Le Comprehensive Crypto x AI Thesis by Cookies 🍪 Here's my thesis on how the maturing crypto x AI infrastructure stack leads to emerging innovative applications Let's dive in on how to navigate this new age as users and as builders 🧵

Maybe we could have a tool that’s able to measure the level of emotional attachment users have to ChatGPT based on their chat history As a safeguard to prevent more severe effects like psychosis / mania

chain paid tools into autonomous workflows

launch an ERC-8004 agent

no code apps + x402

make a payment in order to post on somebody's page

collating a list of x402 use cases I've come across over the past week 🧵

Fintech 3.0 - the quiet re-platforming of money 1/ Traditional fintechs leveraged banking as a service (BaaS) as an interface on top of bank infrastructure. Banks still held the deposits, startups just built better distribution/UX. But the new wave of stablecoins/blockchains is different. It’s not a wrapper. It’s a new platform entirely.

🧵 Can a prediction market exist for inherently subjective and unverifiable outcomes?

Hello world. We’re Solomon. Making billions in idle onchain stable capital productive without breaking composability. A dollar stays a dollar, spends like a dollar… and now it earns. More soon.

The best discussion this week wasn’t about any one company. Markets are looking more and more like casinos: perps on illiquid assets, prediction markets, memecoins, social trading. Is it troubling? Is it an adaptation? Both? We couldn't agree. More on our conversation below:

really in-depth article on @Infinit_Labs’s architecture, the various products and how it relates to Google’s AP2 excited to see how this pans out, am curious to explore the possibilities of Infinit’s DeFi agents interacting with partners of Google’s AP2 program

this in singapore? 👀 @tongnk wdyt

1/ I’ve worked with hundreds of VCs and keep running across the same problem: there’s no simple tool to research rounds and competition. So…we built it. The Venture Codex is a better way to do research before investing and before helping your companies fundraise.

Just how much are we spending on AI? Compared to other massive infrastructure projects, AI is the sixth largest in US history, so far. World War II dwarfs everything else at 37.8% of GDP. World War I consumed 12.3%. The New Deal peaked at 7.7%. Railroads during the Gilded Age reached 6.0%. AI infrastructure today sits at 1.6%, just above the telecom bubble’s 1.2% & well below the major historical mobilizations. Companies like Microsoft, Google, & Meta are investing $140B, $92B, & $71B respectivelyin data centers & GPUs. OpenAI plans to spend $295B in 2030 alone. If we assume OpenAI represents 30% of the market, total AI infrastructure spending would reach $983B annually by 2030, or 2.8% of GDP. To match the railroad era’s 6% of GDP, AI spending would need to reach $2.1T per year by 2030 (6% of projected $35.4T GDP), a 320% increase from today’s $500B. That would require Google, Meta, OpenAI, & Microsoft each investing $500-700B per year, a 5-7x increase from today’s levels. And that should give you a sense of how much we were spending on railroads 150 years ago! tomtunguz.com/llm-impact-gdp…

We’re overthinking AI chat history — just like Chrome did with web history. 20 years ago, Chrome indexed every page users visited and designed detailed UIs for them to explore it. But users didn’t care. They just searched again. Here's what AI tools can learn from this 👇 Chrome’s web history dataset was rich, personal, and full of possibility. Designers explored everything from subway-map visualizations, to timeline UIs, to GoogleReader-style feeds — imagining people would want to explore their own browsing history. When they realized it was just easier, faster, and more intuitive for users to search again, Chrome shifted course. Web history became infrastructure — powering suggestions, surfacing shortcuts, and tucked behind “Command + Y.” Today, AI teams are replaying this same debate and trying to “solve” chat history. But people don’t want to manage their chat history. They just want models to remember intelligently so that every conversation feels smoother, smarter, and more personal. History should live in the background, supercharging personalization, context, and recall across the core product. The best designed products weave history in when it makes sense, but focus on propelling users forward. Full post in comments.

imo the biggest unlock here is surfacing relevant markets to individuals 1. interested in a topic 2. search on google 3. see prediction markets barrier to place a trade can be lowered by allowing payments through google pay

🚨 KALSHI AND GOOGLE FINANCE PARTNER TO ADD PREDICTION MARKETS

are there any prediction market copy trading apps?