24:21Greg Isenberg · Talking Head
Karpathy's autoresearch broke the internet
A 24-minute solo breakdown of the AI experiment-loop tool that went viral — and 10 businesses you can build on top of it.
March 11thThe argument in one line.
Jev is a decision-only AI model that returns a probability for each schema option in about 200 milliseconds, letting businesses classify high-volume data like email or leads in real time for a fraction of a cent.
Read if. Skip if.
READ IF YOU ARE…
- You run or build for a business with a high-volume inbound queue, emails, leads, support tickets, contact forms, currently triaged by hand or by a slow LLM call.
- You're a builder or founder looking for a technical wedge to build a startup around fast, cheap automated decision-making.
- You already use LLMs for classification tasks and want a faster, cheaper alternative for high-volume routing decisions.
- You're curious what a non-chat, non-generative AI model looks like and how it differs from ChatGPT or Claude.
SKIP IF…
- You need an AI model that reasons out loud, holds a conversation, or generates freeform text, Jev is explicitly not that.
- You're looking for a model to make high-stakes financial or trading decisions; the video's own Bitcoin test shows Jev underperforming there.
The full version, fast.
Jev is a new AI model built purely to make decisions, not to talk. You give it an input and a schema of possible answers, and it returns a probability for each option in about 200 milliseconds, with no visible reasoning and no streamed text. Ryan Vogel demos it sorting 1,700 of his own emails by category, priority, spam score, and reply likelihood for a total of 18 cents. The pitch: put Jev at the front of any business's expensive incoming queue, emails, leads, support tickets, so low-confidence items get ignored, mid-confidence items get automated, and high-confidence items get routed straight to a human. It's cheap enough to test for a few dollars, but the video's own Bitcoin-trading experiment shows it's not built for high-stakes judgment calls.
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.
Create a free account →Where the time goes.
00:00 – 02:27
01 · Intro
Greg Isenberg introduces Ryan Vogel and previews a new type of AI model built purely to make decisions.
02:27 – 04:32
02 · What Jev Is and Why It Matters
Ryan defines Jev as a classifier: give it an input and a schema, and it returns a probability for each output option.
04:32 – 07:19
03 · Email Triage Demo
Ryan runs Jev live on his own email inbox, scoring category, priority, spam likelihood, and reply likelihood.
07:19 – 15:46
04 · Jev as an AI Decision Maker
Greg and Ryan build a shared mental model for Jev as a decision model, contrast it against chat models like ChatGPT and Claude, and run a letter-by-letter 'Hello test' to show it isn't a text generator.
15:46 – 20:48
05 · How to Use Jev in a Business
Ryan frames Jev as an 'AI traffic cop' that decides what incoming information is, how important it is, and what should happen next, then walks through real use cases like lead scoring and support ticket routing.
20:48 – 22:51
06 · Startup Idea: Local Services Matching and Instant Quotes
Greg pushes for a startup angle: put Jev at the front of any business's expensive incoming queue, illustrated with a local-services platform that turns a delayed quote into a genuinely instant one.
22:51 – 24:03
07 · Use Case 1: Bitcoin Signal Test and Limits
Ryan shows a live Bitcoin buy/hold/sell test where Jev underperforms a frontier reasoning model, marking the edge of where a pure classifier should and shouldn't be trusted.
24:03 – 25:27
08 · Use Case 2: Auto-Clipping Long Videos
Ryan demos a tool he built in about 10 minutes that transcribes a long video and has Jev score the best short-form clip moments in seconds.
25:27 – 26:18
09 · Use Case 3: Browser Control: Flight Pick in 7.1 Seconds
A clip shows a Jev-powered browser agent picking a flight from Zurich to London in 7.1 seconds.
26:18 – 27:25
10 · How to Get Access
Ryan explains that Jev is on a waitlist, but instant access is available today through the Vercel AI Gateway.
27:25 – 28:24
11 · Closing Thoughts
Greg closes by committing to try Jev on his own email and urging viewers to test it, noting the cost is negligible.
Lines worth screenshotting.
- Jev is a classifier model: you give it an input and an output schema, and it returns a probability for each possible answer instead of generated text.
- Ryan Vogel used Jev to categorize, prioritize, and spam-score 1,700 of his own emails for a total cost of 18 cents.
- That 1,700-email run used 4.2 million input tokens and 500,000 output tokens, start to finish.
- Jev responds in about 200 milliseconds per query regardless of how complex the input or output schema is.
- Unlike a chat model, Jev never shows its reasoning; it returns only the structured numbers and categories defined in the schema.
- A decision from Jev is really a probability distribution: a call that looks like a flat yes might actually be 83% yes, 17% no under the hood.
- A $5 starter credit covered two days of heavy testing for Ryan's team, and $10 could plausibly last about three months.
- The suggested business model is to find a business with an expensive incoming queue, emails, leads, local service requests, and put Jev at the front of it.
- Jev scored a graphic design agency's inbound contact-form leads from 0 to 1, so high-scoring leads get a fast reply and low-scoring ones don't eat the owner's time.
- Jev underperformed at a live Bitcoin buy/hold/sell test, while a frontier reasoning model that cross-references news did better, showing Jev fits routing decisions, not high-stakes calls.
- A separate demo used Jev to score 17 clip-worthy moments in a long video in about three seconds, turning a manual clipping job into a near-instant one.
- In a browser-automation demo, a Jev-driven agent picked a flight from Zurich to London in 7.1 seconds.
- Instant access to Jev runs through the Vercel AI Gateway; direct access is otherwise on a waitlist.
A New Kind of AI That Only Makes Decisions
WHAT TO LEARN
Jev proves that not every AI problem needs a chat model. For fast, cheap, repeatable decisions on a stream of incoming data, a classifier that returns a probability in 200 milliseconds beats a language model that reasons out loud for thirty seconds.
02What Jev Is and Why It Matters
- Jev is a classifier: you hand it an input plus a schema of possible answers, and it returns a probability for each option instead of writing text.
- The 'what color is the iPhone' example shows the mechanic: Jev might return 80% orange, 10% red, 10% blue, all summing to 100%.
03Email Triage Demo
- Running Jev across 1,700 real emails, scoring category, priority, spam likelihood, and reply likelihood, cost a total of 18 cents.
- That run used 4.2 million input tokens and 500,000 output tokens, evidence of how cheap classification-only calls are compared to generative ones.
04Jev as an AI Decision Maker
- Jev never generates or reasons in text: it returns only the numbers and categories defined in your schema, which developers can drop straight into code with no parsing step.
- A live 'Hello test', asking Jev to pick letters one at a time, shows it isn't built or trained for open-ended text generation, reinforcing that it's a decision model, not a chat model.
- Every Jev call resolves in about 200 milliseconds regardless of the input or output schema's complexity, versus up to thirty seconds of streamed output from a chat model.
05How to Use Jev in a Business
- A graphic design agency uses Jev to score inbound contact-form leads from 0 to 1, so a 98% lead gets a fast reply and a vague inquiry gets set aside.
- The 'AI traffic cop' framework: Jev decides what incoming information is, how important it is, and what happens next, routing high-confidence items to a human, mid-confidence to automation, and low-confidence to be ignored.
- Jev is pitched as an advisory layer, not a full replacement for human judgment, on any workflow that involves looking at data and making a fast decision.
06Startup Idea: Local Services Matching and Instant Quotes
- The startup framework: find a business with an expensive incoming queue, leads, tickets, requests, and put a Jev-based classifier at the front of it.
- Example: a local-services matching platform where typing a request gets an instant best-match business instead of a 'we'll email you by end of day' form response.
07Use Case 1: Bitcoin Signal Test and Limits
- A live buy/hold/sell Bitcoin test shows Jev underperforming, while a frontier reasoning model that cross-references live news beats it on that specific task.
- The stated limit: Jev is built for fast routing decisions, not high-stakes judgment calls like trading, where more model intelligence is worth the extra latency and cost.
08Use Case 2: Auto-Clipping Long Videos
- A video-clipping tool built in about 10 minutes transcribes a long video and has Jev score the best short-form moments in around three seconds for 17 scored clips.
09Use Case 3: Browser Control: Flight Pick in 7.1 Seconds
- A Jev-powered browser agent picked a flight from Zurich to London in 7.1 seconds, versus an estimated one to three minutes for other browser-control AI agents.
10How to Get Access
- Direct access to Jev is waitlisted, but instant access is available today through the Vercel AI Gateway.
- The suggested first move: ask your own AI agent which of your daily workflows could be handed to a fast, cheap decision model like Jev.
Terms worth knowing.
- Classifier
- An AI model that sorts an input into predefined categories or scores, rather than generating new text.
- Schema
- The defined set of possible output categories or value ranges you hand the model alongside your input, telling it exactly what shape the answer should take.
- Decision model
- A term used in the video for an AI model, like Jev, that outputs only probabilities and structured choices, with no visible reasoning or generated language.
- AI Gateway
- A Vercel product that provides hosted access to multiple AI models, including Jev, through one API.
- Confidence score
- The probability, expressed as a percentage, that the model assigns to a given answer being correct.
Things they pointed at.
Lines you could clip.
00:00
“Jev is here, and it's a big deal.”
06:53
“We had 4.2 million input tokens and 500,000 output tokens. The entire cost was 18 cents for each one of those emails, all categorized.”
09:05
“You don't really ask Jeff because Jeff isn't a text model.”
10:49
“Jeff seems like a circle guy.”
13:39
“It just gives you a decision. That's the way to think about it. It's a decision model.”
19:41
“It's basically this, you know, AI traffic cop.”
23:28
“I would not put this model in front of like your stock portfolio or Bitcoin or anything like that.”
25:48
“This is Jeff controlling this browser to pick a flight from Zurich to London in 7.1 seconds.”
27:30
“It is dangerously addictive. Once you see the speed and once you see the price, you will just be like, holy cow.”
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.
metaphoranalogystory
Jev is here, and it's a big deal. It was created by Diogo Almeida. Yes, that's the same guy whose research built ChatGPT.
Now, it's such a big deal because it's a whole new way to do AI. So I brought on my friend Ryan, who's on the founding team of OpenCode, to just come on and clearly explain what Jev is, what are some insane use cases, and break down some startup ideas that are now unlocked. As of publishing this, Jev is invite only, but good news, by the end of the episode, you're going to see how you can get access today.
So you're going to want to like, comment, and subscribe right now so your algorithm knows to bring you content like this to get your creative juices flowing in the future. Happy Jev Day, and I'll see you at the end of the episode.
Ryan Vogel, welcome to the pod. By the end of the episode, what are people going to learn? We're going to learn about a new type of AI, a type of AI that we haven't really seen before, but I think it's good.
It's Jev, and people are ready for this new type of classifier AI because we've been so used to just learning and using these LLMs, which are slow, they stream, and I think, as we'll cover today, This AI is fundamentally different in so many different ways with quality, speed, and price that there are so many different usage applications for it that the possibilities are truly endless and it just becomes on the humans again about how creative you can be.
I just have a few things I need from you because I haven't used Jev. I want you to give me the simplest possible explanation of Jev. I want you to give me three or four insane use cases so that people can walk away from this episode with productivity, making money, even boring use cases that could become $10 million businesses, $100 million businesses.
I just want you to put it all together, wrap it in a bow that people understand. You know, if they stick around to the end, that they'll be able to understand why should they care about it. Can you commit to that, Ryan Vogel?
I can. I can. And I'll add one better.
I'll make it entertaining so that way you can actually get excited about it. Because first up, I'm just going to start out with a demo. This is my email.
I'm not afraid to share it. I've been working with email. If you know me at all, you know that I love email because it seems unsolved.
I mean, like. Greg, how many spam emails do you get every day? Like there's too many, right?
There's too many. You can't reply to all of them. And it's just so frustrating.
And some of the email algorithms that exist are good, but it's not the best. But then some people are trying to like take like traditional AI where it's like they're having like a GPT 5 .6 Luna, like kind of read every email and then score it. But that takes time and it's not like instant.
And it's just like, oh, I wish we could just have something that could like instantly categorize all the emails. So this is that this is using Jev. And before I run it, I'm going to break down Jev in a super simple example.
Jev is a classifier at its truest being. That's what it is. I won't get into the architecture and stuff like that, because honestly, I don't even understand it that well.
But essentially, you define an input. Let's say you have this iPhone as an input. Right.
That's the input. And then the output is a schema. So we could have the schema be what color is the iPhone is the question almost.
And it has blue, orange, red, green, yellow as the output options for that question. And the classifier, Jev, then looks at this phone in a text format and says, hmm, what is this orange? Is it red?
It could be red. But then. It says, OK, this is about I'm pretty confident it's 80 percent orange, but it could be 10 percent red or it could be 10 percent blue, which adds up to 100.
And it's the probabilities of those choices. So it's not just going to be a 100 percent affirmative. This is orange.
This is blue. This is red. It's a hey, I'm 80 percent confident that this is orange or this is red.
And the best way to illustrate that is with this email example. So each one of these rows that you see on the table is a full email object. It's got a subject, it's got a description, it's got a body, it's got a sender, all the snazzy email jazz.
And what the input is, is that just entire email object. There's no sugarcoating or any special treatment. It's just the email object.
And we have four outputs. We've got a category, which is an option where basically it can say, is this shopping, work, marketing, finance, security, yada, yada, yada. Then we've got a priority, which it can allow to select from, I think, five different options where it's like low priority, medium, high, important, or urgent, which is like, oh no, you have a missed credit card payment or something like that.
That's obviously urgent. You want to be able to nail that right on the head as soon as that comes in. And then we have a spam score.
This is what I was talking about with those percentages. Obviously, not every email is going to be a true or false when it comes to spam. It's going to be a percentage.
It's a range, if you will. So it's like some emails are more spammy, like this Kickstarter one. It's obviously trying to sell me a bunch of stuff and junk.
I don't really care about that. I signed up for that Kickstarter thing like two years ago. Still haven't been able to unsubscribe from the list since.
And then we've got some like Mercury things. Okay, this is just like a payment thing. It's like, okay, Exxon Enterprise received $22 from Stripe.
That doesn't seem spammy. That seems just like it's informative and it's just informing me that something happened. And then we've got the reply percentage.
This is how much does this warrant your reply? So if we go back here and I'm not going to click on this because this is a real email, but 90 % account violation possibility. This is a user saying, hey, my account seems to be violated somehow.
Jev identified, hey, this user seems to be having some trouble. we should probably warrant a response on this. Now I've already got these all categorized and there are 1 ,700 of these emails.
And this is where we come back where it's so sad because it just takes so much time to run all of these. And it's probably gonna take like 10 hours to do. And then I'm gonna have to go through and probably pick out some of the data.
And oh my God, the price is gonna be so expensive. And oh, it's done. Oh, it didn't cost 18 cents or 1700 emails.
That is the power of Jeff. I can't explain it any better than that. We had 4 .2 million input tokens and 500 ,000 output tokens.
The entire cost was 18 cents for each one of those emails, all categorized, all. I mean, you can see here, they're all categorized. They're all ranked.
They're all given that score. So if you were to imagine, let's say... Ryan, here's what I'm hearing.
I just want to make sure I have a good mental model for what Jev is. And correct me where I'm wrong, okay? So Jev is basically like an AI decision maker.
So you give it some information. In this case, you're giving it the contents of the email and a set of possible choices, like is it spam or not? Jev's going to go ahead and look at that information and choose...
an answer so for example like is this email spam or urgent or no normal um but you can also have it do things like you know is this customer likely to buy or unlikely to buy right exactly right you're almost there that's like 90 correct it makes it it makes a probability of a decision okay so the difference between it making a decision because a decision would be you uh like you submit an API or something like that, and it tells you buy or not to buy.
Technically, what happens on the underside is that percentage. So it would be like 83 % buy, 17 % no buy type of thing. And obviously the answer that is the stronger percentage would win and that would get returned to you, but it's not a 100 % decisive action type of thing.
Okay, so instead of asking... Chat, JPT, Claude, whatever. Read this email and explain what I should do.
The new mental model is use Ask Jev. You ask Jev, read this email and choose a set of actions. So reply or escalate.
And then you get a choice from Jev and that gives you some sort of confidence score. Is that the way to think about it?
The LLMs that we know nowadays have corrupted our minds so much because there's an interesting point you said. You said, ask Jeff. You don't really ask Jeff because Jeff isn't a text model.
What's really interesting, if you look at the actual spec of Jeff, it doesn't generate any text at all, which you're like, okay, that's kind of weird. It obviously generated text because how did you get the data for this, right? That was defined in the schema.
So let me see if I can pull up a little whiteboard here, a little whiteboard action. Not too good at this. So we've got our schema, right?
We'll call it, I don't know. We'll have our email, right? And this will be our email input.
And then we'll do a circle for Jeff. Jeff seems like a circle guy. I would say Jeff.
There's the entertainment you promised. There we go. .
That's just the type of guy that Jeff seems like. Okay. Maybe a tiny circle.
There we go. Tiny circle. Cause it's fast, you know, it's fast and cheap.
Okay. So we've got our email and that goes in to Jeff. It doesn't get asked to Jeff.
It doesn't, you're not asking Jeff, Hey, what should I do with this email? It's just an input, like an eight, like a standard API. And you define a schema up here.
And we'll have a, we'll have like a simple little schema and be like, is spam. And that can be what they call a null, which is a true false, but it's a scale. So it could be a 1 to 0.
Let me format this. Yeah, I told you I wasn't good at whiteboards. I don't know about this.
So it could be a 1 to 0, which means that it could be 0 .31. Or it could be, I don't know, like 9 -0. And that's that percentage.
So if it were to return is spam 0 .90. that would be a 90 % chance that it is spam type of thing. So it doesn't give those definitive answers, but you can infer definitive answers from that sort of choice.
And then, let me get rid of this. Why are we doing JSON? And then we could have a choice, like, let's see, category.
And that would be like marketing.
It could be finance. It could be spam. And it doesn't.
generate the categories itself. It looks at the categories that you've passed into it as a model. Because you pass all of these, essentially this output schema in, and you say, here's the email, here's the output schema, I need you to generate the answer for me.
By the way, a schema is just a fancy word for how a database is organized, right? It's just how the database is organized, but not even the database. It's just how the output is organized.
It's just a fancy way, which is why the develop, why all the developers love it because they're like, oh my gosh, it's actually type safe, which is a whole nother video on everything like that. But it just means that you can take the output that this Jeff model gives you. and instantly use it in code.
Because like this null that it returns is a number object. It's not like text that is a number or something weird that you would have to do some additional data processing on. It just basically gives you this object, which is the structure of the data.
And so like, let's say we pass in this email and we have these two classification categories. So then the model would just evaluate, okay, is this spam? And what's the category?
And it would just return the percentage and the category. So it's not exactly like generating text, like in a traditional, like LLM, like chat GPT. It's not saying, hmm, well, I think this is a spam email from Kickstarter.
So I should probably rate it. Nope. It just says category spam is spam.
90 % type of thing. There's no internal reasoning or anything like that, which is why people are like, well, I don't know if I can trust it because. The whole recent development with AI, as you've probably seen, is the models are reasoning, which is basically just saying the models are speaking out loud to identify possible issues in their sort of thought progression.
And Jev doesn't do that at all. Or it might do that, but it might just do it like really fast on the server. We don't really know.
But from our point of view, it doesn't reason. It doesn't have any other text output. It just gives you the output.
So just shoots it back. It just gives you a decision. Exactly.
That's the way to think about it. That's the way I'm starting to think about it. It's a decision model.
And that's what I pointed it out.
Like right here, like all of these are just decisions. It's not because everyone has started to assimilate AI with LLMs, which is like that next token prediction where it's a conversational agent. This isn't that at all.
This is still AI because it's like machine learning, but it's a decision model strictly. You can't ask it to be like, hey, how are you doing today? Or can you?
So I like to think around with these ideas a little bit. And I was like, okay, it's a decision model, right? Well, I'm a decision model.
When I'm typing on my keyboard, I'm making the decision to type each letter. So like if I were to type hello, I'm making the decision to type H -E -L -L -O. which is technically text, but I'm also making the decision for each key.
So I'm like, what if I can apply that same principle to Jeff? So if we go back to our Excalibur whiteboard here, let's say instead of this category, we just have all of the letters A through Z, right? And each one of those letters is a new one.
So the model can basically predict each letter and say, okay. What's the percentage, what's the decision of this letter based on previous letters? So if it types H -E -L -L, it's like, okay, my next best decision is to type O to complete the word hello.
And I didn't know how it would work, but this is how it worked. So this is me asking it the prompt, what is bigger, a cat or an elephant?
And this is all real time, by the way, so it's very fast. But obviously, it's not as trained in these sort of next letter completion stuff. But it's still fun to see because it's just like, this is cool, but it also shows this isn't a traditional type of LLM where you can talk to it and it's a conversation.
It's a decision -based LLM, which we've kind of learned. So I guess that begs the question around, what should I use Jeb for? Especially the person listening to this.
is someone who wants to build a business, who wants to invest in themselves, who could be a side hustle or their own thing. And they see this and they're like, I notice that this is really interesting. I believe Ryan when I hear him talk.
And I could see that this is a glimpse into the future, but I don't know how to use it. Right. And there's something really interesting about this because this is the first model.
That's been, that can cater to a lot of different applications, which I'll say in a second, but it's also really fast and really cheap. So it's, you don't have this high barrier to entry that we've seen with other AIs where it's like, okay, I've got to dedicate like a thousand dollars a month to this. You could dedicate like $5.
Like when we got, when our open code team got set up on this account, we had like a $5, like, I guess like intro. uh credit i guess on the account we were able to use that for two days without hitting it and we were using it like a ton like all of my demos and everything like that we were using it so it's extremely cheap so you could probably like load 10 bucks on it and be good for like maybe three months but some cool things that you could probably use with this is um i already got my girlfriend working on it because she runs a graphic design agency And she gets a lot of inbound and she needs to know if this inbound is high quality or just like if it's just maybe like solicitation spam because she has a contact form on her website.
So she's using Jev to essentially say, is this a good lead? And it basically does that same sort of category where it's like is good lead and it ranks that on a percentage. So it's like is good lead.
And it ranks it from one or zero to one. So if it's like, if you get a 98 % lead, that's a pretty high lead. And you're probably going to want to reply to that.
But then if you get someone who's like, yeah, I think I might want graphic design, but I'm not too sure. They probably don't know what they want. And that would probably require more effort from you as a business owner or her as the graphic designer to sort of feel out that client.
So you can use Jev to make a lot of the decisions in your business that you might have to do yourself. So like going through, I love the email example just because it's such an easy fix. That way you can go through all your emails and all your historical emails and be like, are there any leads I missed?
Are there any high value clients that I could maybe attack again to see if I can extract more value for them and me? And basically you can kind of think through your workflow and anything where you're looking at some data. It can be any type of data.
If you're looking at some data and thinking, hmm. I have to make a decision on this. You should probably think about adding Jev at that layer.
Obviously not for like 100 % of interactions and stuff like that. It should be a very heavy advisory role. But Jev is really good because it can make those split second interactions.
If you run a business that has a contact form or like for issue triage, let's say you get a lot of support inquiries. and someone comes in and asks you, and they're like, hey, I need help with XYZ product. Jeff can do instant classification and say, okay, let's make the decision.
What product team does this need to get routed to? Let me route it over here. Let me route it over here.
And there's so many different things where if you say, hmm, this is a decision. Maybe I can use Jeff here. I guarantee you will have good results.
And it will be super cheap and fast because it takes around 200 milliseconds per query. to Jev no matter like what the input output structure is. So that is something really shocking too, because AI can take like up to like 30 seconds for some things.
And you normally have to do like streaming where then you wait for the response to be done. And then you've got to like have a listener and it's all this complex stuff. But with Jev, you can just do like a boom, like quick API call and it just works.
So Jev is basically this, you know, . So there's information that needs to come in. And then Jev is going to decide where it should go and what should happen next.
So Jev is basically going to pump out What is this information? How important is it?
What should happen next? And it's either going to go to a human being, in the case of your girlfriend's agency, where it's like, oh my God, this is a lead that she needs to act on right now. This is Coca -Cola.
It's the CMO of Coca -Cola. But if the confidence score was lower, but also a local... business in Orlando, maybe it's you automate it or use an LLM to do something, draft something up or send something.
Or the confidence is so low that you just ignore it. So am I getting that right? Yeah, that's like spot on where you basically think of anything that you would have to make a decision that would need to be quick and fast and maybe like provide feedback to a user and you can do it with that.
So where my brain goes with, you don't know me too well, but I'm all about startup ideas. That's what this podcast is about. Oh, me too.
My brain is always thinking the next way to do something like this. So I'm kind of like, oh, wow. So Jev now exists.
How do I find a business with an expensive queue of incoming information and then just put Jev at the front of that queue? What I mean by that, what do I mean by a queue? You've got like a lot of inbound coming in where people need stuff from you and you need to get them routed to the correct person.
So something that immediately comes to mind, which would require a little bit of architecture, but let's say you run a services aggregation business, like a, like a SAS level on top of a local, a lot of local services stuff in your area. And you type in and you say, hey, I need my driveway power washed, right?
And Jev could take in that information and then it could take in a lot of the input stuff of like all of the other businesses in the area and it could like return percentages of which one would probably be the best fit for you. So you type in a form and then you get an instant match with a company that's like near you. It could require, it's obviously a little bit more complicated under that, but Jev could do stuff like that where whenever you, You know, the forms that you always see when you're trying to sign up for a website and it's like get an instant quote and it's never instant.
And it always is like, we'll email you by end of day. Then a lot of that stuff can be put into like a classifier and it could genuinely be an instant quote that they could get to say, hey, this is a good match. Hey, this isn't a good match.
And Jeff could be used to do that. And that's why the speed of Jeff is nice, because then that client could see you're not wasting the client's time, which. If that client does become your client in the future, that's an insane, insanely good virtue signals to say, hey, we're not trying to waste your time.
We're not trying to waste our time. Let's get this done and work on it together. So what other Jeff use cases do you want to show?
Let me see. I was I was messing around with this and it doesn't seem to be doing well, but I wanted to see if I could hook Jeff up to a Bitcoin signal. So basically every minute it would run and it would have this decision mix right here where it would.
tell me to buy, hold, or sell. And it does not seem to be doing well, which shows that this model is great, but it does have some regressions. I would not put this model in front of like your stock portfolio or Bitcoin or anything like that.
This is just for like routing or other sort of decisions like that where it doesn't need insane and model intelligence. Like I did a test with this with GPT -6 Astra, the OpenAI's latest frontier model. And it did a little bit better than this because it cross referenced some news information and everything like that.
But that's, it's, it's completely, uh, it's not apples to apples comparisons, apples to oranges, because it's just a different type of model. So that's where it's like, this is something that a classifier and decision maker could be used to do, but it's not the best in all of the situations and everything. And I also, let me see if I can find it.
Yeah, right here. So I made a little, I made a YouTube video here where everyone who makes content is aware of this issue, where you make content and you make like a longer form YouTube video or something like that, but you want clips. And the cool part about this is, so this right here, I'm dragging and dropping in a video file.
And what this process is going to do, and I'll explain it really quick, is it's going to transcribe the video. and get a like word level transcript of it. And then it's going to pass that entire thing into Jev with some different classifier decisions to find the best clips.
And we'll get to see how quickly it works. Paste it in. It prepares audio and scores 17 moments and around like three seconds.
And each of these moments are like one of the interesting parts of the video. They're not like the filler text where I'm like, so I'm going to set this up. It's like.
Let's go ahead and watch that. It's flying. It's absolutely flying.
We've got 1 .1 million tokens, yada, yada, yada. So it allowed me in this demo to be able to find the best clips that I could publish on short form content. So honestly, and I worked on this for maybe 10 minutes.
So if you worked on this and iterated on this to create your own startup with this type of idea, you could probably get pretty far, especially if you combined it with other different AI agent types. So that way you could have a really good clipping sort of feel on it. But there's so many different ideas that you could come up with this.
And honestly, the best way that I've thought about it is if you just think about it for like a night in the morning, you'll be buzzing with ideas like, oh, I could do this. I could do this. I don't know if I already showed this one, but the browser use for browser control with Jev is pretty insane, too.
I'm going to play this clip right here. This is in real time done by the browser control guys. or the browser use guys, where this is Jeff controlling this browser to pick a flight from Zurich to London in 7 .1 seconds.
Let's watch it.
This is all real time, by the way. So selecting the dates and it found a flight in 7 .1 seconds. If you asked any other sort of like browser use AI agent, this probably would have taken a minute, two minutes, even three minutes in the same type of regard.
Yeah, that's a big deal. That's a really big deal. If people want to get set up with Jeff, how do they do it?
So Jeff right now is on a wait list, but by the time this video drops, it might be out in general accessibility. But if you want instant access to it, you can go to the Vercel gateway. And they have Jev available on it right away.
So you can just instantly start testing it out. They've added some stuff into their AI package so you can start messing around with it. But honestly, if you ask your AI agent and drop it this link and the type safe AI to say, hey, how can I start experimenting with Jev?
You can probably get started right away. And that's a great way to get started to any type of AI agent. You could talk to it about your business and with Jev and say, hey.
What sort of workflows do I do on the daily basis that could benefit from a decision maker like Jeff? That's a huge tip. I appreciate that.
I'll include the link in the show notes, in the description, where you can go in and play around with this. I'll also include links where you can follow Ryan. He's got a criminally underfollowed YouTube channel.
I think it's like a thousand subs. I know. It's crazy.
So I'll include that as well. Ryan, thank you so much for coming on. You know what I'm doing after this?
I'm going to this Vercel link. I'm going to play with Jev. I'm going to start classifying some stuff.
Let me caution you, though. It is dangerously addictive. Once you see the speed and once you see the price, you will just be like, holy cow.
And to all of you guys watching at home or listening, please just try it out. It's so cheap. You won't even notice.
It'll be like one one -thousandth of a cent type of thing to test it out. It is so cheap. please test it out.
This is a new type of AI. If you've ever done any sort of classification or if you just want to build something for your own email or other system, try it out. It's so fun to use and the experience with it, it's just going to be mind blowing because I don't think we've seen AI this fast in a long time.
All right. Can't wait to play with it. Thanks everyone for your time.
Ryan, you're a legend and I'll see you next time. See ya.
The Hook
The bait, then the rug-pull.
Greg Isenberg opens by calling Jev a genuinely new category of AI, then turns the mic over to Ryan Vogel to prove it live against his own inbox.
Named ideas worth stealing.
03:16model
Classifier IO Model (Input, Schema, Output)
- Input: any object
- Schema: the defined output options
- Output: a probability per option, summing to 100%
Jev's whole architecture in three steps: hand it an object, hand it a schema of possible answers, and it returns a probability for each one.
Steal forany place you'd otherwise write an if/else rule or manually tag incoming data
12:02concept
Decision Model vs. Chat Model
- You send: a question in words vs. an input plus a schema
- You get: streamed text vs. numbers and categories
- Reasoning: thinks out loud vs. none shown
- Speed: up to ~30s streamed vs. ~200ms, one call
- In your code: parse the text first vs. type-safe, use as-is
A side-by-side comparison of how ChatGPT or Claude behave versus how Jev behaves on the same five dimensions.
Steal fordeciding whether a workflow actually needs a chat model or just a classifier
17:21model
The AI Traffic Cop
- Incoming: form leads, support tickets, emails
- Jev asks: what is it? how important? what next?
- High confidence: a human, now
- Middle confidence: automate it or an LLM drafts a reply
- Low confidence: ignore it
A routing framework where Jev triages incoming information into three lanes by confidence score, staying in an advisory role.
Steal forany inbound queue: support desk, lead intake, contact forms
21:39concept
Put Jev at the Front of the Queue
- Find a business with an expensive incoming queue
- Put a Jev-based classifier at the front of it
- Return an instant match or quote instead of a delayed one
The startup-idea framework Greg pushes Ryan toward: any business with a slow manual intake process is a candidate for a Jev-based instant-matching layer.
Steal forlocal services marketplaces, quote-based service businesses, lead routing tools
How they asked for the click.
VERBAL ASK
26:16link
“if you want instant access to it, you can go to the Vercel gateway”
Ryan names the exact access path, the Vercel AI Gateway, paired with a waitlist for direct access, and Greg closes by committing on camera to go try it himself right after the recording.
MENTIONED ON CAMERA
Visual structure at a glance.
