Jev 1.13: 6 Demos From Startup Scoring to Doom

What Are People Building With Jev 1.13?

The demos people shared around Jev range from practical work tools to games that make its decision process easy to follow. You can watch it break down competitor advertising, plan an interface from a natural-language request, score a startup idea, compare structured decisions, play Doom, and navigate toward a target page on Wikipedia.

The examples with the widest visible reach focus on work. Matthew Berman's competitor ad analysis leads with 1,044,431 views on X, followed by michaeltefula's interface-planning demo with 49,900. Together, they offer a useful starting point for anyone searching for concrete Jev projects rather than a general model description. The other demos widen the picture by applying structured choices to evaluation, gameplay, and link navigation.

How Can Jev Analyze Competitor Ads?

“Break down competitor ads” shows Matthew Berman using Jev to examine advertising through four specific lenses: hooks, offers, formats, and the fit between an ad and its landing page. That makes the demo immediately understandable. Instead of producing a broad reaction to an advertisement, the analysis is organized around parts that someone can inspect separately.

The hook asks how the ad attracts attention. The offer focuses the review on what is being presented to the audience. Format separates the way the message is packaged from the message itself. Landing-page fit connects the advertisement to the page people reach afterward. Those categories turn competitor research into a structured exercise, which is why this example is useful for marketers and builders who want to see Jev applied to recognizable work.

Berman, who shared the demo as @TheMattBerman, gives the model a task grounded in choices people already make when reviewing campaigns. The interest lies in seeing those choices separated cleanly enough to compare. It is the strongest first example because it pairs a practical use with the largest stated audience among these demos.

Can Jev Plan An Interface From A Request?

“Build an interface from a request” moves from analysis into product planning. In the demo shared by @michaeltefula, Jev takes a natural-language request and turns it into a plan for an interface. The request provides the starting point, while the resulting plan gives the proposed interface a more explicit structure.

This is interesting for anyone who has an idea for a tool but thinks about it in ordinary language rather than interface terminology. The demo shows a path from describing what someone wants to organizing what should be built. That makes it relevant to early design conversations, feature planning, and creator workflows where a written request needs to become something more concrete.

The focus is the plan itself. That keeps the example easy to judge: a viewer can compare the original request with the proposed interface plan and decide whether the structure reflects the intended experience. Among the shared work demos, it offers the most direct look at Jev translating an open request into an organized build direction.

Can Jev Evaluate A Startup Idea?

“Is this startup stupid?” turns evaluation into a playful, focused tool. The demo from @gajananrx asks someone to enter a startup idea, then returns scores for novelty, money potential, buildability, and appeal. Each category answers a different question about the same proposal.

Novelty considers whether the idea feels distinct. Money potential gives the commercial side its own place. Buildability brings the discussion back to whether the idea can be made, while appeal focuses on whether people might want it. Presenting those dimensions separately makes the result easier to read than a single overall reaction.

The demo is especially useful as an example of product design around structured judgment. A broad question about a startup becomes four named criteria that can be viewed together. The playful title makes the experience approachable, but the underlying format is practical: enter an idea, receive a compact assessment, and see which dimension shapes the result.

How Does Jev Handle Structured Decisions?

“Compare structured decisions” places Jev and a language model beside the same structured request. The value of the demo comes from the shared task. Because both systems work through the same request, viewers have a common frame for examining how each one approaches it.

This example is less about a specialized application and more about making a decision process visible. A structured request gives the comparison defined parts, so the viewer can follow how each response addresses the same prompt. That makes it a helpful companion to the ad and startup demos, where Jev is also used to organize judgments across named considerations.

For someone searching specifically for what Jev does differently in a decision-oriented setting, this is the most direct demonstration. It does not require a marketing campaign, product concept, or game environment to make the comparison understandable. The request and the two resulting approaches provide the complete setup.

Can Jev Play Doom By Reading Game State?

“Jev plays Doom” shows an AI reacting to game state and following instructions inside Doom. The setting makes each decision visible through what happens next on screen. Instead of reading an analysis after the fact, viewers can watch the relationship between the current situation, an instruction, and the resulting action.

Doom is useful for this kind of demonstration because the state keeps changing. Each action leads into another situation that calls for a new response. The demo therefore presents Jev in a continuing sequence of choices rather than as a single answer to a single request.

This is the clearest game example among the shared projects. It gives people an intuitive way to observe instruction following without requiring them to understand an interface plan or a set of business criteria. The interest comes from watching decisions unfold in an environment where progress is visual and immediate.

Can Jev Navigate A Wikipedia Link Race?

“A race through Wikipedia” gives Jev a target page and shows it choosing links to reach that destination. Every move is constrained by the links available on the current page, so the route develops one choice at a time.

The demo turns navigation into a readable sequence. Viewers can see the page Jev is on, the link it selects, and how that selection changes the path toward the target. Unlike the Doom demo, the environment consists of connected information pages rather than game state, but both examples make repeated decisions visible through their consequences.

For searchers interested in Jev projects beyond work tools, the Wikipedia race is a concise example of goal-directed link selection. It also complements the structured-decision comparison: one presents a shared written request, while the other presents a changing route with a defined destination. Together with the work and game demos, it shows why people are experimenting with Jev across several kinds of decisions.

Which Jev Demo Should You Watch First?

Start with “Break down competitor ads” for the most viewed practical example, then watch “Build an interface from a request” to see a written idea become an interface plan. “Is this startup stupid?” is the best next stop for a compact scoring experience built around clear criteria.

Choose “Compare structured decisions” when you want a direct side-by-side request, “Jev plays Doom” when you want decisions shown through gameplay, and “A race through Wikipedia” when link-by-link navigation is more interesting. These early demos cannot show everything Jev can do, but they give searchers several concrete projects to examine.