New http://tbpn.com live on ▲. TBPN is one of the coolest things in tech!
Daily Edition
2026-03-13
X / Twitter
13
Thanks for this opportunity @VaibhavSisinty, fun to chat about so many products in 30 minutes!
Vaibhav Sisinty: Just sat down with @joshwoodward, VP of Google Labs.
The man behind Gemini, NotebookLM & Google AI Studio.
Time's top 100 most influential AI figures.
Here's what actually stuck:
- 25-30% of Google's code is written by AI.
- Search is evolving into doing.
- Personal
Vaibhav Sisinty: Just sat down with @joshwoodward, VP of Google Labs.
The man behind Gemini, NotebookLM & Google AI Studio.
Time's top 100 most influential AI figures.
Here's what actually stuck:
- 25-30% of Google's code is written by AI.
- Search is evolving into doing.
- Personal
The frontier AI use-cases probably use 100X or so more tokens than the frontier use-cases did a year ago.
We’re already seeing the explosion of coding agents that do vastly more complex and longer running projects, which is already using up an insane amount of inference capacity. And this is a small percentage of total knowledge work.
This same architecture is about to come for the rest of knowledge work, where agents will effectively have their own computer to work with and ability to write and run code for many task, and comb through gobs of data to do their work. The token usage for these agents will be insane.
It’s about to get very interesting in inference land.
Suhail: The run on inference capacity is coming. You have been warned.
We’re already seeing the explosion of coding agents that do vastly more complex and longer running projects, which is already using up an insane amount of inference capacity. And this is a small percentage of total knowledge work.
This same architecture is about to come for the rest of knowledge work, where agents will effectively have their own computer to work with and ability to write and run code for many task, and comb through gobs of data to do their work. The token usage for these agents will be insane.
It’s about to get very interesting in inference land.
Suhail: The run on inference capacity is coming. You have been warned.
Wow.
@GarryTan (@ycombinator's CEO) just dropped the ultimate cheat code for software engineers. 🔥
He just open-sourced gstack, his personal toolkit that transforms Claude Code from a basic chatbot into an entire virtual engineering department.
Instead of asking Claude to "build a feature" and hoping for the best, gstack lets you summon specific "brains" on demand:
→ The Visionary:
/plan-ceo-review acts like Brian Chesky. It stops you from building boring features and pushes for magic.
→ The Architect:
/plan-eng-review draws sequence diagrams and state machines before coding.
→ The Paranoid Reviewer:
/review looks for N+1 queries and stale reads.
→ The QA Lead:
/qa literally logs into your staging environment, clicks around, takes screenshots, and gives your app a health score in 60 seconds.
The QA tool alone is built on Bun and Playwright, running 20x faster than Claude’s native Chrome MCP with zero context bloat.
He used this exact setup to ship 100 PRs a week for the last 50 days.
Get the repo in 🧵 ↓
@GarryTan (@ycombinator's CEO) just dropped the ultimate cheat code for software engineers. 🔥
He just open-sourced gstack, his personal toolkit that transforms Claude Code from a basic chatbot into an entire virtual engineering department.
Instead of asking Claude to "build a feature" and hoping for the best, gstack lets you summon specific "brains" on demand:
→ The Visionary:
/plan-ceo-review acts like Brian Chesky. It stops you from building boring features and pushes for magic.
→ The Architect:
/plan-eng-review draws sequence diagrams and state machines before coding.
→ The Paranoid Reviewer:
/review looks for N+1 queries and stale reads.
→ The QA Lead:
/qa literally logs into your staging environment, clicks around, takes screenshots, and gives your app a health score in 60 seconds.
The QA tool alone is built on Bun and Playwright, running 20x faster than Claude’s native Chrome MCP with zero context bloat.
He used this exact setup to ship 100 PRs a week for the last 50 days.
Get the repo in 🧵 ↓
The only way out of cringe valley is to post
Jonathan Brebner: Cringe is the path to glory.
Jonathan Brebner: Cringe is the path to glory.
Obsession is the mother of invention
A founder's best strengths are very often their biggest weaknesses.
If you work with them—as an investor, employee, or board member—your job is not to fix them. It's to understand exactly who they are and decide if you can live with that.
Don't look for folks who are:
- Ambitious BUT grounded.
- Decisive BUT collaborative
- Scrappy BUT refined.
That person doesn't exist.
If you work with them—as an investor, employee, or board member—your job is not to fix them. It's to understand exactly who they are and decide if you can live with that.
Don't look for folks who are:
- Ambitious BUT grounded.
- Decisive BUT collaborative
- Scrappy BUT refined.
That person doesn't exist.
I know there is some overlap between open source and anti-AI activists, but I have a hard time reconciling it. My million+ open source LOC were always intended as a gift to the world. Yes, I would make arguments about how it would strengthen our communities, and the GPL would prevent outright exploitation by our competitors, but those were to allay fears of my partners to allow me to make the gift.
AI training on the code magnifies the value of the gift. I am enthusiastic about it!
Some people do look at open source as a tool for social change, career advancement, or reputation building, but those are all downstream of the gift.
Rich Whitehouse: Genuinely devastating take to see from someone who popularized the GPL across so many communities. Fails to appreciate the social and cultural importance of the license.
AI training on the code magnifies the value of the gift. I am enthusiastic about it!
Some people do look at open source as a tool for social change, career advancement, or reputation building, but those are all downstream of the gift.
Rich Whitehouse: Genuinely devastating take to see from someone who popularized the GPL across so many communities. Fails to appreciate the social and cultural importance of the license.
I know there is some overlap between open source and anti-AI activists, but I have a hard time reconciling it. My million+ open source LOC were always intended as a gift to the world. Yes, I would make arguments about how it would strengthen our communities, and the GPL would prevent outright exploitation by our competitors, but those were to allay fears of my partners to allow me to make the gift.
AI training on the code magnifies the value of the gift. I am enthusiastic about it!
Some people do look at open source as a tool for social change, career advancement, or reputation building, but those are all downstream of the gift.
Rich Whitehouse: Genuinely devastating take to see from someone who popularized the GPL across so many communities. Fails to appreciate the social and cultural importance of the license.
AI training on the code magnifies the value of the gift. I am enthusiastic about it!
Some people do look at open source as a tool for social change, career advancement, or reputation building, but those are all downstream of the gift.
Rich Whitehouse: Genuinely devastating take to see from someone who popularized the GPL across so many communities. Fails to appreciate the social and cultural importance of the license.
Software is no longer "eating the world." AI is eating software.
The old moats are collapsing overnight. We are entering the Economy of Action, where the primary actors aren’t humans clicking buttons, but autonomous AI agents transacting in milliseconds.
But there is a massive problem: Software cannot secure an AI-driven world.
Software security is probabilistic. In a war of code vs. code, the attacker only needs to be right once. AI has turned "good enough" into "extinct."
At @Ledger, we’ve spent 12 years preparing for this collision of Blockchain and AI.
We are moving the "root of trust" out of the reach of code and into the Physics of Trust. By anchoring identity and governance in immutable silicon—Secure Elements logically isolated from the OS—we create a physical barrier that AI simply cannot cross.
Ledger is the weapon for the Agentic Economy.
Pascal Gauthier @Ledger: http://x.com/i/article/2032483015113281540
The old moats are collapsing overnight. We are entering the Economy of Action, where the primary actors aren’t humans clicking buttons, but autonomous AI agents transacting in milliseconds.
But there is a massive problem: Software cannot secure an AI-driven world.
Software security is probabilistic. In a war of code vs. code, the attacker only needs to be right once. AI has turned "good enough" into "extinct."
At @Ledger, we’ve spent 12 years preparing for this collision of Blockchain and AI.
We are moving the "root of trust" out of the reach of code and into the Physics of Trust. By anchoring identity and governance in immutable silicon—Secure Elements logically isolated from the OS—we create a physical barrier that AI simply cannot cross.
Ledger is the weapon for the Agentic Economy.
Pascal Gauthier @Ledger: http://x.com/i/article/2032483015113281540
"For the last three years, the best growth strategy in AI has been to just lie."
@synthesiaIO co-founder Victor Riparbelli (@vriparbelli) calls it "the biggest fog of war in AI right now."
"Just make a video of a product demo that's what you want to work eventually, go viral, pay a lot of influencers to push that message out for you."
The tactic can juice numbers even as it disappoints customers, but Victor hopes that startups taking a more honest approach win in the long run, as customers don't renew contracts on hype.
"We'll have to see in a couple of years how many of them, even the things that have grown super fast, how much of that is people dreaming, and hoping they can do something, versus them actually implementing it in a real workflow."
Catch the interview on The Upstarts Podcast
YouTube: https://www.youtube.com/watch?v=MnwO_f93AEI&t=590s
Apple: https://podcasts.apple.com/us/podcast/synthesias-victor-riparbelli-using-ai-to-replace-boring/id1875709419?i=1000754835347
Spotify: https://open.spotify.com/episode/1LsNFM5ORaq6bt4W0u5vex
Alex Konrad: Synthesia founder Victor Riparbelli (@vriparbelli) has always been drawn to counterculture ideas. "Life's more interesting at the fringes," he says.
The problem with fringe ideas? "Most of them are wrong. But the few that are right are very, very valuable."
At @synthesiaIO,
@synthesiaIO co-founder Victor Riparbelli (@vriparbelli) calls it "the biggest fog of war in AI right now."
"Just make a video of a product demo that's what you want to work eventually, go viral, pay a lot of influencers to push that message out for you."
The tactic can juice numbers even as it disappoints customers, but Victor hopes that startups taking a more honest approach win in the long run, as customers don't renew contracts on hype.
"We'll have to see in a couple of years how many of them, even the things that have grown super fast, how much of that is people dreaming, and hoping they can do something, versus them actually implementing it in a real workflow."
Catch the interview on The Upstarts Podcast
YouTube: https://www.youtube.com/watch?v=MnwO_f93AEI&t=590s
Apple: https://podcasts.apple.com/us/podcast/synthesias-victor-riparbelli-using-ai-to-replace-boring/id1875709419?i=1000754835347
Spotify: https://open.spotify.com/episode/1LsNFM5ORaq6bt4W0u5vex
Alex Konrad: Synthesia founder Victor Riparbelli (@vriparbelli) has always been drawn to counterculture ideas. "Life's more interesting at the fringes," he says.
The problem with fringe ideas? "Most of them are wrong. But the few that are right are very, very valuable."
At @synthesiaIO,
my three favorite claude code shortcuts:
1. `!` prefix runs bash inline. the command + output land in context
2. `ctrl+s` stashes your draft. type something else, submit, and it pops back
3. `ctrl+g` opens the prompt (or plan) in $EDITOR for bigger edits
1. `!` prefix runs bash inline. the command + output land in context
2. `ctrl+s` stashes your draft. type something else, submit, and it pops back
3. `ctrl+g` opens the prompt (or plan) in $EDITOR for bigger edits
Opus 4.6 1M context is now the default model for Max, Team and Enterprise users. Enjoy 🎉
Claude: 1 million context window: Now generally available for Claude Opus 4.6 and Claude Sonnet 4.6.
Claude: 1 million context window: Now generally available for Claude Opus 4.6 and Claude Sonnet 4.6.
YouTube
3
The MAD Podcast with Matt Turck
Mathematical Renaissance #ai #podcast
Watch Video
The MAD Podcast with Matt Turck
The Agent Harness Eats the Foundation Model #ai #podcast
Watch Video
No Priors: AI, Machine Learning, Tech, & Startups