ChatGPT's app has generated $2B

👋 Good morning/evening (wherever you are). It’s Friday.

ChatGPT’s mobile app is on another level.

Since launching in May 2023, ChatGPT’s mobile app has blown away the competition:

• $2B in global consumer spending, nearly 30x the combined lifetime mobile revenue of Claude, Copilot and Grok

• $1.35B earned in 2025 so far, up 673% from the same period last year

• Averaging $193M per month in 2025, 53x more than its next closest competitor

• 690M installs worldwide, with the U.S. accounting for 38% of all revenue

• Average spend per download: $2.91 globally and $10 in the U.S.

Scroll up, read that last bullet again.

ChatGPT’s $10 per U.S. download is exceptionally rare for a freemium app.

How rare?

It ranks among the highest monetizing mobile products ever recorded in any category.

OK let’s keep going ↓

🔥 Creative pulse

In case you missed it, here’s 3 videos you’ll watch twice (or thrice):

1. Steve McQueen…unreal quality. (Miki Durán)

2. Hong Kong backstreets…simply beautiful. (Mateo Willis)

3. Athletes in training mode…looks too real. (Andrew Rivard)

PS: If you’re passionate (and obsessed) about AI creative…guess what?! I just launched an AI creative club. It’s only $5/month. Skip the Big Mac and use it for this instead. I’m serious, you’ll thank me later. Click here to join.

💡 Here’s what you should know

• Meta’s consumer AI app reports ongoing issues

  Months after launch, Meta’s standalone AI app continues to experience technical problems, with users noting performance and reliability concerns.

• Meta to restructure AI unit for fourth time in six months

  The Information reports Meta will split its Superintelligence Labs into four groups, including a new TBD Lab, product and infrastructure teams, and the FAIR research lab.

• Otter.ai faces lawsuit over secret recordings

  A federal suit claims Otter.ai recorded private virtual meetings without consent to train its AI transcription service.

• NVIDIA releases huge dataset to improve speech AI

  The Granary dataset includes 1 million hours of audio and supports new NVIDIA models for accurate, fast transcription and translation across 25 languages.

• AI moves into the apartment market

  From virtual leasing agents to automated lease reviews and vendor payments, AI is starting to streamline operations for large multifamily operators, though adoption remains early and costs high.

• Turn websites into music with ElevenMusic

  Drop in a URL and let AI spin it into a jingle, a rap or even a movie trailer theme.

• Coinbase devs highlight new AI use case for Ethereum

  For the first time, Coinbase developers pointed to autonomous AI agents as a potential major user of Ethereum for stablecoin transactions in commerce, content, and logistics.

💰 The numbers

• TeraWulf signed a $3.7B, 10-year AI hosting deal with Fluidstack, with Google providing a $1.8B lease guarantee and taking an 8% equity stake.

• Galaxy Digital secured a $1.4B project financing agreement to speed up construction of its Helios AI datacenter infrastructure with CoreWeave.

• Eli Lilly signed a $1.3B AI drug discovery deal with Superluminal for obesity and cardiometabolic treatments.

• Xpand raised $6M to launch AI-powered autonomous “store-in-a-box” retail units, starting with its first smart store in Vienna.

🧠 Thought starters

This new Fortune piece on China’s power grid and AI is interesting (it’s paywalled), but here’s what matters:

• U.S. data centers already consume 4.4% of national electricity, projected to reach 6.7 to 12% by 2028 (Goldman Sachs)

• Reserve margins in most U.S. grids are 13 to 15%, leaving little headroom for rapid AI-driven load growth

• Grid delays are so severe that some hyperscalers are building their own power plants to bypass local constraints

• McKinsey projects 6.7 trillion dollars will be needed in global data center buildout by 2030 to meet AI demand

• China maintains far higher spare capacity with national reserve margins in the mid-20% range after decades of deliberate overbuilding and integrated national planning

Why it matters:

China has energy to spare, while the U.S. is already near its grid limits. That allows China to add AI data centers faster and cheaper, without political fights or years-long permitting.

This is not just about keeping the lights on. Energy capacity determines where the largest models are trained, who owns the resulting IP, and which markets lead the AI ecosystem. If U.S. operators cannot secure power at scale, workloads will shift to friendlier grids, taking technical leadership and economic leverage with them.

The result is greater dependence on hyperscalers and foreign infrastructure, reducing U.S. control over costs, security, and innovation. In the end, AI’s edge may be measured in megawatts and deployment speed as much as in GPUs and model size.

Context note:

Fortune has been owned since 2018 by Thai billionaire Chatchaval Jiaravanon, whose family has significant business ties in China. That does not discredit the reporting, but the framing leans more favorably toward China’s position.

😂 Meme of the day

Thanks for reading,

Eddie

Join the AI Creative Club (you’ll find me there). It’s not only for artists and filmmakers…it’s for anyone who recognizes this once-in-a-lifetime opportunity to learn as much as you can about AI to multiply your impact...to become superhuman. Remember, creativity is just intelligence having fun. This club is for ultra curious people, and if you’ve made it this far…you’re pretty damn curious and probably the smartest one in your circle of friends. It would be my honor to have you in there.

Are we friends on LinkedIn and X? Let’s connect. DM me or hit the follow button.

P.S. If this was valuable, forward it to a friend. If you’re that smart friend, subscribe here.

P.P.S. Interested in reaching other ambitious readers like you? To become a sponsor, reply to this email.