LogoThread Easy
  • 탐색
  • 스레드 작성
LogoThread Easy

트위터 스레드의 올인원 파트너

© 2025 Thread Easy All Rights Reserved.

탐색

Newest first — browse tweet threads

Keep on to blur preview images; turn off to show them clearly

So go out, and develop some fomo. And remember, do not raise around Christmas because everyone is OOO. 

If you want more tips like this, give me a follow and head over to my substack (link in bio).

So go out, and develop some fomo. And remember, do not raise around Christmas because everyone is OOO. If you want more tips like this, give me a follow and head over to my substack (link in bio).

stealth consumer bio | prev founding team @a16z @speedrun | writer of the bird’s nest, https://t.co/3fgQ6avUsx

avatar for robin
robin
Wed Dec 17 19:29:17
The Hierarchy of Fundraising FOMO

Most founders I advise don't understand that they need to create a sense of urgency and desire. Emotional passion for the hunt. Most investors are sheep but even highly rational ones can get caught up in deal fever. 

Here's how to do that 👇

The Hierarchy of Fundraising FOMO Most founders I advise don't understand that they need to create a sense of urgency and desire. Emotional passion for the hunt. Most investors are sheep but even highly rational ones can get caught up in deal fever. Here's how to do that 👇

Your job is to create that deal fever because it'll mean better terms for your company. Of course the best thing you can do is have a great company with stellar metrics that's growing. Or be a master storyteller with an in-meta team going after a compelling market.

avatar for robin
robin
Wed Dec 17 19:29:13
i’m kinda unsure if everyone’s default subagents control flow is right

like why does the subagent always immediately self-destruct after sending its response??

sometimes i’d like the orchestrator to chat back to the subagent to refine stuff if needed without respawning a full Task and keeping that context

rn we give the orchestrator one shot to specify the Task, the subagent one shot to complete it, no feedback

i think the self-destructing and separate subagent makes sense but some offloading and feedback might too as an alternative pattern

my agent control flow friends, thoughts??

multi-agent orchestration 

so i think it’s useful for the orchestrator

i’m kinda unsure if everyone’s default subagents control flow is right like why does the subagent always immediately self-destruct after sending its response?? sometimes i’d like the orchestrator to chat back to the subagent to refine stuff if needed without respawning a full Task and keeping that context rn we give the orchestrator one shot to specify the Task, the subagent one shot to complete it, no feedback i think the self-destructing and separate subagent makes sense but some offloading and feedback might too as an alternative pattern my agent control flow friends, thoughts?? multi-agent orchestration so i think it’s useful for the orchestrator

agents, harnesses, and evals @LangChainAI, prev @awscloud, phd cs @ temple

avatar for Viv
Viv
Wed Dec 17 19:28:39
Saw this comment in a youtube video and wow! 

Such a thought provoking statement. Easier said than done though.

Saw this comment in a youtube video and wow! Such a thought provoking statement. Easier said than done though.

👨‍💻 Building apps. Tech: rails, flutter.

avatar for Rinas
Rinas
Wed Dec 17 19:26:43
I’m running these 3 businesses: 

- LinkTree for entrepreneurs 
- Google Analytics for businesses 
- A marketplace for startups

I’ve seen thousands of businesses.

One thing I realized is there are millions ways to make money online.

You don’t need a trendy idea
You don’t need AI in your biz name
You don’t need a fancy landing page

Solve a problem at least 1,000 ppl have. Don’t quit until you’ve tried everything to reach them all. Repeat for 10 problems. That’s it.

I’m running these 3 businesses: - LinkTree for entrepreneurs - Google Analytics for businesses - A marketplace for startups I’ve seen thousands of businesses. One thing I realized is there are millions ways to make money online. You don’t need a trendy idea You don’t need AI in your biz name You don’t need a fancy landing page Solve a problem at least 1,000 ppl have. Don’t quit until you’ve tried everything to reach them all. Repeat for 10 problems. That’s it.

💻 https://t.co/Y30jsaHwz9 $30K/m ⚡️ https://t.co/vatLDmi9UG $21K/m 📈 https://t.co/3EDxln5mdi $17K/m ⭐️ https://t.co/MZc8tG9xWi $17K/m 🍜 https://t.co/r07EpGSYJ2 $1K/m 🧬 https://t.co/SfrVXVtmdA $0/m 🧾 https://t.co/7olaOzV8Xd $0/m +20 https://t.co/4zCWHGJp1S

avatar for Marc Lou
Marc Lou
Wed Dec 17 19:25:27
the stuff that makes agents actually work is often unsexy —> writing prompts/skills by hand, reading traces, comparing outputs.  but if it’s unsexy at least there’s tooling to make it easy to do

in that last 10-20% of getting agents right, we wanna make it super easy for you to pick the best model, harness, prompts etc with tools like this

the stuff that makes agents actually work is often unsexy —> writing prompts/skills by hand, reading traces, comparing outputs. but if it’s unsexy at least there’s tooling to make it easy to do in that last 10-20% of getting agents right, we wanna make it super easy for you to pick the best model, harness, prompts etc with tools like this

agents, harnesses, and evals @LangChainAI, prev @awscloud, phd cs @ temple

avatar for Viv
Viv
Wed Dec 17 19:22:56
  • Previous
  • 1
  • More pages
  • 496
  • 497
  • 498
  • More pages
  • 5634
  • Next