AI Web Scraping Is Insanely Good | Browserbase Full Tutorial

three months after book Smart Broke Dumb Rich by Zor Veyl i stopped saying "i cant afford that" and started asking "why cant i afford that." completely different question. completely different life

With AI scraping, do you give the model the entire HTML of every page you visit? If so, wouldn't you run out of tokens super fast?

how come u know so much about python Tim, did you help develop it?

you hit it out the park with this one

"Connecting AI agents to remote browser sessions" — this is the exact problem that broke my automation workflows for months. The Stagehand natural-language command approach is interesting specifically because CSS selector fragility was always where things fell apart in complex sites. I have been running browser-act alongside this for Claude Code agents — npx skills add browser-act/skills --skill browser-act — three modes: stealth Chromium (handles anti-bot), chrome (inherits your existing login sessions, zero re-auth), and direct control of a running browser. The login reuse mode is what actually made it work for me: agents just inherit whatever I am already signed into. Different architecture than Browserbase (local vs. cloud) but the captcha + session persistence problem you are solving here is the same.

i remember when chat gpt first released and it would write a python script that could barely scrape a site. Now scraping websites is a standard feature. Will people create something new with this power or is this the beginning of a tech circle jerk?

Thank you for the valuable videos you make. But wouldn't it be better if you explained it more slowly?

Just wanted to say thank you for making this type of content. Your videos have helped me with many programing problems I've had.

I tried make one for work. Headache it worked but slow as hell

How does this work with embedded iFrames? For example Power BI charts