Cole Medin Puts Jev's System 1 Decisions to Work in His Game, PR Triage and LLM Router
He argues a decision-only model doesn't replace your LLMs, it takes over the classification and routing steps inside your AI workflows, and proves it with his own indie game, Archon's PR pipeline, and a four-model router.
Jev, TypeSafe's new 'System 1' model, replaces the classification and routing steps inside AI workflows with a purpose-built decision engine that is far faster and cheaper than any LLM, without competing with LLMs for reasoning or writing.
Who This Is For
Read if. Skip if.
READ IF YOU ARE…
You're building AI agents or coding workflows with classification, routing, or triage steps currently handled by an LLM call or hand-written regex.
You maintain an automation, like support routing, PR review, game AI, or browser automation, where decision speed and cost matter more than generated text.
You're trying to figure out whether 'System 1' decision models are a genuine new category worth testing or just repackaged classifiers.
SKIP IF…
You need a model that writes or converses. Jev never generates free text, so it can't replace a chatbot or a content-writing model.
Your workflow has no discrete decision points, so there's nothing here to route or classify.
TL;DR
The full version, fast.
Jev is a new kind of AI model from TypeSafe that never generates text: it takes a described situation plus a set of multiple-choice questions and returns a calibrated decision with a confidence score, trained with a new method called RLCD instead of the RLHF used for chatbots. Cole Medin argues its value isn't inventing classification but making it general purpose: the same model routes customer support tickets, plays his own indie game in real time with no LLM latency, triages pull requests in his Archon tool, and picks which LLM should handle a given prompt, all for a fraction of a cent per decision. His conclusion: Jev doesn't replace LLMs, it takes over the decision points inside AI workflows so LLMs stay free for reasoning and writing.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Cole Medin explains his model-launch fatigue, then introduces Jev as a genuinely new category: a 'System 1' model that only makes decisions and never generates text, trained with a new RLCD method instead of RLHF.
01:40 – 02:57
02 · How Jev Works
Jev takes a situation plus a set of multiple-choice questions and returns a decision with a confidence score for each; it can never produce freeform text.
02:57 – 04:14
03 · Inputs and Decisions
A worked customer-support example: Jev routes a Stripe billing complaint to the billing department at 64% confidence and scores the customer's sentiment as frustrated but civil.
04:14 – 05:32
04 · Speed, Cost and Reliability
TypeSafe's benefits claims: 0% malformed structured output, 20-200x faster than LLMs, 40-1,000x cheaper, shown on a logarithmic cost chart against models like GPT and Claude.
05:32 – 06:58
05 · Jev Plays My Game (Insane)
Cole Medin shows Jev playing his own in-development indie game in real time, hands off the keyboard, making attack/dodge/movement decisions with confidence scores as fast as a human.
06:58 – 08:54
06 · Sponsor - Firecrawl
Sponsor segment for Firecrawl, a context API for AI agents, demoed as an MCP server added to Claude Code for cleaner web search results.
08:54 – 10:42
07 · PR Reviews with Jev
In Cole Medin's Archon tool, a PR-triage workflow uses Jev only for the classify and route steps, then hands off to a full LLM for light or deep code review before reporting results.
10:42 – 12:18
08 · Using Jev for Model Routing
Jev routes prompts between four LLM tiers (strong, coding, open, fast) in the TypeSafe playground; dozens of test routes cost four tenths of a penny at roughly two tenths of a second each.
12:18 – 14:53
09 · Community Demos
Third-party demos: Jev playing Doom on X, a browser-automation tool (jev-ultrafast), Jev vs. LLMs playing Pong in real time, and the community-curated 'Awesome Jev' project list.
14:53 – 17:13
10 · More Than a Glorified Classifier?
Cole Medin addresses the main criticism (classification models aren't new) and argues Jev's generality, demonstrated with a silly example asking which country has the coolest buildings, is the real advance.
Atomic Insights
Lines worth screenshotting.
Jev never generates text: every input is a described situation plus multiple-choice questions, and every output is a decision with a confidence score.
Jev is trained with RLCD (Reinforcement Learning for Calibrated Decisions) instead of RLHF, the method behind chatbots like GPT and Claude.
TypeSafe claims Jev has a 0% malformed-output rate, versus the occasional broken JSON large language models produce on structured output tasks.
Jev makes decisions 20 to 200 times faster and 40 to 1,000 times cheaper than large language models, on a logarithmic cost curve.
Cole Medin used Jev to play his own in-development indie game in real time, hands off the keyboard, at a speed no LLM could match.
A routing workflow that picked between four different LLMs based on query difficulty cost four tenths of a penny across dozens of test calls, averaging two tenths of a second per decision.
In Cole Medin's Archon tool, Jev handles only the classification and routing steps of a pull-request pipeline, while a full LLM still performs the actual code review.
The most defensible use case for a System 1 model isn't replacing LLMs, it's taking over the decision points a workflow used to hand to a slow LLM call or a brittle regex.
Community demos already have Jev playing Doom and Pong in real time, and driving browser-automation tools like jev-ultrafast.
The main criticism of Jev is that classification models aren't new; TypeSafe's counter is that Jev is general purpose rather than trained for one narrow task.
Jev is available now through OpenRouter or directly through typesafe.ai.
Takeaway
When to trade an LLM call for a decision-only model
SYSTEM DESIGN
The lesson isn't about Jev specifically, it's about recognizing which steps in an AI pipeline are actually decisions in disguise, and swapping the slow generative call for a purpose-built one.
01Why Jev Is Different
A new model category can trade text generation for pure decision-making, which matters wherever a workflow step already collapses to picking one of a few options.
Constant AI-model hype makes it easy to tune out real category shifts; the way to tell the difference is checking whether a claimed innovation solves a problem you actually have, not whether it's new.
02How Jev Works
A model trained on a different objective, like calibrated decisions instead of human-preferred text, behaves differently even when it looks similar to an LLM from the outside.
Structured output that never fails to parse removes an entire class of workflow bugs that come from an LLM occasionally returning malformed JSON.
03Inputs and Decisions
Splitting a task into one situation description plus several parallel multiple-choice questions is a reusable pattern for any decision step, independent of which model answers it.
A confidence score attached to a decision is what lets a pipeline decide when to trust the automated choice versus escalate to a human or a stronger model.
04Speed, Cost and Reliability
Order-of-magnitude cost and latency differences, not small percentage gains, are what justify introducing a second, specialized model into a pipeline instead of just tuning prompts on the one you already use.
Cheap, fast decisions change what's economically possible: workflows that need hundreds or thousands of decisions per run only become viable once each one costs a fraction of a cent.
05Jev Plays My Game (Insane)
Real-time interactive systems like games and live agents expose latency limits an LLM can't cross, no matter how good its reasoning is, because the decision has to land in milliseconds.
Testing a decision model on a fun, visible use case like a game is also a stress test: if it can react at human speed continuously, its reliability claims get harder to dismiss.
07PR Reviews with Jev
The most defensible use of a decision-only model in a coding pipeline is routing and severity classification, not the substantive review itself, which still needs full LLM reasoning.
Tiering how much review a change gets based on a cheap upfront classification keeps expensive LLM review calls reserved for the changes that actually need them.
08Using Jev for Model Routing
Using a second lightweight model to route requests between multiple LLMs is more reliable and far cheaper than using another full LLM call just to decide which model to call.
Routing decisions this cheap, fractions of a cent and sub-second, make it viable to re-evaluate the routing choice on every single request instead of hardcoding a rule.
09Community Demos
Once a fast decision model exists, early adopters gravitate to games, browser automation, and other real-time interactive demos first, because that's where LLM latency was most visibly broken before.
A public demo where a decision model visibly outperforms a well-known LLM at the same real-time task is a more convincing benchmark than a leaderboard score.
10More Than a Glorified Classifier?
The honest response to 'this isn't new' is to concede the specific technique existed before, then show why making it general-purpose instead of task-specific is the actual advance.
A model that gives the same confident answer to a subjective, repeated query is demonstrating consistency, not necessarily correctness, and it's worth telling those two apart before trusting its output.
Glossary
Terms worth knowing.
System 1 model
TypeSafe's name for a new model category, including Jev, that only makes fast, calibrated decisions from multiple-choice options instead of generating free text.
RLCD (Reinforcement Learning for Calibrated Decisions)
The training method TypeSafe built for Jev, aimed at well-calibrated confidence scores for decisions rather than optimizing for human-preferred text.
RLHF (Reinforcement Learning with Human Feedback)
The standard training method for chat-style large language models like GPT and Claude, tuned to produce text people rate as good responses.
Archon
Cole Medin's open-source AI coding workflow tool, used in the video to run a pull-request triage pipeline that hands classification and routing steps to Jev.
“I've been kind of burnt out with all the new LLMs coming out... I want something new, and we have something new and genuinely innovative here with Jev.”
honest framing of model-launch fatigue before the pitch lands→ TikTok hook↗ Tweet quote
08:13
“It costed me four tenths of a penny to do all this routing with Jev, and the average time it took was two tenths of a second to make each one of these routing decisions.”
“It's not like you're going to totally swap all your large language models for Jev right now... the best workflows... going forward, it's going to be a combination of Jev for the decision making and LLMs for the other reasoning.”
the thesis statement of the whole video→ IG reel cold open↗ Tweet quote
The Script
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
17px
metaphorstory
So we have a new AI model that was just released within the last week called Jev, but Jev is special. Jev is not just another large language model. It introduces an entirely new class of AI models called System 1 models, which are master decision makers.
Now that might not sound super exciting at face value, but having a master decision maker is actually incredibly useful, especially because of how fast and cheap Jev is. We'll of course talk about that as well. And I'm genuinely excited for this like I haven't been excited for an AI model for a while now I gotta be honest I've been kind of burnt out with all the new LLMs coming out because every single time we have a new model there's always the rush to incorporate it in our workflows there's all the hype on the internet for look what this does for your second brain look what kind of beautiful websites you can make look at these beautiful scenes that I'm generating in Blender and you see that time and time again it's just like man I want something new and we have something new and genuinely innovative here with Jev.
Now Jev has been out for a little bit so you might have already seen a video or two on it but I've specifically waited to make content on Jev because I wanted to actually build with it as I try it out initially before I go and just talk about it. And so of course we'll start with a high level overview of how Jev works, its limitations, and the really incredible benefits but then I want to get into how I've been building with Jev and show you some really cool things I've been using with it already and then there's one other thing I want to hit on which is there's actually a good amount of criticism for Jev as well.
The idea that it's not really truly innovative and it's more just taking ideas that already exist in the industry for classification models. It's kind of true. It's something interesting to talk about.
And so we'll hit on that as well. So the most important thing to understand with Jev is it is not a large language model. You can't have a conversation with it like you can with ChatGBT, Claude, Gemini, because it doesn't generate text.
Its sole responsibility is to take a bunch of input around a... situation and generate a decision. And so with this, there's a brand new training algorithm that they called RLCD, Reinforcement Learning for Calibrated Decisions.
This is different from RLHF, or Reinforcement Learning with Human Feedback, which is the algorithm used to train generative AI models. And so the promise with GEV and System 1 models is that we are solving a lot of reliability and hallucination issues we see with generative AI, specifically for making decisions.
And I love what they say here. Our claim may sound too good to be true, but the bitterest lesson in AI is that optimizing for the right task gets you an unfair advantage. And I agree with this 100%.
I love how they sign off here. May your intelligence be ever reliable. And if you follow my channel, you know that that's a really important thing for me that I'm always chasing.
Like when I'm building Archon and my AI coding workflows, I'm all about adding as much determinism in these workflows as we possibly can. And Jev is another tool. in our tool belt to do that.
Now before I go and explain more with Jev, I want to show you what it looks like in practice. So you can see what the inputs and outputs look like. It's quite different from a large language model because it's never going to be freeform text, it's always going to be a decision.
In fact, it looks a lot like structured output that we have with LLMs, though there are a lot of differences and we'll talk about that as well. So the best way to describe the input to Jev is it's really a two -part input. You have the situation that it is analyzing and then you have a set of questions that each have a multiple choice answer.
It is never generating free form text. Remember, it is always making a decision based on the situation. And so for this basic example right here, you can imagine Jev is integrated into some kind of customer support agent.
And so this person, they're having trouble connecting their Stripe account, they're losing sales, and they need help ASAP. And so we're asking Jev a series of questions that it's going to answer in parallel. And it's always gonna pick from the criteria that we give it, multiple choice.
And so the output, quite simply here is going to be its answer, its choice for each one of the questions, as well as its confidence score. And so here it made a routing decision.
It's 64 % confident that it should go to the billing department, this complaint. And for the frustration, a score of one, that means that the customer is deemed frustrated, but civil. So doing sentiment analysis as well.
Okay, so that's cool, Cole, but large language models can already do this. They can already output structured JSON to make decisions in a workflow. And yes, that is true, but there are three.
massive benefits to Jev. It's the perfect trifecta. Jev is faster, more cost -effective, and more reliable when it comes to making decisions compared to large language models.
At least what they claim and what I've seen, it has a 0 % failure rate for structured output. So it never has any kind of malformed JSON that would make a future step in your workflow fail. And Jev is incredibly cost -effective.
This graph right here is logarithmic. And so Jev is dozens, even hundreds of times more cost -effective. than all of the best large language models.
There are a couple of models that do actually make better decisions if you give them the time than Jev, but it is incredibly cheap. Like look at the cost per million tokens of Jev compared to something like GPT -6 Astra or Cloud Fable 5 .1. And so if you combine the incredible cost effectiveness with the speed that we have with Jev, it's able to make decisions 20 to 200 times faster.
It's 40 to 1 ,000 times cheaper. That together means that you can build these systems having a lot of intelligence, making snap decisions extremely fast, and even making hundreds or thousands of decisions in parallel. So there are a million different use cases for Jev that are super powerful, very practical, but the coolest one that I've been working with right now, I wanna show you just really quickly, is using Jev to play video games exactly as a user would, making decisions as quickly as us or even faster.
And so I've been experimenting a lot. with using Jeb within my AI software factory to test out games as I'm building them as the large language model is actually writing the code. And so you can see in real time, it's making all these decisions with confidence scores for all the different actions that I'm proposing for it.
And it seems like I'm playing. It's attacking, moving to enemies, dodging, but I have hands off the keyboard. This is so cool to watch.
And I've tried to get large language models to do this kind of thing, but it just doesn't work because they can't process things fast enough. And it would be way - too expensive so this by the way is actually my game this is running on localhost right now this isn't just some like twitter demo that i have up though i'm going to show some of those as well the important thing here is that jev can't create this game but it can definitely interact with it in a way that a large language model never could so it's not like you're going to totally swap all your large language models for jev right now it doesn't work that way you're just going to put jev in your automations where you have those decision points or any kind of classification step and so the best workflows for AI coding or any kind of business use case going forward.
It's going to be a combination of Jev for the decision making and LLMs for the other reasoning. The sponsor of today's video is Firecrawl. Every AI agent that I build eventually needs access to the web.
And there are a lot of agents out there that have these capabilities out of the box like Cloud Code. But if I'm building my own AI agent with Pydantic AI or Laingraph or Pi, I have none of that. And even if you are using Cloud Code, the search capabilities built right in are very inefficient and token heavy if you haven't noticed before.
And Firecrawl has the solution for this. They call it the context API for AI agents. That's exactly how I use it.
And they have an MCP server that makes it extremely easy to bring their context API into any AI coding assistant or other AI agent. So for example, with Cloud Code here, I just copy this command, go into a new terminal, paste it in just a single line to get the MCP server added. So now when I go into Cloud...
first time i simply have to do slash mcp to then set up the authentication and then i'm good to go so i'm showing you the full flow here in just like 20 seconds authorize and then back over to the terminal authentication is successful and i can now start sending in my requests and the search capabilities of firecall are powerful it's not just a google search it's able to directly generate queries that answer my question instead of just performing a really broad web search like you'd usually see in something like cloud code so right here i asked what are people running into when upgrading to Pydantic AI version 2.
A specific but powerful example because it has to look through a lot of context to answer this, but it's able to do so in only three calls to the Firecrawl MCP server. So Firecrawl gives me exactly the context I need. It can also give me the full page as clean markdown.
The MCP server is easy to use anywhere and they also have an SDK if we want to build Firecrawl directly into our custom agent tools. So it's super easy to use whether you're building your own agent or using something out of the box like a coding agent. And if Firecrawl is free to get started with a thousand credits a month, and you don't even need an API key to use their MCP server.
I'll have a link to them in the description. Now, of course, I've been doing a lot of testing with this myself, building larger AI coding workflows with my open source tool, Archon, combining Jev with LLMs, using the right model for the right step. And so with this Jev PR triage workflow, essentially what it does is we have classification and routing at the start that figures out what kind of review we need to...
perform on a pull request and then go and do that review. And of course, for the first two steps here, classification and routing, I'm going to be using Jeb because we're just making decisions here. And so it's very cost effective, this workflow, because of course, Jeb itself is cost effective.
But then also we get to decide what kind of review we're doing, because we don't always need a super deep AI review on every single pull request. And so this is just one simple example with a lot of testing that I've been doing with our So also let me know in the comments if you want me to make more content on this.
I'm definitely going to continue to explore using Jev within AI coding workflows for testing things like I showed you with my game, for classification steps with things like issue triaging and pull request review. The possibilities are endless. And if you've been following my channel in Archon and you're curious, this is the exact workflow that you just saw in the Archon UI.
So I just simply call this Python script that makes the classification with Jev. And so for... my Jev usage, I'm going directly through OpenRouter.
So OpenRouter was super fast to make the Jev model available along with all the other LLMs that you can use there. And then also if you want, you can go directly through typesafe .ai. Typesafe is the company that created Jev.
And so of course I'll link to this in the description. By the way, they're not sponsoring this video at all. I am genuinely excited for this model.
I hope that you are too, just going through some of the use cases with me here. And if you're not sold yet, let me show you some more use cases. So this is some...
experimentation I've been doing myself. I've seen a lot of other people on the internet do this as well, using Jev as an LLM router. It's a really common use case where you have a bunch of different LLMs that you want to pass the right requests to, right?
Like sometimes for the sake of cost, the simpler requests go to the faster model, ones that require deeper reasoning, we want to send to the strong model. Traditionally, you've used yet another LLM to make the routing decision. But again, with Jev, even with tiny LLMs, it is going to be faster and cheaper.
and of course more reliable so the situation we give as input to jev is the query that we want to route and then the multiple choice that it has to answer is which one of these models should we route it to the strong the coding the open or the fast and this is just a quick visualization i put together to show you all the testing that i've been doing but for a deeper question it routes to the strong model with a confidence of 100 percent this, you know, convert this bash to PowerShell, a little bit of a coding task.
It has a 98 % confidence going to the coding model and some faster ones here, like convert 72 Fahrenheit to Celsius. Yeah, this can definitely be handled by a cheaper model. Like GPT 5 .6 Luna is the faster one here on open router compared to, I don't know, like what's the strong one here?
Yeah, Claude Sonnet 5, for example. Now these numbers aren't the best. This is just a really small subset of all the testing that I've been doing.
But the really cool thing to show you here is that out of the dozens of the tests that I have visualized here. It costed me four tenths of a penny to do all this routing with Jev, and the average time it took was two tenths of a second to make each one of these routing decisions.
Super cool. Okay, so that's enough of my testing. I hope you liked it, but let me show you really quickly what other people have been sharing on the internet as well.
So this person on X posted using Jev to play the classic game Doom, and it actually looks a lot like my own testing with my own game, where we have the decisions that are being displayed on one side right here in real time as Jev is playing the game. And then of course the game itself.
It's so cool how it's able to play something like this. And yeah, it's a basic game because you can't just give it like millions of decisions, but this is still incredibly impressive. And of course, I'll link to all of these resources in the description.
Another really powerful use case for Jev that you probably thought of at this point is using Jev for browser automation. So more traditional tools like Playwrights and Vercel's Asian Browser Seal, lie, it's always driven by an LLM.
And I use these every single day as I'm building web apps, full stack apps, but it's always slow, right? Like the slowest part of my AI coding workflow for any kind of full stack app is always when it has to validate things visually and navigate the browser. But now we don't need an LLM to do it.
We can use Jev because every single situation is the current layout of the site. And it just has to decide with multiple choice, the next action to take, like click this button or type in this input. And so this is just one really cool open source repo that I've seen.
There's probably going to be a lot of Jev browser use tools released in the next couple of weeks, but this is one of them. It works incredibly well. And then I also found this really cool visualization of Jev playing Pong.
And so that's the top row right here. The game is slowed down basically to the rate that the model can handle. And Jev can pretty much handle it at a human rate.
And then other LLMs down here, like 3 .8 Flash, Cloud Haiku 4 .5, you can see the game has to be slowed down. a lot for it to actually process where the paddle needs to be as the ball is coming. And then one last resource I want to show really quick is this open source repo that curates a list of projects and just general use cases for Jev.
So I'll scroll down in the readme to current coverage. We got classification and routing. Of course, that's going to be the most common one.
I mean, that's literally what Jev is made for. But then using that for agent decisions, verifications and guardrails, calibration and research, games and simulation, finance and trading. There are so many cool...
use cases to poke around here. So yeah, I'll link to it in the description. Just check this out.
You just get your imagination going here as you go through these different use cases, like I'm trying to do for you in this video. Because when you think of Jev as this glorified classification model, it's not very exciting at first, but once you realize what you can do with it, man, the world becomes your oyster here. And speaking of Jev being a glorified classification model, that's the last thing I want to hit on really quickly, because it's the biggest criticism that I've seen.
for Jev, and I've actually seen it quite a bit. A lot of people say that we've had the idea of a classification model in the AI industry for decades, and so we're just reinventing the wheel here. And it's true to an extent, especially when we use it for very basic things like this example I have in the TypeSafe playground.
But what really makes Jev powerful is how general it is. Like, I think the best way to describe it is it feels like there's still the intelligence of a large language model operating behind the scenes. producing the structured output, answering our different questions.
Like I just had this really silly one right here, which country has the coolest buildings? I gave it a few options and it's actually says Germany with 63 % confidence, a very opinionated thing. I mean, who knows what it's going off of, but if I run this over and over and over again, the numbers change a little bit, but it actually always says Germany, which is interesting.
Very cool. So anyway, anyway, I've built a lot of classification models in the past, but it's always for a very specific task and you have a very specific data set so i've used you know like tensorflow and pytorch to build classification where you give it a chess position and it says who's winning or it looks at an animal and identifies which animal it is but then if you have it do a different task it can't do it at all because it's so specific but with jev it's classification in the general sense you can give any kind of situation for customer support or opinions on countries if you really want and it's able to give you a response here now this is kind of a silly example, but for anything more objective, like what kind of pull request review level does this need?
Or what model should we route to here? Or what's the next best action in this video game? Like Jev can just handle any of that, and it's so accurate.
So I hope that you found this interesting. All the super cool use cases for Jev, I would encourage you to try this right now, either through OpenRouter. They also have a waitlist that I was able to get into within a day.
I'll link to that in the description as well. And I will certainly be doing a lot more content. especially with Archon and how I'm using it in my AI coding workflows.
So stay tuned for that. So with that, if you appreciated this video and you're looking forward to more things on AI coding and Jev and system one models, I would really appreciate a like and a subscribe. And with that, I will see you in the next video.
The Hook
The bait, then the rug-pull.
Cole Medin says he's gone numb to AI model launches this year, but Jev pulled him out of it: a model that can't write a single sentence, only make fast, calibrated decisions. He spends the video proving that claim on his own indie game, his PR pipeline, and his LLM router before addressing the obvious pushback: haven't we had classification models for years?
Frameworks
Named ideas worth stealing.
02:57model
Situation + Multiple-Choice Questions
Situation description
Set of typed multiple-choice questions
Every Jev call describes a situation once, then asks parallel multiple-choice questions; the response returns one choice plus a confidence score per question.
Steal forany workflow gate currently handled by an LLM call or a regex
10:42model
Strong / Coding / Open / Fast router
strong
coding
open
fast
Jev classifies an incoming prompt against four LLM tiers and routes to the cheapest model that can still handle it.
Steal forany multi-model agent stack trying to cut per-request cost
08:54model
Classify -> Route -> Review -> Report
classify
route
light-review or deep-review
report
Archon's PR workflow uses Jev only for the first two decision steps, then hands off to a full LLM for the actual code review.
Steal forCI/PR bots that don't need full LLM review on every diff
CTA Breakdown
How they asked for the click.
VERBAL ASK
16:52subscribe
“I would really appreciate a like and a subscribe.”
Soft, single-sentence ask at the very end after the full demo, no upsell or urgency, tied to wanting more content on Jev and Archon.
Add Modern Creator as a preferred source and Google shows you more of our breakdowns in Search, Top Stories, and AI Overviews. It only changes what you see, and you can undo it in your Google settings anytime.
Add to Preferred SourcesOpens your Google source preferences with us pre-loaded. Tick the box and you're done.
A lightweight alternative to built-in computer-use tools: one markdown skill and a handful of Python scripts that let a coding agent click, type, and screenshot its way through any desktop task.
A creator known for teaching people to build their own AI tools explains why he's now just handing you one: an open source pipeline that turns a single planning document into shipped code nobody reviewed.
A software engineer who hasn't handwritten a line of code in over a year opens up the actual skill library and two-loop system that runs his AI-assisted development process end to end.
Boris Cherny told a Y Combinator crowd to delete your AI layer every six months. Cole Medin ran the actual benchmark to find out which parts of that advice hold up and which don't.
00:19
00:32
00:45
00:58
01:11
01:23
01:49
02:02
02:15
02:28
02:41
02:54
03:07
03:20
03:33
03:45
03:58
04:24
04:32
04:50
05:01
05:12
05:42
06:08
06:20
06:33
06:46
07:12
07:25
07:38
07:52
08:04
08:17
08:30
08:42
09:08
09:21
09:34
09:47
10:00
10:13
10:26
10:52
11:04
11:17
11:30
11:43
11:56
12:09
12:35
12:48
13:01
13:14
13:32
13:39
13:52
14:05
14:18
14:31
14:44
15:10
15:23
15:36
15:49
16:01
16:14
16:27
16:40
16:53
13:01
14:08
23:07
23:00
Video of the Day20:00
21:55