In partnership with
In Today’s Issue:
📈 OpenAI is tracking a massive compute expansion to 1.9 gigawatts of power
💸 OpenAI begins testing "Sponsored" boxes for Free and $8/month "Go" users
🏛️ Meta, Microsoft, and Perplexity sign multi-million dollar agreements
🚪 Thinking Machines ofounder Barret Zoph departs Mira Murati’s startup
✨ And more AI goodness…
Dear Readers,
OpenAI just crossed $20 billion in annualized revenue- a tenfold jump in two years - yet they're still burning $8 billion annually and won't turn a profit until 2029, which raises the question every AI watcher should be asking: are we witnessing the birth of a new tech titan or the most expensive experiment in Silicon Valley history?
Today's issue digs into this tension as OpenAI simultaneously announces ads are coming to ChatGPT, a move that could reshape how 800 million weekly users experience AI assistance. Meanwhile, the ripple effects are already devastating traditional software: Anthropic's Claude Cowork launch sent Intuit tumbling 16% and dragged the entire SaaS sector down 15% year-to-date - AI agents are no longer a future threat, they're a present-day market mover.
We've also got leadership drama at Thinking Machines Lab, Wikipedia finally getting paid by AI giants for its treasure trove of human-verified data, and a fascinating look at Demis Hassabis, the mind steering Google's AI ambitions. Let's dive in.
All the best,
🚀 Ads Arrive Inside ChatGPT Experience
OpenAI is beginning US tests of ads inside ChatGPT, placing them in clearly labeled boxes below answers while promising they won’t influence responses or rely on selling user data. Ads will target conversation topics (not personal profiles), appear only for free and $8/month Go users, and avoid sensitive areas like health, politics, or minors, while Plus, Pro, and Enterprise stay ad-free. With 800+ million weekly users and mounting monetization pressure amid rivals like Google Gemini, OpenAI’s leadership, led by Fidji Simo, is betting ads can fund growth without eroding trust, a balance even Sam Altman has warned is critical.
🚨 Startup Turmoil Sparks Leadership Exit
Leaders at Thinking Machines Lab confronted cofounder Barret Zoph over an alleged workplace relationship deemed “serious misconduct,” which ultimately preceded his termination. The fallout strained his partnership with Mira Murati, accelerated internal departures, and comes amid broader strategic misalignment at the fast-growing startup. Zoph has since landed at OpenAI, despite concerns reportedly not shared by its leadership, while Thinking Machines faces churn during talks of raising capital at a potential $50B valuation.
🧠 Wikipedia Monetizes AI Data Access
Wikipedia is officially getting paid by major AI companies like Meta, Microsoft, and Perplexity as demand for clean, human-verified training data explodes. Through new agreements led by the Wikimedia Foundation, AI firms gain high-speed, structured access to Wikipedia’s 65+ million articles—helping offset soaring server costs caused by automated scraping, even as human readership dropped ~8% last year. The move marks a strategic shift: Wikipedia stays open, but the biggest machine users now help pay the bills.
The Man Behind Google's AI Machine | Demis Hassabis Interview
From Research Preview to Revenue Rocket:
How OpenAI Built the Fastest-Growing AI Business in History!
The Takeaway
👉 OpenAI's annualized revenue exploded from $2B (2023) to $20B+ (2025) – a 10x increase in two years that directly tracked its compute expansion from 0.2 to 1.9 gigawatts
👉 Advertising is coming to ChatGPT's free and $8 Go tiers in the U.S., while Plus, Pro, and Enterprise remain ad-free – OpenAI promises ads won't influence responses or sell user data
👉 Despite record revenue, OpenAI is burning approximately $8B annually on compute costs with projected losses of $14B by 2026 – profitability isn't expected until 2029
👉 The 2026 priority is "practical adoption" in healthcare, science, and enterprise, with new monetization models including licensing, IP agreements, and outcome-based pricing on the horizon
OpenAI has officially crossed $20 billion in annualized revenue for 2025, marking a 233% increase from 2024 and cementing its place as the fastest-scaling tech company ever. For context: revenue grew from $2 billion in 2023 to $6 billion in 2024 to over $20 billion in 2025 – a tenfold jump in just two years.
CFO Sarah Friar spelled out the strategy in a new blog post: the business model scales directly with the value intelligence delivers. What started as subscriptions has evolved into a sophisticated ecosystem including consumer tiers, enterprise APIs, commerce integrations, and now – advertising. OpenAI announced it will begin testing ads within ChatGPT in the coming weeks, targeting free and Go-tier users while keeping paid plans ad-free.
The compute story is equally wild. OpenAI increased compute capacity from 0.2 gigawatts in 2023 to roughly 1.9 gigawatts in 2025 – and Friar insists more compute would have meant even faster growth.
But here's the tension that matters: the company expects to burn $8 billion in cash in 2025 on compute and other costs, with projected losses of $14 billion by 2026. Revenue is exploding, but so are expenses.
For 2026, the focus shifts to practical adoption – especially in health, science, and enterprise. Can OpenAI convert unprecedented usage into sustainable profits?
Why it matters: OpenAI is proving that AI can generate massive revenue at unprecedented speed, but the race between income and infrastructure costs will define whether this becomes a sustainable business or the most expensive experiment in tech history. Every AI company is watching this playbook unfold.
Sources:
🔗 https://openai.com/index/a-business-that-scales-with-the-value-of-intelligence/
AI in HR? It’s happening now.
Deel's free 2026 trends report cuts through all the hype and lays out what HR teams can really expect in 2026. You’ll learn about the shifts happening now, the skill gaps you can't ignore, and resilience strategies that aren't just buzzwords. Plus you’ll get a practical toolkit that helps you implement it all without another costly and time-consuming transformation project.
AI Agents Crushing Software Stocks
The software industry just got a wake-up call it won't forget anytime soon. When Anthropic dropped Claude Cowork on January 12, investors didn't just take notice – they panicked. Intuit tumbled 16% in its worst week since 2022, while Adobe and Salesforce both sank more than 11%. The SaaS sector is now down 15% year-to-date, marking its roughest January since 2022.
Here's what happened: Claude Cowork uses sophisticated multi-agent orchestration – a lead agent handles high-level planning while spawning sub-agents for specific tasks. Think reconciling fifty receipts against a budget spreadsheet without human hand-holding. A significant portion was actually built by Anthropic's developers using Claude Code in just ten days. AI building AI tools – that's the recursive future arriving faster than anyone predicted.
This moment is pivotal. The classic "build versus buy" calculation that governed enterprise software decisions for decades has fundamentally shifted. When domain experts can create sophisticated systems in hours rather than months, traditional SaaS subscriptions start feeling like expensive baggage.
While chipmakers expect profit growth of 59% in 2026, software earnings growth is slowing to 14%. The divergence tells the story: infrastructure wins, middleware worries.
Stop guessing. Start scaling.
See the top-performing Facebook ads in your niche and replicate them using AI. Gethookd shows you what’s actually working so you can increase ROI and scale ad spend with confidence.
