Peter Yang

Practical AI tutorials and interviews for busy people | Get my best AI skills and guides at

[Peter Yang @petergyang]

"If you want to build great agents, you have to let the model cook."

Here’s my new episode with @imjaredz (Builder in Residence at @cognition), where he shared what he learned from watching the best engineers build with AI agents, including:

→ Why fewer rules make agents work better
→ How to get agents to check their own work
→ Live demo: Build one agent to manage an agent team

Some quotes from Jared:

“Our goal is not for you to token max. Our goal is for you to ROI max.”

“The trick is to break up the work to keep the context window small for each sub-agent.”

“More Devin sessions now start from other Devins than directly from humans.”

📌 Watch now: https://t.co/jP1GoKtMnX

Thanks to our sponsors:

@RiversidedotFM: All-in-one AI studio for podcasts and video https://t.co/uWnS6aiMPE

@WisprFlow: 4x faster than typing with your voice https://t.co/oqHJ8bN3ll

[Peter Yang @petergyang]

"The mouse and keyboard are antiquated."

@imjaredz (Builder in Residence at Cognition) on his most radical prediction:

"I think we're going to return to a world where we just have a wooden desk without a keyboard or mouse.

I just want a button and a Wispr Flow [for voice dictation]. I think that's all we need."

📌 Watch the full episode: https://t.co/jP1GoKtMnX

[Peter Yang @petergyang]

Seeing this quite often today.

My wild guess is 90%+ of people are using GPT 5.6 Sol and <10% are using Terra or Luna.

[Peter Yang @petergyang]

Literally the first time I see GPT do this (it's for Sol)

[Peter Yang @petergyang]

There's a tendancy when community sentiment turns to communicate less and in a more corporate way.

I think it's important to do the reverse - communicate in a more human way, be transparent with the community about what's going on, figure out solutions together, etc.

[Peter Yang @petergyang]

Anthropic makes great models too but I don't understand why they communicate like this.

I think they need to engage the community more directly like what OpenAI is doing.

[Claude @claudeai]

We're extending Claude Fable 5 access on all paid plans, as well as keeping Claude Code’s weekly rate limits 50% higher, through July 19.

[Peter Yang @petergyang]

@andruyeung shit i think you just convinced me to go

[Peter Yang @petergyang]

@Teknium So sorry to hear this may he rest in doggy heaven

[Peter Yang @petergyang]

"If you want to build great agents, you have to let the model cook."

Here’s my new episode with @imjaredz (Builder in Residence at @cognition), where he shared what he learned from watching the best engineers build with AI agents, including:

→ Why fewer rules make agents work better
→ How to get agents to check their own work
→ Live demo: Build one agent to manage an agent team

Some quotes from Jared:

“Our goal is not for you to token max. Our goal is for you to ROI max.”

“The trick is to break up the work to keep the context window small for each sub-agent.”

“More Devin sessions now start from other Devins than directly from humans.”

📌 Watch now: https://t.co/jP1GoKtMnX

Thanks to our sponsors:

@RiversidedotFM: All-in-one AI studio for podcasts and video https://t.co/uWnS6aiMPE

@WisprFlow: 4x faster than typing with your voice https://t.co/oqHJ8bN3ll

[Peter Yang @petergyang]

@scaling01 wtf is a juice value

[Peter Yang @petergyang]

@imjaredz All of us talking to AI 1 year from now

[Peter Yang @petergyang]

"The mouse and keyboard are antiquated."

@imjaredz (Builder in Residence at Cognition) on his most radical prediction:

"I think we're going to return to a world where we just have a wooden desk without a keyboard or mouse.

I just want a button and a Wispr Flow [for voice dictation]. I think that's all we need."

📌 Watch the full episode: https://t.co/jP1GoKtMnX

[Peter Yang @petergyang]

"If you want to build great agents, you have to let the model cook."

Here’s my new episode with @imjaredz (Builder in Residence at @cognition), where he shared what he learned from watching the best engineers build with AI agents, including:

→ Why fewer rules make agents work better
→ How to get agents to check their own work
→ Live demo: Build one agent to manage an agent team

Some quotes from Jared:

“Our goal is not for you to token max. Our goal is for you to ROI max.”

“The trick is to break up the work to keep the context window small for each sub-agent.”

“More Devin sessions now start from other Devins than directly from humans.”

📌 Watch now: https://t.co/jP1GoKtMnX

Thanks to our sponsors:

@RiversidedotFM: All-in-one AI studio for podcasts and video https://t.co/uWnS6aiMPE

@WisprFlow: 4x faster than typing with your voice https://t.co/oqHJ8bN3ll

[Peter Yang @petergyang]

@kevinyien @nurijanian I'd love to collaborate on this if you're down and then share it with my audience. Can make it open source?

[Peter Yang @petergyang]

What is an example of a popular piece of writing (maybe an X article or long twee?) that upon a closer look was written with AI and full of slop?

[Peter Yang @petergyang]

Thank you Claude :) Just extend it permanently yeah?

[Claude @claudeai]

We're extending Claude Fable 5 access on all paid plans, as well as keeping Claude Code’s weekly rate limits 50% higher, through July 19.

[Peter Yang @petergyang]

I keep seeing this in Claude Design and while I get it shouldn't the product just become better at managing context automatically?

I'm not a designer but design is a very iterative and explorative process.

Seeing messages like this makes me not want to explore variations, etc.

[Peter Yang @petergyang]

@gabrielchua Haha this is awesome can you open source it? It can be like smash bros

[Peter Yang @petergyang]

@1809157Gaurav I'ts not looking good

[Peter Yang @petergyang]

They’re playing so badly even up one man

[Peter Yang @petergyang]

@imjaredz @cognition Was my pleasure Jared! Keep cooking

[Peter Yang @petergyang]

"If you want to build great agents, you have to let the model cook."

Here’s my new episode with @imjaredz (Builder in Residence at @cognition), where he shared what he learned from watching the best engineers build with AI agents, including:

→ Why fewer rules make agents work better
→ How to get agents to check their own work
→ Live demo: Build one agent to manage an agent team

Some quotes from Jared:

“Our goal is not for you to token max. Our goal is for you to ROI max.”

“The trick is to break up the work to keep the context window small for each sub-agent.”

“More Devin sessions now start from other Devins than directly from humans.”

📌 Watch now: https://t.co/jP1GoKtMnX

Thanks to our sponsors:

@RiversidedotFM: All-in-one AI studio for podcasts and video https://t.co/uWnS6aiMPE

@WisprFlow: 4x faster than typing with your voice https://t.co/oqHJ8bN3ll

[Peter Yang @petergyang]

Also available on:

Spotify:
https://t.co/PBcsQBzZ24

Apple:
https://t.co/xBhMbt2o77

Newsletter:
https://t.co/PGdOa77MFf

[Peter Yang @petergyang]

@JJEnglert How do you do this again? You’re running 5.6 in Claude code?