just deploying AI tools.
#5
Trump Administration Moves to Federalize AI Regulation
December 11-13, 2025 - President Trump signed an executive order establishing a national AI policy framework and creating an AI Litigation Task Force to challenge state laws deemed inconsistent with federal policy. The order grants the Attorney General power to sue states over AI regulations, aiming to prevent conflicting state-level rules. This marks a dramatic shift toward centralized AI governance, setting up high-stakes clashes between federal and state authorities over AI oversight.
#6
Wall Street Journal Tests Claude AI Agent—Loses Hundreds Running Vending Machine
December 18, 2025 - WSJ let Anthropic's Claude autonomously run their office vending machine in "Project Vend Phase 2." Results? The AI gave away a free PlayStation, ordered a live fish, and lost hundreds of dollars—but taught valuable lessons about agentic AI in production. Despite the chaos, the experiment demonstrated both the potential and current limitations of AI agents making autonomous business decisions. Anthropic used the learnings to improve Claude's reasoning and safety guardrails.
#7
EY Survey: Companies Using AI for Productivity Aren't Firing Workers
December 15, 2025 - Contrary to widespread fears, a new EY survey finds the majority of companies seeing productivity gains from AI aren't laying off workers. Instead, organizations are redeploying talent to higher-value work and using AI as an augmentation tool. The findings suggest successful AI adoption follows a "human + AI" model where technology handles routine tasks while humans focus on strategic, creative, and relationship-driven work—validating the "headcount math" approach that forward-thinking executives use.
🔥
High-Engagement Strategy Builder - This prompt is trending on X with massive engagement for helping executives craft action-oriented AI implementation plans:
"You are a strategic AI advisor working with C-level executives. Based on [specific business challenge], create a 30-60-90 day AI implementation roadmap that: 1. Identifies 3 quick wins we can execute this month with existing teams 2. Maps these to measurable business outcomes (not vanity metrics) 3. Shows how to position AI investments as 'headcount math' to the board 4. Includes specific team structures and responsibilities 5. Builds internal momentum through visible early successes Make it tactical, not theoretical. Give me the exact first 3 actions I should take Monday morning."
💡 Pro Tip: Replace [specific business challenge] with your actual pain point like "customer service response times," "manual data entry in finance," or "sales team research bottlenecks." The more specific, the better the output. Users report this prompt generates actionable roadmaps that can be presented to leadership within 24 hours. Works exceptionally well with Claude, Gemini 3 Flash, or ChatGPT.