Ron Gross

Let me explain what just happened, because I don’t think people realize how INSANE this is. > Cortical Labs put 200,000 real human brain cells onto a silicon chip and trained them to play Doom in just one week. > Each CL1 system costs $35,000. > A rack of 30 units consumes only 850–1,000 watts combined. > The human brain operates on 20 watts. > Large AI training clusters burn through megawatts. >Backed by In-Q-Tel. 115 units began shipping in 2025. > Cortical Labs is selling “Wetware as a Service” through Cortical Cloud, letting developers deploy code remotely to living human neurons with no lab required, > priced like a software subscription but powered by real brain cells grown from adult skin and blood samples. > it isn’t about gaming, it’s about biological computing that could eventually outperform traditional silicon in energy efficiency and adaptability. This is getting really scary and we’re still at the very beginning.

750 million users. 10 billion tokens per minute. $185 billion in capex. The numbers Google just reported are massive. But zoom out and you’re watching a four-horse race where every horse is running on a different track. Gemini just crossed 750 million MAUs. ChatGPT has roughly 900 million weekly active users. Claude has around 30 million. Meta AI crossed 1 billion. Four companies, four wildly different user counts, and the scoreboard that actually matters looks nothing like the headline numbers. Start with how Google gets its users. Gemini went from 650 million to 750 million MAUs in one quarter. That 100 million user jump happened in the same window where Google replaced the default assistant on 580 million Android devices, baked Gemini into Search (2 billion AI Overview users), and launched Gemini 3, which Pichai called “the fastest adoption of any model in our history.” When you own the default position on 3+ billion devices and the world’s most visited website, adding 100 million users is the floor, not the ceiling. ChatGPT grew to 900 million weekly users by making people deliberately download an app or type in a URL. That’s a fundamentally different acquisition motion. ChatGPT processes 2.5 billion prompts per day and pulls 5.8 billion monthly web visits. Gemini gets 1.2 billion monthly visits on 750 million “active” users, which means the average Gemini MAU visits the web product 1.6 times per month. The engagement gap is still enormous. Claude tells the most interesting story of the four. Roughly 30 million monthly users generating $9 billion in annualized revenue by end of 2025, with 80% coming from enterprise and API across 300,000 business customers. Claude Code alone is approaching $1 billion in annualized revenue. Anthropic is projecting $20 to $26 billion ARR in 2026. Compare that to OpenAI expecting to close 2025 with over $13 billion in total revenue from 900 million weekly users and 10 million Plus subscribers. Anthropic generates $9 billion ARR from 30 million users. OpenAI generates $13 billion from 900 million. The revenue per user gap between those two companies tells you everything about where enterprise AI spending is actually flowing. Google launched AI Plus at $7.99/month, half the price of ChatGPT Plus, and still hasn’t disclosed Gemini’s paid conversion rate. Meta AI crossed 1 billion MAUs by embedding into WhatsApp, Instagram, and Facebook, generates zero direct AI revenue, and Zuckerberg told shareholders the plan is to eventually “insert paid recommendations” or offer subscriptions. A billion users with no monetization path beyond future ads. The capex tells you who’s betting on what. Alphabet is spending $175 to $185 billion in 2026, more than doubling the $91.4 billion from 2025. Meta is spending $115 to $135 billion. Both are building infrastructure to serve billions of free users they haven’t figured out how to monetize through AI yet. OpenAI raised $40 billion in its Series F and is building Stargate targeting $500 billion over four years. Anthropic raised $13 billion in its Series F at a $183 billion valuation that’s since been marked up to roughly $350 billion on the back of $15 billion from Microsoft and Nvidia. Each company is winning exactly one race. Google wins cloud throughput and distribution. OpenAI wins consumer engagement and willingness to pay. Anthropic wins enterprise revenue efficiency. Meta wins a user count that nobody’s figured out how to monetize. The question Wall Street hasn’t answered: which of those four races determines who captures the economics of AI over the next five years.

In the coming weeks, we plan to start testing ads in ChatGPT free and Go tiers. We’re sharing our principles early on how we’ll approach ads–guided by putting user trust and transparency first as we work to make AI accessible to everyone. What matters most: - Responses in ChatGPT will not be influenced by ads. - Ads are always separate and clearly labeled. - Your conversations are private from advertisers. - Plus, Pro, Business, and Enterprise tiers will not have ads.

Can you read 900 words per minute? Try it.

True, ancient Peter Thiel technique.

Delete your product backlog every six months. If it’s important, it will emerge again. If not, you’ll save your team the mental anguish of rehashing something that will never see the light of day.

Group chats in ChatGPT are now piloting in Japan, New Zealand, South Korea, and Taiwan. A new way to collaborate with friends, family, or coworkers and ChatGPT in the same conversation.

Does your imposter syndrome ever go away or do you just scale it with your ambition?