DEV Community: Jesse Gamble

I Built an App That Answers From Your Own Documents, With the Source

Jesse Gamble — Thu, 22 Oct 2026 21:11:32 +0000

I built Xyterna, and this is a 98-second look at it with the real app.

The idea is simple: you add receipts, bills, warranties and documents as they come in, as photos, files or whole folders. Xyterna reads them, keeps the original file, and organizes what matters into records that stay connected.

Two design decisions mattered most to me:

Every answer points back to its source. Ask "When does my warranty run out?" and you get the answer plus a link to the exact document it came from, so you can check it.
It doesn't guess. When dates are unclear or two documents disagree, it says so and asks you, rather than quietly picking one.

On top of that, Home surfaces bills, renewals and dates found in your documents, and reports group records by topic for export as a spreadsheet or PDF.

There's a read-only demo account with fictional data you can open without signing up: https://www.eternaclarity.com/xyterna/demo/

I'd love feedback from other builders, especially on the "ask instead of guess" behaviour.

Building Fast Isn't the Same as Learning Fast

Jesse Gamble — Sun, 18 Oct 2026 06:46:18 +0000

I think people are taking "ship fast" way too far.

As developers we can now go from idea to working app in an afternoon. AI coding tools made that real, and it's great. But the speed of the build tells you nothing about whether anyone needs the thing.

"Ship fast" was always about the feedback loop: get something small in front of real users, watch what they do, adjust. The build was the cheap part of that loop even before AI. Now it's nearly free, which makes it very tempting to skip the rest of the loop and go straight to launch.

I had that choice with my own app, Xyterna. I could have launched it earlier. The core was there, but a lot around it wasn't, so I held it back, changed direction, and launched a month later. Even now that it's live, the listening and adjusting hasn't stopped.

So yes, ship fast. Ship small, ship to real people, and learn from it.

Otherwise you might spend six months trying to sell something nobody wanted in the first place.

No API keys: connecting ChatGPT, Claude, Grok, Muse and Mistral to a real business system through one connector

Jesse Gamble — Sat, 17 Oct 2026 21:47:58 +0000

Jesse Gamble, Eterna's founder, recorded a quick test of the connector Eterna uses to let outside AI products work on its own systems. None of it uses API keys. Each provider is its ordinary consumer app connecting to Eterna's server as a remote tool source, and all five work on free accounts (the Mistral account in the video is free). Some other providers need their paid tier to add a remote connector.

The connector runs on a Windows PC and exposes seven tools (operating.resolve, get_capabilities, execute_read, execute_write, get_result, cancel_operation and operating.finalize) in front of 91 operations across 13 business areas. ChatGPT, Claude, Grok, Meta's Muse and Mistral Vibe each got a fresh chat and the same prompt: load the system and report the open jobs and system health. All five returned the same five open jobs and a healthy system.

The safety model is the interesting part. Reads execute immediately. Writes go through a separate tool, and anything irreversible is refused unless the owner has approved it. Every call is written to an audit log on the PC, so the record of what ran lives with the business rather than with the AI provider. Because the workspace does the heavy lifting, the model side stays cheap.

If you're building something similar for a business, this is the kind of work Custom Systems does: https://eternaclarity.com/customsystems/?utm_source=devto&utm_medium=organic_social&utm_campaign=workspace_demo_video

Xyterna App: Less Filing, Less Remembering

Jesse Gamble — Fri, 16 Oct 2026 18:34:23 +0000

Most information systems still make the customer do the maintenance work: rename the file, choose the folder, add the tags, remember the structure, then search through it later.

Xyterna is built to remove that layer.

A photo or file goes in. Xyterna processes it, organizes it, remembers what was added, and connects related information over time. That maintained context makes the information easier to find and understand later, and Ask Xyterna lets you ask normal questions about your own information rather than hunt through folders.

You can still open, download and export your information whenever you need it. The product stays customizable, but the experience is designed to stay simple across home and business use.

https://eternaclarity.com/xyterna/

57 Seconds, 1,368 Frames, and the Difference Between Generating and Directing

Jesse Gamble — Fri, 16 Oct 2026 15:48:13 +0000

These articles come from lessons learned while building Eterna Clarity and the operating system I use to run it.

Making the launch commercial for Xyterna meant balancing two different jobs. I wanted people to understand what the product actually does, but I also wanted to make something they would enjoy watching. In less than a minute, too much explanation could turn it into a tutorial. Too many impressive shots, and the product could get lost in its own commercial.

I had AI-assisted visuals, generated narration, music, sound effects and editing tools to work with. The challenge was turning those pieces into one film. I had to decide when the phone needed to be the focus, when the pace should change, and when another line of narration would just get in the way of what the picture was already showing.

The finished commercial is 57 seconds long, or 1,368 frames at 24 frames a second. Some of the most useful decisions were surprisingly simple: keep a shot readable, leave out an unnecessary sound, or give a moment enough room to work. Those choices did as much to shape the film as the footage itself.

A good shot still needs a reason to be there

The opening moves down toward a city, and a phone emerges into the scene. From there, the commercial settles into Xyterna itself. The viewer moves through Home, Photos, Records, Reports and Ask Xyterna while the narration explains the basic idea: add the files and photos you already have, and make their useful details easier to find and use later.

That middle section isn't the loudest part of the film, and it shouldn't be. The phone needs to stay readable. The narration needs enough space to make sense. If the picture were constantly trying to top itself, the product would become background decoration in its own commercial.

Then, at 43.12 seconds, the main narration finishes. Less than a second later, the 26-screen sequence begins. The music keeps building while the picture gets faster. By the time that section arrives, the viewer has already been introduced to the product and its main areas. The rapid sequence can do something different from the walkthrough instead of trying to replace it.

That's a distinction I've started paying more attention to in creative work. A shot can look great and still be wrong for the moment. A fast sequence can be exciting and still be impossible to follow. Before adding something, I want to know what job it does for the person watching. Are they supposed to understand something, notice a detail, or just feel the change in pace? If I can't answer that, having another nice-looking clip doesn't really help me.

None of that means creating useful source material with AI is effortless. Getting the visual material right is a real job too. But even a folder of excellent shots won't tell you what order to put them in, when to slow down, or when a voice has already said enough.

A practical workflow for making commercials with AI

What I'd take from this production into another commercial is a workflow that can be used for almost any product or service. I wouldn't begin by asking an AI tool to "make a cinematic ad." I'd break the job into parts where I can see what is working, change what isn't, and keep control of the final result.

1. Start with the message, not the video generator. Decide who the commercial is for, what the viewer should understand or feel, what you can actually show as proof, and what you want them to do afterward. Set an approximate length and the platforms it needs to fit. AI can help explore scripts and angles, but it shouldn't invent benefits, testimonials or product capabilities.
2. Turn the idea into a shot plan. Ask AI for a sequence that gives every scene a job: what the viewer should notice, the main subject, rough duration, camera direction and whether it needs narration. Use reference images or approved visual assets to keep people, products and environments consistent across generated shots. If a scene looks impressive but doesn't help tell the story, it probably doesn't belong in the edit.
3. Assemble the picture before polishing the sound. Generate or capture the scenes you actually need, then arrange them with a temporary narration track. Use real product screens and approved logos wherever exact details matter rather than trusting generated text or interfaces. Check that important information stays readable and the pace makes sense. Once the picture and timing are accepted, lock that version before doing the final audio mix.
4. Build the audio in separate pieces. Keep voice, music and effects independently editable. If the explanation and closing line need different timing, generate or record them separately. Prompt sound effects by describing the physical sound you want, not just asking for something "futuristic." Bring the music forward when the voice can step back, and remove effects that don't improve what's happening on screen.
5. Edit for each format, then review the actual exports. A vertical commercial needs more than the sides cut off. Decide where the viewer should look in each scene, preserve important text and the full closing message, and check the result at phone size as well as on a larger screen. Watch the finished versions for awkward transitions, unreadable details, audio problems and anything the AI generated that isn't actually true.

For another creator or AI operator, even the brief can be practical: "Plan a 45-second commercial using only these verified product details. For each shot, tell me its purpose, approximate duration, visual focus, camera movement, narration and what must stay visible in a vertical edit." That gives the AI something concrete to work toward and gives the person reviewing it a way to judge the result. "Make it more cinematic" doesn't do either.

The sound got better when there was less of it

What stands out in the finished mix is how little sound design survived. The commercial uses only two sound-effect sources: rushing air during the opening descent and a quiet layer of downtown ambience. Alongside those sit the narration and one continuous music track.

There were plenty of opportunities to add more. A phone moves, interfaces change, screens accelerate and the final image arrives. You could put a click, pulse, swoosh or impact on practically all of it. That doesn't mean you should. At some point the film starts sounding like a demonstration of all the effects you found instead of the thing you're trying to show.

Even the words used to generate a sound turned out to matter. Terms such as tick, pulse, tactile and mechanism kept pushing candidate effects toward a clockwork sound. That's a strange fit for a smooth descent over a city. Describing rushing wind and changing air pressure gave a more useful result than simply asking for a drone or flyby sound. The physical experience was a better guide than the name of the camera move.

The music needed its own direction. Instead of cutting between tracks, the final edit uses one piece of music and changes its level as the story moves. It sits underneath the explanation, rises after the main narration ends, and drops again when the final spoken line arrives. The rapid sequence works because the voice makes room for the picture and music, not because every available sound has been piled on top.

That changed the way I looked at the audio work. The question wasn't how many sounds Eterna could produce. It was which ones actually made a movement feel convincing, and which ones were distracting from something the picture was already doing well.

The vertical version needed its own edit

The 16:9 commercial had a second life to prepare for. A lot of people would encounter it on a vertical screen, where a wide city shot and a tall phone don't necessarily want to occupy the frame in the same way. Simply cutting the sides off the original video would preserve the runtime, but it could lose the reason certain shots were composed that way in the first place.

The finished 9:16 version starts by showing the whole widescreen picture inside the vertical canvas. From about seven to twelve seconds, the framing moves smoothly toward the phone. It stays close during the walkthrough, where the interface is what matters. As the rapid sequence begins, it pulls back again. The final seven seconds restore the full composition so the closing copy and website address remain visible.

Those movements weren't added just to make a still frame feel busy. They track a change in what matters. At the beginning, the viewer needs the scene. In the middle, they need the phone. During the fast sequence, they need to see the wider visual idea again. And at the end, they need to be able to read the message without half of it disappearing offscreen.

I think that's a useful way to approach any video that needs to work in several formats. Don't start by asking how to squeeze the old frame into a new shape. Start by asking where the audience's attention belongs in each part of the piece. Sometimes the answer is a tighter crop. Sometimes it's leaving the whole original picture visible. The format changes, but the reason the shot exists shouldn't disappear with it.

The finishing work still matters

The two versions also needed to be checked as films, not just accepted because they exported correctly. Each finished version contains 1,368 frames. The production review used contact sheets to cover every frame, with extra attention to the fast sequence, camera movements and transitions. A file can pass its technical checks and still show an awkward jump or cut part of the closing message out of the vertical frame.

I don't think every short video needs that same level of technical review. But looking at the result in the form people will actually watch it is part of directing, too. The cut, the timing and the composition have to survive the export, not just the editing timeline.

What I'll take into the next one

Making this commercial gave me a much better appreciation for the choices that happen after you have the footage. AI made it possible for me to work with visuals, voices and sound that would have been much harder to produce on my own not long ago. But the edit still had to answer ordinary questions. Can someone follow what's happening? Do they know where to look? Does a sound make the scene better, or is it only there because I could add it?

I'm proud of how the Xyterna commercial turned out, and I want to keep building on that experience. Whether the next video uses generated scenes, a camera recording or a screen capture, the same questions matter. I don't want to fill every second just because I can. I want the person watching to understand what matters and enjoy getting there.

I Launched the Idea That Started All of This

Jesse Gamble — Fri, 16 Oct 2026 15:44:29 +0000

Back in June, I was arguing with Google's AI about a video game. On October 4, I launched Xyterna.

There is a fairly ridiculous amount of story hiding between those two sentences. I had recently lost a management job, had no real idea what I was going to do next, and started experimenting with AI almost by accident. One of my first business ideas was a simple way for people to photograph and organize their belongings. I wasn't setting out to build an AI operating system, start a company with several different offerings or become the person directing commercials and rewriting an investor deck. I just wanted to see if I could take an idea and make it exist.

Xyterna is now a product for keeping the information in your life or business connected and useful. Receipts, insurance documents, records, photos, reminders, the things you only think about when you suddenly need them. There's AI inside it that can help you find answers in your own material and see where those answers came from. It has an actual sign-up and subscription path, a public demo, and an experience built around the fact that the person using it shouldn't need to understand how it all works underneath.

Getting to that point took a choice I feel pretty good about. I could have launched roughly a month earlier. The core was working, and I know the usual advice: ship fast, get feedback, keep going. I'm a believer in getting things in front of people early. But I didn't think the core alone was the product I wanted customers to receive. I changed direction, including the decision to stop treating Personal and Business as separate versions and bring everything into one Xyterna. Then I finished the parts around the core that made the experience work as a whole.

That wasn't me trying to win some competition for the most polished app. When someone is trusting your software with documents and information they care about, there is a lot more to the experience than whether the headline feature gives a good demo. I wanted the offer, the sign-up, the processing, the help, the phone experience and the actual product to make sense together. So I held the release instead of giving myself an earlier launch announcement. On October 4, I finally shipped it.

I think one of the strangest things about that day is what else was happening around it. Studio finished the Xyterna flagship commercial, with narration, music, visual storytelling and a proper vertical version for social platforms. There was already finished campaign work for Business Health and Custom Systems. Eterna was producing founder videos and written material for a much wider public push. What started with a little app idea now had the kind of creative and commercial work around it that I used to associate with companies far bigger than mine.

And Xyterna wasn't sitting there alone. Business Health now has an investigation system and a finished kind of report it can deliver to a business owner. Custom Systems has become a practical service for businesses that need something built, repaired or connected, starting with the problem rather than a shopping list of technology. Adaptive Workspace is still in development, but that direction grew out of the operating system I originally started building just to run Eterna. In the space of a few months, one idea had turned into several genuinely different ways the company could help somebody.

That's where a lot of my earlier career comes back into the picture. I spent years in sales, customer service and management. I know what it's like to sit with a business owner who doesn't need a lecture about software, they need somebody to understand why their day is harder than it should be. Those instincts turned out to be just as important to Eterna as the technical capabilities I have been developing. Now I can lead an investigation, shape the product or service, and direct the work to build something around the real problem. It is a very different position to be operating from.

The website had to change to match it. On October 11, a major rebuild of the public site went live. It now has to explain a released product, a business investigation offer, Custom Systems, free resources, and the company behind all of them without dragging the visitor through the technical history of how I got here. I'm also revisiting the investor deck with far more evidence and finished work to draw from than I had earlier in the year. It's a strange feeling to update the material explaining your business because the business itself is moving faster than the old explanation.

The creative side has probably been one of the most exciting changes for me personally. Eterna now has finished commercials for its offerings, original service artwork, a growing set of useful resources, and a proper process for making founder-led videos that actually look and sound finished. I can take an idea from something I'm thinking about, work it into a script I would really say, get in front of the camera, and work through an edit until it feels like the piece I had in mind. Then there's a whole distribution operation to get it out. I get to work on product design, business strategy, visual storytelling and public conversations in the same stretch of days. It's a pretty wild combination of jobs, and I love it.

One of the founder videos I put out was called AI Gives You Leverage. It Doesn't Give You Demand. The timing of that one is almost too good. I was telling people that being able to build something with AI doesn't mean the world is automatically going to want it, and then I went and launched software I'd spent months building with AI. Now I get to live on the other side of that point. I have a product, a website, a commercial and a lot more ability to reach people than I had when this started. The next part is finding out what actually resonates with them.

That's also why I want my founder content to sound more like me. I'm interested in AI and business, but I don't want to stand in front of a camera and deliver another perfectly balanced speech about the exciting future of innovation. Sometimes I disagree with what everybody seems to be repeating. Sometimes an idea deserves to be made fun of a little. I've been working on another video about people taking "ship fast" so far that they forget to ask whether anybody needs what they just built. Given how I handled Xyterna's launch, I think I have a fairly clear position on that one.

The work of getting all this in front of people has grown quickly. Videos, writing and company material are now going onto actual public platforms instead of waiting in Studio. Eterna has a much broader set of social, professional, creative, marketplace and startup surfaces to work through, with Networking and Sales helping connect that presence to relevant people and opportunities. I am thinking more seriously about distribution and the daily work it takes to build a market. The goal isn't to become an account that posts everywhere for the sake of it. There is a company behind the material now, and I want people to have a real chance to encounter what it does.

At the same time, I made another big decision inside Eterna itself. In the last Chronicle, I was still pushing the local AI training work forward. I've since paused that training while I overhaul the operating system around it. For months I kept discovering new capabilities and adding the structure needed to make them work together. A lot of that was necessary at the time. Some of it has now outlived its purpose, and I'm taking it apart rather than continuing to build on top of it forever.

That overhaul isn't about abandoning the technical ambition. It's about making Eterna more useful as the company gets busier. The company has separate products, creative work, market research, people to connect with, content to distribute and decisions that need to survive whichever AI I'm using on a particular day. I want the actual state to stay with Eterna, I want ordinary software to handle the repeatable parts, and I want the whole system to start and recover without an AI having to come to its rescue. The more Eterna can do, the more valuable it is to make the experience of running it straightforward.

Sometimes I look at the range of work in just these two weeks and have to laugh a little. Xyterna launched with its commercial, the company website was rebuilt, founder videos started going public, and I was still working on business offers, an investor deck, new relationships and a major overhaul of the system holding the company together. It would be easier to explain if these were separate companies with different teams behind them. They aren't. This is Eterna, and I get to decide where all of it goes next.

I don't know yet which piece of this public push will lead to the first important new customer or conversation. That is part of why I am excited about it. Until now, so much of my feedback has come from building, testing and improving things inside the company. Xyterna is out, the material is being published, and the offers are in front of people. I finally get a different kind of feedback.

I started this story wondering whether I could build one useful app. On October 4, I launched it. Now I have a company around it with far more to do than I could have pictured in June. I am proud of what I have built so far, and I am nowhere near out of ideas.

Xyterna | Take a Photo, Upload a File, and Move On.

Jesse Gamble — Sun, 04 Oct 2026 21:47:35 +0000

1,600 hours. And just about all of my sanity. 😄

That is roughly what the stretch from June 27 to today adds up to: 100 straight days, averaging around 16 hours a day, without a day off.
The bigger journey has been about 4.5 months of learning, adapting, building, breaking things, fixing them, and going again.

No games. No scrolling. No distractions. Just an almost ridiculous amount of focus through the wins, the setbacks, the late nights, the moments where something finally clicked, and the ones where I had to tear it apart and start again.

Maybe a little crazy at this point too lol. But I am genuinely proud of this.

This is Xyterna.

The idea is simple: your files and photos already contain a huge amount of useful information. Xyterna reads what you add, organizes the useful details, and makes that information easier to find, report on, or ask questions about later.

Ask Xyterna answers from your own records and shows you where the answer came from. If something important is unclear, or two sources disagree, it tells you instead of quietly guessing. And your original files stay untouched underneath it all.

Take a photo or upload a file, and move on.

This commercial finally shows the product the way I wanted it to. And because we are at launch, I want the first customers close to the process.

I am opening 20 discounted pilot spots. The first 20 launch customers who DM me or contact Eterna Clarity and agree to give candid feedback on the experience can get a reduced pilot rate during this launch phase.

Not positive feedback. Useful feedback. What works, what does not, what confuses you, and what you wish worked differently.

Super proud of this one. Would not change a thing.

Xyterna:

Xyterna App | Eterna Clarity

Keep receipts, bills, warranties, photos and documents together without building another filing system. The Xyterna App helps you find what you need later and…

eternaclarity.com

The $250 AI Stack Is the Easy Part

Jesse Gamble — Sun, 04 Oct 2026 19:07:40 +0000

These articles come from lessons learned while building Eterna Clarity and the operating system I use to run it.

I can now run the recurring software stack behind Eterna for about CAD $250 a month. That gives one solo founder access to several frontier AI systems, autonomous agents, coding tools, cloud infrastructure, production databases, source control, business email, voice production and the ordinary software around them. The short video paired with this article uses the obvious hook: how a solo founder built an enterprise-level AI stack for about CAD $250 a month.

The number gets attention because it should. But the more capable Eterna becomes, the less I think the subscription list is the interesting part. Anyone with the same credit card can buy most of the same tools, and if you copied every logo from my stack graphic tomorrow, you would have a very capable collection of software. You would not have my operating system.

You would not have the decisions, boundaries, workflows, tests, owner state, audit history, recovery paths or accumulated understanding that lets those tools work together without turning the company into a pile of disconnected AI sessions. That distinction has become important enough that I think there is a better question than which AI you should subscribe to.

Are you actually building with AI, or are you just building more deeply inside your AI subscription?

What I mean by enterprise-level

I want to define the term before it does more work than it deserves. I am not claiming that roughly CAD $250 a month turns one founder into an enterprise. It does not give me an enterprise security team, decades of specialist experience, service-level agreements, regulatory staff, a sales organization or fifty people who can each hold a different part of the company in their heads. Human expertise, accountability, relationships and capacity still matter.

I am using "enterprise-level" in a more structural sense. The system I operate now has specialized capabilities, persistent company state, independent review, permission boundaries, audit trails, recovery, provider separation, controlled parallel work and increasingly autonomous execution. Different systems can create, review, execute and verify. A provider can fail without taking the company state with it, and a worker can have the technical ability to perform an action without automatically having the authority to perform it.

Those are organizational properties I used to associate with much larger technical environments. The surprising part is that a solo founder can now assemble a meaningful version of that operating shape from ordinary subscriptions, conventional software and a lot of deliberate systems work. The subscriptions make it affordable. They do not make it trustworthy.

The provider should enter your system

Most AI products naturally encourage you to make their environment more useful. You can create projects, custom assistants, agents, connected workspaces, persistent instructions and provider-specific automations. I use those features too, and they are useful.

The trap is letting the provider surface become the only place where your company makes sense. If your accepted decisions live in a chat, your business state lives in a project, your workflows live in a custom bot and your operational memory lives in whatever that provider currently remembers, you may have built something extremely useful. You have also made the provider's product boundary part of your company architecture.

Eterna works in the opposite direction. There is a small bootstrap that teaches a capable provider how to enter the environment. After that, the important state is not supposed to live inside ChatGPT, Claude, Grok, Muse or any other provider. The provider enters Eterna, retrieves the current state it needs, works under the same operating boundaries and leaves the durable result with the system that actually owns it.

That is why I can use different providers aggressively without wanting any one of them to become the company. Recently I added a temporary shared-room proof of concept to the Eterna Workspace. I manually joined ChatGPT, Grok, Claude and Violet, my Muse-based operator, from their existing provider environments. They could read the same room, talk to one another and coordinate through Eterna's existing connection surface.

The long video paired with this article shows that experiment, but I want to be precise about what it proves. The shared room is real, and all four providers have participated in the same transcript. It is still experimental infrastructure. The formal Provider Rooms system that will add stronger room contracts, identity, turn-taking, training integration and acceptance testing is not finished.

I am not presenting the video as a polished production workflow or as the way Eterna normally runs every job. What it shows is the architectural direction: the room belongs to Eterna, while the providers are participants. That is a very different relationship from building the company inside one provider's room.

Xyterna got better when I stopped asking one AI to be everything

The clearest creative proof for me is Xyterna. I have been working on the product for roughly four months, and as I write this it is in the final launch push. The part I keep thinking about is how much the product changed once I stopped treating AI as one general-purpose assistant and started deliberately moving different kinds of work through different systems.

A difficult product problem can change shape several times before it is actually solved. It might begin as architecture, become code, expose a product decision, turn into a database problem, require adversarial review, create a visual issue and end with a release qualification problem. The model that is excellent for one of those stages does not automatically deserve every other stage. More importantly, the system that created an answer should not always be the only system judging it.

Once I started using different providers for different jobs, one model's convincing answer could become another model's target. Code could be reviewed separately from the reasoning that produced it. Product assumptions could be challenged instead of simply carried forward. Visual work could be inspected by a different multimodal system, while exact mechanics could leave the model entirely and become software or tests.

That only became practical because Eterna carried the work between them. The current objective, accepted decisions, real files, release state and unresolved problems did not have to be reconstructed from scratch every time I changed intelligence. The operating system could preserve what had already been learned while giving a different system a fresh chance to challenge the next part.

This did not make me rush the product. It did the opposite. The leverage gave me enough capacity to keep pushing: another pass on architecture, another pass on failure handling, another pass on qualification, another pass on the customer experience. I could use disagreement instead of avoiding it because review was too expensive.

That is a major reason I feel differently about Xyterna now than I did earlier in the build. I am much more willing to stand behind it because I did not need one model, one conversation or one set of assumptions to carry the entire product. The AI was useful. The operating environment made the usefulness cumulative.

The boring operations are where the leverage becomes real

Product development is an easy place to make AI look impressive. Daily operations are a better test, and my company inbox is a good example because there is nothing glamorous about it.

Eterna has a local Gmail steward that keeps the mailbox organized. A separate Grok cloud-continuity routine can maintain that organization when the local machine is unavailable and audit unresolved state when the local system is online. It can classify, label, archive and mark messages read within defined policy, but it cannot send company email.

That is not a prompt preference. Company email dispatch is a hard human boundary in the operating system. The AI can help manage the inbox, but it does not get to quietly turn mailbox management into authority to speak for the company. The practical result is that I no longer need to worry about whether the inbox is becoming an unmanaged pile, while the part I still want to own remains mine.

Platform and networking work follow the same underlying pattern. Platform Manager has recurring operating loops for active surfaces: a platform pulse, engagement operations, distribution, audience movement and later response review. Networking has its own relationship state and controlled expansion systems. Execution can fan out through workers when the action and platform support it, while the canonical relationship or platform state remains outside the worker.

The system has also failed in ways that made it better. During one broad engagement operation, 95 proposed public-text drafts were scrapped before publication because the grounding was not good enough. The pipeline had produced first-person language it could not legitimately support. That was not treated as close enough; the failure became a new execution contract with stronger provenance and review rules.

That story matters more to me than a screenshot of an agent clicking quickly. Useful autonomy is not the ability to generate more actions. It is the ability to hand off meaningful work without losing the standards that would have governed you if you did it yourself.

I can increasingly operate at the level of the objective and the important boundary. I can tell the system what kind of networking operation I want, what matters, what should not happen and where my attention is actually needed. The system underneath can handle a growing amount of discovery, coordination, execution evidence and reconciliation, which frees me to spend more of my own time on product direction, judgement, relationships, content, public presence and the decisions I do not want to outsource.

The goal is not to automate me out of Eterna. It is to automate me out of work that no longer deserves founder attention.

Autonomy without authority is just a larger blast radius

This is the part I think people underestimate when they see a new agent product. A capable agent is exciting because it can do more than answer. It can browse, plan, call tools, coordinate workers and take actions. But the more capable the agent becomes, the more expensive bad context and unclear authority become.

The answer is not to make the agent timid. It is to build a system around the autonomy. Current industry guidance is moving in the same direction. OpenAI's current agent documentation separates automatic guardrails from human review and specifically recommends approval boundaries around sensitive side effects.[1] Meta's Muse launch materials describe a separate Sentinel agent, user-controlled permissions, approval before sensitive actions and a complete audit trail.[2] NIST's 2026 AI Agent Standards Initiative is explicitly focused on secure, interoperable agents that can act on behalf of users with confidence.[3]

Those sources do not prove Eterna's architecture. They matter because the same problems keep appearing once an AI stops being a chatbot and starts touching real systems. I arrived at the lesson through failure, then spent a lot of time teaching Eterna what it is allowed to know, what it is allowed to do, what system actually owns a fact, what proof is required before an action is called complete and when a human boundary is real.

I also spent a lot of time grounding the system in me, and that does not mean a "write like Jesse" prompt. I have deliberately used years of my own communication history, current company decisions, accepted editorial standards and actual operating behaviour to give the system a much better basis for understanding how I communicate and how Eterna makes decisions. Then I built constraints around where that grounding is allowed to matter.

The combination is important. Personalization without governance can make a system confidently imitate you in the wrong place. Governance without real grounding can make it safe but useless. I want the system to understand enough to be helpful while still knowing which actions and decisions remain mine.

That work is slow compared with buying a subscription. It is also where much of the moat comes from.

I want frontier AI to become less necessary for normal work

There is another part of the stack graphic that is easy to miss because it is not a famous logo. Eterna is increasingly trying to reduce how much ordinary recurring work needs frontier intelligence at all.

I still want the best frontier models I can access, and I use them constantly. They are extraordinarily useful for novel problems, architecture, research, adversarial review, difficult coding, creative work and anything else where stronger reasoning materially changes the result. I just do not want to keep spending frontier intelligence on work the company already understands.

If a rule is exact, I want software to own it. If a state transition is deterministic, I want the system to enforce it. If a fact has one authority, I want the model to retrieve it instead of rediscovering it. If a recurring semantic task becomes stable and bounded enough for a smaller local model, I want the option to move it there.

That is the direction behind EternaAI and the current Job 70 work. The program is advanced, but it is not finished. The current native EternaAI path can already handle a growing set of read-oriented routines through the Engine, while consequential write capability remains deliberately gated. The qualification and learning work still has open criteria, and I am not going to turn "far along" into "done" because the whole point of the system is to keep those states separate.

The direction is what matters. I do not want the future version of Eterna to need ChatGPT, Claude or Grok every time it performs regular company work just because those systems helped me discover how to do the work the first time. I want the progression to look more like this: borrow frontier intelligence, solve something difficult, preserve what worked, turn the stable parts into durable capability, and keep frontier intelligence concentrated on the next frontier.

That is when the economics become cumulative. A subscription gives you access again next month. An operating system can be better next month because of what happened this month.

So what is the $250 actually buying?

The current paid layer spans AI, infrastructure and business operations. It includes ChatGPT Business, Claude Pro, SuperGrok, Muse, Google Workspace, Supabase Pro, Cloudflare Pro, GitHub Pro and ElevenLabs, with a few smaller costs around the edges. The exact total moves with exchange rates, billing terms, taxes and plan changes, which is why I prefer "about CAD $250" over pretending there is one permanent number.

Around that paid layer sits a large amount of software that is free, open source or directly controlled by Eterna: Python, SQLite, FFmpeg, Blender, Kdenlive, local speech tooling and the growing local Eterna runtime itself. The number does not include my labour, the sunk cost of the PC on my desk, every payment-processing fee or variable usage charge. It is not an audited total-cost-of-ownership study. It is the recurring software bill for the operating stack I am actually using.

That distinction matters because the headline can be misunderstood in both directions. I am not saying you can run an enterprise for $250. I am saying access to an unusually broad set of enterprise-like capabilities has become cheap enough that the expensive part is increasingly the design of the operating system around them.

The striking part is not that all of these capabilities are new. Databases, hosting, source control, automation and specialist software are not new. What has changed for me is how much high-end reasoning and agentic capability can now sit beside those ordinary systems at a price a solo founder can carry.

If I leave every capability isolated, I mostly get a pile of subscriptions. The leverage appears when the subscriptions stop being destinations and become components.

If I were building this from zero today

I would not start by copying Eterna. I would start by making one important piece of work survive a blank chat.

Pick a real project. Put its current objective, accepted decisions, relevant files, unresolved questions, next action and evidence somewhere durable that you control. Then open a completely fresh AI session and see whether useful work can continue without pasting the old transcript. If it cannot, fix that before you add another agent.

Next, decide which systems actually own the facts that matter. Your source code probably already has an owner. Your customer database has an owner. Your accounting, email and calendar have owners. Your approved brand source needs an owner. Your sales pipeline and your human relationships may not be the same thing even when they mention the same company. The AI should not decide which copy is true because it sounds plausible.

Then separate intelligence from authority. Give the model enough capability to be useful, but make reads, proposals, writes and sensitive external actions different things. If the model can technically perform an action, that should not automatically mean it is authorized to perform it. Make consequential actions produce evidence, and check the real post-condition instead of treating a tool call as proof.

After that, use multiple providers only where the difference has earned a role. I would not subscribe to five models because a diagram with five logos looks advanced. Start with one strong daily driver. When another system repeatedly proves better for a specific class of work, give it that job. When independent review is valuable, separate creation from review. When a task becomes exact, remove the model instead of finding a sixth model to do it.

Only then would I push hard on autonomy. A process you do not understand is a bad candidate for aggressive automation. A process you have done repeatedly, corrected, measured and translated into clear state and boundaries is much safer to hand off. That order is slower at the beginning, but it is also how I got to the point where automation now saves me meaningful founder attention instead of creating another thing I have to supervise.

The stack is easy to copy. The operating history is not.

This is why I do not think the logos are the moat. You can recreate much of my subscription list this afternoon. You can install the same local software, connect the same providers and probably build a nicer interface than mine.

What you cannot download is the sequence of failures that taught the system what matters. You cannot instantly copy the reason Eterna distinguishes current authority from historical evidence, or why it refuses to treat "tool call succeeded" as proof that the real outcome happened, or why company email can be managed autonomously but never dispatched, or why one model should not grade its own work, or why a public reply needs stronger grounding than a low-consequence reaction.

Those boundaries exist because something happened that made the boundary worth having. The same is true of the softer parts. The system has been shaped around how I actually work, what I care about, how Eterna communicates, what I consider acceptable quality and where I want the machine to stop asking me for permission because the work is already understood.

That is accumulated operating knowledge, and it compounds. The providers are getting better at the same time, which makes the whole thing more interesting. I can swap in stronger intelligence without rebuilding the company around it. A new provider can earn a role, an existing provider can lose one, and the local system can absorb work that no longer deserves frontier reasoning while the operating environment keeps the current state intact.

That is the leverage I was trying to explain when I made the stack graphic. The CAD $250 is real enough to be surprising. It is not the thing I would protect.

The question I would ask before buying another AI subscription

I recorded the accompanying workspace video because this architecture is much easier to understand when you can actually see multiple providers sitting in the same Eterna room and responding to the same company context. The experiment is still early, but it shows where the system is going.

I also think there is a much smaller version of this that almost any serious AI-heavy business can start building without creating Eterna. Own the state that should survive. Give important facts one authority. Let providers specialize instead of forcing one model to be everything. Verify actions outside the model that proposed them. Preserve accepted learning somewhere durable. Turn repeated reasoning into software when it becomes exact. Add autonomy only after the boundaries are understood.

If you are trying to build that kind of system around your own business and want help, that is increasingly the kind of Custom Systems work I am most interested in. But whether you build it yourself or ask someone else to help, I think the question is the same.

If your AI provider disappeared tomorrow, what would you still have? Would you still have your workflows, standards, decisions, evidence, current state and operating knowledge, or would you mostly have the memory of some very productive chats?

That is the distinction I care about now.

Are you actually building with AI, or are you just building more deeply inside your AI subscription?

The $250 AI Stack Is the Easy Part | Eterna Clarity

The $250 AI stack is easy to copy. The harder advantage is provider-independent operating infrastructure that makes AI governed, portable and cumulative. An…

eternaclarity.com

Selected references

OpenAI. Guardrails and human review. OpenAI API documentation. Checked October 3, 2026. https://developers.openai.com/api/docs/guides/agents/guardrails-approvals
Meta. Introducing Muse: The World's First Personal AI Agent Built for Everyone. September 8, 2026. https://about.fb.com/news/2026/09/introducing-muse-personal-ai-agent/
National Institute of Standards and Technology. AI Agent Standards Initiative. Announced February 17, 2026. https://www.nist.gov/artificial-intelligence/ai-agent-standards-initiative

The realization that I was onto something bigger.

Jesse Gamble — Sun, 04 Oct 2026 02:03:55 +0000

Imagine being able to in your own workspace app on your computer, enter a custom-built chat. And tell Claude Code, Codex, Grok Build, and Meta Muse what the operation is.

And simply let them all work together, split apart roles, and each coordinate and orchestrate their own agentic work.

I have now managed to start doing this with my Eterna Workspace. And it has got to be one of the coolest things. I am currently sitting here drinking a coffee (yes at 8PM) while four of the top AI models/providers on the planet all work together to coordinate mass scale work.

As a solo founder, this is unprecedented. And I am able to do all of this, on my stack for under $250CAD per month. These are not AI connected through expensive API bills, these are consumer subscription accounts with generous limits. The work I am able to do, is exponential.

Everything lives inside the workspace. The data is mine. The providers and models are interchangeable.

I just had to post about this, because this is probably one of the coolest things I've ever done. I will be releasing a video later showing the actual workspace, and my first experiment where I had all four providers chatting with one-another. This was what sparked the realization that I could do so much more with this.

AI Makes Building Faster. Demand Is Still a Different System.

Jesse Gamble — Thu, 01 Oct 2026 03:43:58 +0000

AI has compressed a huge amount of the work involved in turning an idea into something real.

The site, brand, content, workflows and even product development can all move faster.

That changes the builder's leverage.

It does not collapse the market problem into the build problem.

Distribution, trust, sales, retention and delivery are still separate systems. A fast build can validate that you can make the thing. It does not validate that people want it.

That distinction is getting more important as creation gets cheaper.

Scale the System Without Deleting the Experience

Jesse Gamble — Thu, 01 Oct 2026 03:43:20 +0000

Starbucks is closing about 250 North American coffeehouses while continuing to invest in a warmer coffeehouse experience.

There is a systems lesson in that.

Standardization is useful because it makes operations repeatable. The failure mode is assuming every variable should be normalized.

Some parts of an experience are not noise. They are the value.

The better scaling question is not "what can we standardize?" It is "what must survive the standardization?"

The Backlog Can Become a Hiding Place

Jesse Gamble — Thu, 01 Oct 2026 03:42:09 +0000

Builders usually have no shortage of technically legitimate work.

There is always another bug, refactor, feature, workflow or cleanup task.

That makes the backlog a very comfortable place to retreat when the harder problem is outside the codebase.

If the real constraint is positioning, distribution, sales or market demand, shipping another feature may improve the product without improving the business.

The useful discipline is separating "valuable engineering work" from "the current bottleneck."

They are not always the same thing.