Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Programming is about to get pretty crazy 😅 I just "programmed" a Stripe customer checkout by providing a spec of of what it should do (left side). Then let our agent use tools and figure out the rest up to the deployment Took like 90 seconds or so? 😅

Vasek Mlejnsky

6,521 subscribers

11,858 views • 3 years ago •via X (Twitter)

Comedy Science & Technology Education

Anya Rossi• Live Now

Private livecam show

5 Comments

Jason Laster3 years ago

@auchenberg @stripe This dx feels right.

~/fabio$3 years ago

@stripe Nice but it won't work, just another hyped tool based on a hyped new tech.

Philipp Tsipman3 years ago

What tool is this? New DevBook product?

Vasek Mlejnsky3 years ago

Hey! New project. Leave your email here if you want to get notified when we launch We're open-sourcing the repo right now

Renaissance Man3 years ago

@stripe @OruOgbo

Related Videos

Tesla Robotaxi Unsupervised gets into a hotel parking lot that blocked off by an Amazon delivery vehicle, another vehicle, and a large Sysco delivery truck. It took Robotaxi a few seconds to figure out what it was going to do and then it went to the side of the hotel to drop me off, and then went behind the hotel to get back out and on to it's next pickup. Pretty impressive!

Tesla Robotaxi Unsupervised gets into a hotel parking lot that blocked off by an Amazon delivery vehicle, another vehicle, and a large Sysco delivery truck. It took Robotaxi a few seconds to figure out what it was going to do and then it went to the side of the hotel to drop me off, and then went behind the hotel to get back out and on to it's next pickup. Pretty impressive!

JC Christopher

33,931 views • 1 month ago

I can struggle to articulate interactions verbally before seeing them, so it's helpful to use a familiar tool like Origami to figure out just enough of what to say, and then prompt the rest

I can struggle to articulate interactions verbally before seeing them, so it's helpful to use a familiar tool like Origami to figure out just enough of what to say, and then prompt the rest

Julian Francis Kelly

24,068 views • 1 month ago

Bensky: And then we have the moment where all of my information is released. All of these other survivors — all of our information is literally out there for the world to see. And you’re like, “Wow, the DOJ really — our government doesn’t care.” And so then you’re wrestling with, like, “Do they just not care? Is it incompetence? Is this, like, you know, a cover-up? A major cover-up?” And I didn’t want to believe that it was a cover-up. When I first got here, I was pretty naive and green about all of it.

Bensky: And then we have the moment where all of my information is released. All of these other survivors — all of our information is literally out there for the world to see. And you’re like, “Wow, the DOJ really — our government doesn’t care.” And so then you’re wrestling with, like, “Do they just not care? Is it incompetence? Is this, like, you know, a cover-up? A major cover-up?” And I didn’t want to believe that it was a cover-up. When I first got here, I was pretty naive and green about all of it.

Acyn

52,336 views • 3 months ago

Should the dog have a service vest? Should the guy be less of a dick? Is there an actual disability or is she just a liberal? I can’t decide who’s in the wrong here, so just fight about it and let me know. 🤔😅😭

Should the dog have a service vest? Should the guy be less of a dick? Is there an actual disability or is she just a liberal? I can’t decide who’s in the wrong here, so just fight about it and let me know. 🤔😅😭

Zero Tolerance Policy

10,399 views • 1 month ago

One reason I wish I could train Sora locally is because about 25% of generations glitch up pretty hard and turn into what I call "Nightmare World." I'll use the same prompt over and over and get like 5 Nightmare World generations in a row. Then suddenly I'll get a perfect generation. I still can't figure out why this happens or how to fix it. A simple thumbs up/thumbs down button would help weed out Nightmare World (or explore it, if you want). I've tried negative prompts, simplifying prompts, refreshing, and all kinds of stuff but I can't figure out how to eliminate this glitch. One theory I have is this is the AI equivalent of "sketching" - After several abysmal generations I'll suddenly, without altering the prompt, get one that's better than ever before. Like the animation upgrades suddenly. So then I'll just keep running it and running it and get a really good streak of beautiful animation. And then out of nowhere it will switch back to Nightmare World again. It's a very interesting phenomenon that really should be looked into!

One reason I wish I could train Sora locally is because about 25% of generations glitch up pretty hard and turn into what I call "Nightmare World." I'll use the same prompt over and over and get like 5 Nightmare World generations in a row. Then suddenly I'll get a perfect generation. I still can't figure out why this happens or how to fix it. A simple thumbs up/thumbs down button would help weed out Nightmare World (or explore it, if you want). I've tried negative prompts, simplifying prompts, refreshing, and all kinds of stuff but I can't figure out how to eliminate this glitch. One theory I have is this is the AI equivalent of "sketching" - After several abysmal generations I'll suddenly, without altering the prompt, get one that's better than ever before. Like the animation upgrades suddenly. So then I'll just keep running it and running it and get a really good streak of beautiful animation. And then out of nowhere it will switch back to Nightmare World again. It's a very interesting phenomenon that really should be looked into!

Linda

47,631 views • 4 months ago

Elsa Jean says the absolute worst thing a guy can do when a girl is close to clim@xing is "automatically going har6er." claiming communication is the key that guys should be asking questions like “tell me what you like” and actually paying attention instead of going cr@zy. 👀 “I think it’s really important for a guy to be like tell me what you want, tell me if you like it like and actually listen to what she saying and I think it’s all about communication. Don’t get so crazy into it that can’t even listen to what she’s saying and then another fuck up thing guys use to do is when a girl is c$mming, they automatically like let me just go even har6er. Don’t just start going cr@zy”

Elsa Jean says the absolute worst thing a guy can do when a girl is close to clim@xing is "automatically going har6er." claiming communication is the key that guys should be asking questions like “tell me what you like” and actually paying attention instead of going cr@zy. 👀 “I think it’s really important for a guy to be like tell me what you want, tell me if you like it like and actually listen to what she saying and I think it’s all about communication. Don’t get so crazy into it that can’t even listen to what she’s saying and then another fuck up thing guys use to do is when a girl is c$mming, they automatically like let me just go even har6er. Don’t just start going cr@zy”

wfx

14,495 views • 15 days ago

Because they don't care, and/or get kickbacks. It doesn't take a brainiac to figure it out. The question is, what do we do about these cronies?

Because they don't care, and/or get kickbacks. It doesn't take a brainiac to figure it out. The question is, what do we do about these cronies?

Krista Monroe

33,349 views • 1 year ago

The pinnacle of visual art 😅 My husband gave me this, and it took me 5 minutes to figure out what it’s for

The pinnacle of visual art 😅 My husband gave me this, and it took me 5 minutes to figure out what it’s for

vadina atria

4,501,163 views • 12 days ago

"I bet 90%+ of the country can’t afford to use it." A American guy commented seriously.😅

"I bet 90%+ of the country can’t afford to use it." A American guy commented seriously.😅

ShanghaiPanda （ Account deactivation）

13,586 views • 11 months ago

Finally!! I get to use a power tool. I just need to figure it out first 😅 #japanesegirl #japan #kawaii #japanese #powertools #dirtyjobs #sanding #floor #refinishing

Finally!! I get to use a power tool. I just need to figure it out first 😅 #japanesegirl #japan #kawaii #japanese #powertools #dirtyjobs #sanding #floor #refinishing

Happy Nats

60,051 views • 1 year ago

Here is what openai's new voice mode looks like with RAG. This lets you connect the voice agent to external sources to like company files, a website or a database. The world of customer support just changed forever Check it out

Here is what openai's new voice mode looks like with RAG. This lets you connect the voice agent to external sources to like company files, a website or a database. The world of customer support just changed forever Check it out

Yasser

298,032 views • 1 year ago

Bash is all you need! Which is why I'm introducing my holiday project: just-bash just-bash is a pretty complete implementation of bash in TypeScript designed to be used as a bash tool by AI agents. Because it turns out agents love exploring data via shell scripts, even beyond coding. It comes with grep, sed, awk and the 99th percentile features that an agent like Claude Code or Cursor would use. In fact, Claude Code can use it for secure bash execution. In the package - A bash-tool for AI SDK - A binary for use by yourself or your coding agents - An overlay filesystem to feed files to your agent securely - A Vercel Sandbox compatible API, so you can quickly upgrade to a real VM if you need to run binaries - An example AI agent that explores the just-bash code base using just-bash - I imported the Oils shell bash compatibility suite and just-bash passes a very good chunk What is interesting about this codebase: It was essentially entirely written by Opus 4.5. Coding agents love bash and they are good at reproducing it. They are also great at text-book recursive descent parsers and AST tweet-walk interpreters. That said, it is, like, a lot of code and I didn't read it all 😅. This is very much a hack, but it also seems to be _really_ useful. I haven't really found anything agents want to use that it doesn't support and it's fast and secure (caveats apply). It doesn't have write access to your computer and the filesystem is given a root that the agent cannot escape from. Find it at Related: Our recent blog post how we migrated our data analysis agent to bash tools and achieved incredible quality improvements The video shows the example agent investigating the just-bash code base

Bash is all you need! Which is why I'm introducing my holiday project: just-bash just-bash is a pretty complete implementation of bash in TypeScript designed to be used as a bash tool by AI agents. Because it turns out agents love exploring data via shell scripts, even beyond coding. It comes with grep, sed, awk and the 99th percentile features that an agent like Claude Code or Cursor would use. In fact, Claude Code can use it for secure bash execution. In the package - A bash-tool for AI SDK - A binary for use by yourself or your coding agents - An overlay filesystem to feed files to your agent securely - A Vercel Sandbox compatible API, so you can quickly upgrade to a real VM if you need to run binaries - An example AI agent that explores the just-bash code base using just-bash - I imported the Oils shell bash compatibility suite and just-bash passes a very good chunk What is interesting about this codebase: It was essentially entirely written by Opus 4.5. Coding agents love bash and they are good at reproducing it. They are also great at text-book recursive descent parsers and AST tweet-walk interpreters. That said, it is, like, a lot of code and I didn't read it all 😅. This is very much a hack, but it also seems to be _really_ useful. I haven't really found anything agents want to use that it doesn't support and it's fast and secure (caveats apply). It doesn't have write access to your computer and the filesystem is given a root that the agent cannot escape from. Find it at Related: Our recent blog post how we migrated our data analysis agent to bash tools and achieved incredible quality improvements The video shows the example agent investigating the just-bash code base

Malte Ubl

124,713 views • 5 months ago

HERMES AGENT IS STARTING TO FEEL LIKE A DIFFERENT CATEGORY OF AI TOOL. Not just a chatbot or a coding assistant, but a server side agent that remembers workflows, builds skills, and gets better the more you use it.

HERMES AGENT IS STARTING TO FEEL LIKE A DIFFERENT CATEGORY OF AI TOOL. Not just a chatbot or a coding assistant, but a server side agent that remembers workflows, builds skills, and gets better the more you use it.

0xMarioNawfal

49,486 views • 1 month ago

I just love it when one minute, the timeline is so quiet and then it will be rocked on a random Sunday by series of ayudas like this hahaha 😅🫶

I just love it when one minute, the timeline is so quiet and then it will be rocked on a random Sunday by series of ayudas like this hahaha 😅🫶

˚☾˚｡⋆

12,894 views • 5 months ago

"I think it took me half a second to register that it didn't go in, and then I just screamed..." — .Nebraska Men's Basketball players' reactions to the missed half court shot by Vanderbilt 😅😅

"I think it took me half a second to register that it didn't go in, and then I just screamed..." — .Nebraska Men's Basketball players' reactions to the missed half court shot by Vanderbilt 😅😅

CBS Sports College Basketball 🏀

54,893 views • 2 months ago

An LLM with native access to find and run agents is a huge thing! This large language model (called ASI:One and developed by Fetch.ai) can do this natively when you show it a problem: • Reason about the problem • Identify live agents and tools that can help solve it • Dynamically invoke these agents and tools Honestly, this is pretty nuts! You don't need to set up anything beforehand. You don't need to write any code. Here is the best part: ASI:One has access to any agent published on Agentverse, an open directory of agents! This gives the model access to *millions* of tools it can execute dynamically when solving a task. Here is a simple example (the one in the attached video): • I went to Google Maps • I copied the latitude and longitude of a location • I asked ASI:One about relevant places around that location • ASI:One went to the marketplace • It identified and called the Google Places agent • It returned a list of museums, restaurants, and landmarks near my location Pretty awesome! You can try this now by going to You don't need to pay or register to test it out. If you want to get ideas of what's possible, check Agentverse, the agent marketplace: The model will have access to all of those tools. This is pretty awesome and opens up a whole new world of possibilities! Thanks to the team for partnering with me on this post!

An LLM with native access to find and run agents is a huge thing! This large language model (called ASI:One and developed by Fetch.ai) can do this natively when you show it a problem: • Reason about the problem • Identify live agents and tools that can help solve it • Dynamically invoke these agents and tools Honestly, this is pretty nuts! You don't need to set up anything beforehand. You don't need to write any code. Here is the best part: ASI:One has access to any agent published on Agentverse, an open directory of agents! This gives the model access to millions of tools it can execute dynamically when solving a task. Here is a simple example (the one in the attached video): • I went to Google Maps • I copied the latitude and longitude of a location • I asked ASI:One about relevant places around that location • ASI:One went to the marketplace • It identified and called the Google Places agent • It returned a list of museums, restaurants, and landmarks near my location Pretty awesome! You can try this now by going to You don't need to pay or register to test it out. If you want to get ideas of what's possible, check Agentverse, the agent marketplace: The model will have access to all of those tools. This is pretty awesome and opens up a whole new world of possibilities! Thanks to the team for partnering with me on this post!

Santiago

43,782 views • 1 year ago

Yes here is my 10 minute breathless rant about why I'm so excited about Notion Workers + Custom Agents... Context: I spent this afternoon building a custom agent to help me manage Shiori (a side project I shipped last weekend). I gave the custom agent everything it needs to understand what's happening in my product (email, log drain, sentry alerts, stripe payments, etc) and to do work on my behalf (access to coding agents). In an afternoon of tinkering, this agent can: - Diagnose bug reports proactively by looking through past email conversations, system logs, and database records - Draft replies to user questions with the correct answer based on past email threads, or help me proactively reach out to churning paid users - Self-construct a database of feature requests with an understanding of who is requesting the feature and how they're using the product today - Answer any question I have about how people use the app and what I should be thinking about next - Initiate Claude Code workflows to open PRs proactively in the background when someone sends a bug report or feature request This custom agent is now my "Side Project Chief of Staff" (I don't really know what a chief of staff does but this sounds right). I didn't write a single line of the worker code because I didn't need to: models are so good that I can link to the Workers readme, yap my desired outcome into a microphone, and I get a super-personal and highly-capable AI agent out the other side. So fucking cool. The future is now! I'm excited to see what everyone makes.

Yes here is my 10 minute breathless rant about why I'm so excited about Notion Workers + Custom Agents... Context: I spent this afternoon building a custom agent to help me manage Shiori (a side project I shipped last weekend). I gave the custom agent everything it needs to understand what's happening in my product (email, log drain, sentry alerts, stripe payments, etc) and to do work on my behalf (access to coding agents). In an afternoon of tinkering, this agent can: - Diagnose bug reports proactively by looking through past email conversations, system logs, and database records - Draft replies to user questions with the correct answer based on past email threads, or help me proactively reach out to churning paid users - Self-construct a database of feature requests with an understanding of who is requesting the feature and how they're using the product today - Answer any question I have about how people use the app and what I should be thinking about next - Initiate Claude Code workflows to open PRs proactively in the background when someone sends a bug report or feature request This custom agent is now my "Side Project Chief of Staff" (I don't really know what a chief of staff does but this sounds right). I didn't write a single line of the worker code because I didn't need to: models are so good that I can link to the Workers readme, yap my desired outcome into a microphone, and I get a super-personal and highly-capable AI agent out the other side. So fucking cool. The future is now! I'm excited to see what everyone makes.

Brian Lovin

181,088 views • 3 months ago

This employee got a call about a DoorDash order that was already delivered. The customer ordered a steak and cheese, a meatball, and a pepperoni and salami sandwich. When another employee made it, the store was out of steak and mayo so she just made another meatball sandwich without consulting with the customer first. The customer is upset and says that if they were out of steak then they should have called and let her know so she could get something else instead of having the decision made for her. The employee defended the other employee and claims they aren’t allowed to call the customer and that the situation was appropriate. When the customer gets mad she then tells her to come down to the store as if to provoke a confrontation. Was the customer wrong for ordering from a third party site that doesn’t show their inventory in real time or should the employee have let the door dasher know so they could call the customer?

This employee got a call about a DoorDash order that was already delivered. The customer ordered a steak and cheese, a meatball, and a pepperoni and salami sandwich. When another employee made it, the store was out of steak and mayo so she just made another meatball sandwich without consulting with the customer first. The customer is upset and says that if they were out of steak then they should have called and let her know so she could get something else instead of having the decision made for her. The employee defended the other employee and claims they aren’t allowed to call the customer and that the situation was appropriate. When the customer gets mad she then tells her to come down to the store as if to provoke a confrontation. Was the customer wrong for ordering from a third party site that doesn’t show their inventory in real time or should the employee have let the door dasher know so they could call the customer?

SonnyBoy🇺🇸

155,245 views • 1 month ago

Nano Banana on LovartAI is so simple and practical. Select a region, hit tab, and type what you want. Let the agent do the rest... Free to use all weekend 🧵

Nano Banana on LovartAI is so simple and practical. Select a region, hit tab, and type what you want. Let the agent do the rest... Free to use all weekend 🧵

Heather Cooper

18,820 views • 9 months ago

Tony Gwynn shares what it took to become one of the greatest hitters of all time and how to master the mental game. "It took me a while to figure it out, but after about 4 or 5 years, I started to realize that this thing was more mental than it was physical." He didn't figure it out alone. "Those first 5 years, I had great conversations with a lot of great players. Pete Rose, Willie Stargell, Henry Aaron, Mike Schmidt. I had a chance to sit down and talk to 'em - pick their brain, find out how they do what they do." The greats learn from the greats. They never stop learning. They never stop wanting to get better. "After about 4 years, I started to realize that this is more mental than physical. So what I need to do is get an approach that I think will work and then just trust the approach. Trust what you practice on. Trust it." That's the key. Create the plan then work the plan. And when it's not working, you still trust yourself. "During the winter, I can hit in a cage and be mechanically sound. But the first time I hit live in spring training, I'll be lucky if I get a ball out of the cage." "For younger guys, that's a sign of, 'Oh, I'm not getting out of the cage,' so they try to speed things up, do things quicker." Trust your approach and stay under control. Don't let frustration change who you are. The game is mental. The greats know that and they learn how to build a process that works for them. They adapt as needed, but they learn the art of focus. Focusing on how to master and trust the process. (🎥Special Collections )

Tony Gwynn shares what it took to become one of the greatest hitters of all time and how to master the mental game. "It took me a while to figure it out, but after about 4 or 5 years, I started to realize that this thing was more mental than it was physical." He didn't figure it out alone. "Those first 5 years, I had great conversations with a lot of great players. Pete Rose, Willie Stargell, Henry Aaron, Mike Schmidt. I had a chance to sit down and talk to 'em - pick their brain, find out how they do what they do." The greats learn from the greats. They never stop learning. They never stop wanting to get better. "After about 4 years, I started to realize that this is more mental than physical. So what I need to do is get an approach that I think will work and then just trust the approach. Trust what you practice on. Trust it." That's the key. Create the plan then work the plan. And when it's not working, you still trust yourself. "During the winter, I can hit in a cage and be mechanically sound. But the first time I hit live in spring training, I'll be lucky if I get a ball out of the cage." "For younger guys, that's a sign of, 'Oh, I'm not getting out of the cage,' so they try to speed things up, do things quicker." Trust your approach and stay under control. Don't let frustration change who you are. The game is mental. The greats know that and they learn how to build a process that works for them. They adapt as needed, but they learn the art of focus. Focusing on how to master and trust the process. (🎥Special Collections )

Coach AJ 🎯 Mental Fitness

95,120 views • 3 months ago