AI and Layoffs: Automation, Augmentation, and the Future of Work

The conversation about AI and layoffs has become one of the most charged debates of the decade, and most of it is conducted in slogans. At ETAPX, we build AI products every day — across GLSRM, Ocsidian, and Whistlr — and we have a different, more grounded view of what AI does to work: it is not a machine that deletes jobs on a schedule, but a tool whose impact depends entirely on the intentions of the people who design it. This is our honest, clear-eyed position on AI and the future of work.

We are not neutral observers writing from a distance. We ship agentic systems. We watch the AI industry move in real time because tracking it is literally one of our products. We have made the build-versus-buy calls, the automate-versus-augment calls, and the harder call of deciding what a tool should refuse to do even when it technically could. So when the headlines announce that AI is "coming for your job," we want to answer the only way we know how: specifically, and from inside the work.

Why ETAPX Has Standing in the AI and Jobs Debate

A lot of commentary on AI and employment comes from people who do not build with the technology. They either oversell it — every task automated, every worker obsolete by next quarter — or they dismiss it as a passing hype cycle. Both positions are comfortable because they require no contact with the actual difficulty of making AI useful.

ETAPX lives in that difficulty. We build three products in one connected ecosystem, and AI runs through all of them in different ways. GLSRM is an intelligence platform for the AI industry itself — it tracks the launches, the labs, the leaderboards, and the research that matters, distilling a chaotic firehose into a clean, continuously updated surface. Ocsidian is an agentic game creation platform where AI-powered workflows help a single creator do work that used to require an entire studio. Whistlr is the social layer that connects people across the ecosystem with one account.

That gives us a vantage point most opinion pieces lack. We see how fast the frontier actually moves because we measure it. We see what agents can and cannot do because we ship them. And we see what happens to a creator's workflow when you hand them a capable AI collaborator — because that is the entire premise of Ocsidian. This article is built on that, not on speculation.

"AI does not have intentions. The companies that build it do. The layoffs conversation is really a conversation about what those companies choose to optimize for — and that is a choice, not a law of physics."

— ETAPX Product Team

The Two Stories We Tell About AI and Work

There are essentially two dominant narratives, and they are both incomplete.

The first is the replacement story. In this version, AI is a cost-cutting engine. Every capability gain is a headcount reduction waiting to happen. The protagonist is a spreadsheet, and the plot is subtraction. This story is seductive to a certain kind of executive because it is legible: fewer people, lower costs, same output, better margins. It treats labor as a line item and intelligence as a commodity to be bought wholesale.

The second is the augmentation story. Here, AI is a force multiplier. It removes drudgery, compresses the distance between idea and execution, and lets people operate at a higher level. The protagonist is a person who is now more capable, and the plot is addition. This story is popular with technologists because it flatters the technology and, conveniently, the people building it.

The truth is that both stories are real, and which one plays out is not determined by the AI. It is determined by the company deploying it. The same model can be used to fire half a team or to make the whole team dramatically more effective. The technology is genuinely neutral on this question. The org chart is not.

Automation and augmentation are not the same decision

It is worth being precise, because these words get used interchangeably and they are not the same thing. Automation means handing a task to a machine so a human never has to touch it again. Augmentation means giving a human a tool that makes them better at a task they still own. The first removes the person from the loop. The second keeps them in it and raises the ceiling on what they can do.

Most real work is a blend. There are parts of any job that genuinely should be automated — they are repetitive, low-judgment, and soul-deadening. And there are parts that should never be fully automated, because they require taste, accountability, context, or care. The skill — and the ethical choice — is in knowing which is which.

How ETAPX Actually Thinks About Automation Versus Augmentation

Our default is augmentation. Not because automation is evil, but because the things worth building are the things that make a person more powerful, not the things that make a person unnecessary. When we look at a workflow, we ask a specific sequence of questions before we point AI at it.

1. Who owns the outcome? If a human is accountable for the result, the human stays in the loop. AI can draft, suggest, accelerate, and check — but the person who answers for the work makes the final call.
2. Is this drudgery or is this judgment? Drudgery is fair game for automation. Judgment is where we augment, because removing the human removes the very thing that made the work good.
3. Does removing the person remove the learning? Many "inefficient" tasks are how people build expertise. Automate the wrong one and you hollow out the pipeline of skilled people you will need in five years.
4. What is the failure mode? If the AI is wrong, who notices and how fast? Augmentation keeps a human positioned to catch failures. Full automation hides them.
5. Does this make the product better, or just cheaper? Cheaper is not a goal we find interesting on its own. Better is. If automation only lowers cost while degrading the experience, we do not ship it.

That last question is the one that quietly governs everything. ETAPX is user-first and craft-driven, and craft is the part of work that resists automation by nature. You cannot automate taste. You can only give people with taste better instruments.

What this looks like inside Ocsidian

Ocsidian is the clearest expression of our philosophy because game creation is one of the most labor-intensive creative disciplines that exists. Building a playable 3D world traditionally requires a long chain of specialists: scene composition, physics tuning, rendering and post-processing, scripting, asset management, animation, navigation meshes for AI pathfinding, and deployment. Each of those is a discipline, and most solo creators simply cannot do all of them.

The BlackFrost Engine — Ocsidian's proprietary runtime and editor core — handles the technical complexity so creators can focus on the experience they want to build. It runs Havok physics natively in the editor and in exported projects. It has a TypeScript-first scripting system with hot reload. It offers visual node editors for materials, particles, geometry, and render graphs. And on top of all that, it layers AI-powered agentic workflows.

Here is the part that matters for the layoffs conversation. The agentic workflows in Ocsidian are not designed to remove the creator. They are designed to remove the gap between a creator's vision and their technical ceiling. The platform's own framing is deliberate: "from idea to playable world — with speed, taste, and full creative control." The phrase "full creative control" is not marketing decoration. It is a design constraint. The AI accelerates; the human decides. That is augmentation made literal.

"We did not build Ocsidian so that no one would need to make games. We built it so that the person with the idea — and no studio behind them — could finally make the thing in their head."

— ETAPX Engineering

The Old Way Versus the New Way

To understand whether AI is destroying or expanding opportunity, it helps to be concrete about what changes.

In the old way, making a game meant either joining a studio with a payroll full of specialists, or learning every one of those specialties yourself over years, or assembling a team and the funding to pay them. The barrier was not imagination. It was the sheer surface area of technical skill required before you could ship anything. Thousands of people with great ideas never made a game because the ladder to the first rung was too tall.

In the new way, that surface area collapses. A solo creator can scaffold a project from a template that ships pre-configured with the engine, physics, and editor tooling. They can use AI agents to handle parts of the work that would have required hiring. They can hit play and preview the result in real time. The barrier moves from "can you operate the entire pipeline" to "do you have something worth making."

Now, here is the question the replacement story refuses to ask: did that destroy jobs, or did it create creators? When you lower the barrier to making something, you do not just redistribute a fixed amount of work. You expand who gets to participate. The number of people who can credibly make a game goes up, not down. Some of those people will build studios. Some will hire. Some will collaborate. The pie does not stay the same size — it grows, and it grows in the direction of people who were previously locked out.

This is the pattern history keeps repeating

Every major tool that lowered a creative or technical barrier produced the same pattern. Desktop publishing did not eliminate publishing; it created millions of small publishers. The digital camera did not end photography; it produced more photographers than the world had ever seen. The spreadsheet did not eliminate financial analysis; it made analysts vastly more capable and created entire new categories of analytical work. In each case, the doom prediction focused on the jobs that existed, and missed the work that did not exist yet.

We are not naive about this. The transitions were real, and they were hard on specific people in specific roles. That is the honest tension, and we will get to it directly. But the macro pattern is consistent enough that betting against it requires believing this time is fundamentally different — and "this time is different" is one of the most expensive sentences in the history of forecasting.

The build-versus-buy decision is also a values decision

There is a quieter version of the automation question that rarely makes the headlines but shapes everything downstream: build or buy. When a company needs a capability, it can develop the depth in-house — which means hiring, training, and growing people — or it can buy a black-box solution that promises to do the job with no humans attached. The buy path looks efficient on a quarterly basis. It is also how organizations quietly hollow themselves out.

ETAPX builds in-house on purpose, and the layoffs conversation is part of why. When you build your own engine, your own platforms, and your own connective tissue, you are making a structural commitment to having skilled people who understand the work deeply. That commitment is the opposite of the replacement instinct. A company that buys every capability off the shelf never needs to develop anyone, and so it never has anyone to develop. A company that builds is, almost by definition, a company that invests in people. The architecture of the business and the philosophy toward labor turn out to be the same decision viewed from two angles.

The Honest Tensions We Refuse to Wave Away

A piece that only told the optimistic story would be propaganda, and we have no interest in writing propaganda. There are real tensions in the AI-and-work conversation, and pretending they do not exist is how you lose the trust of the very people you claim to serve. So here they are, stated plainly.

• Displacement is real even when net employment grows. "The economy creates new jobs" is cold comfort to a specific person whose specific role was automated this quarter. Aggregate optimism does not pay an individual's rent. The transition costs land on real people, often the ones with the least cushion.
• The timeline can be brutal. New categories of work emerge, but not always fast enough to absorb the people displaced from old ones. The gap between "job removed" and "new job available" is where real suffering lives. Speed matters, and AI capability is currently moving faster than most retraining systems.
• Augmentation can quietly become replacement. A tool sold as "helping you be more productive" can become the justification for expecting one person to do the work of three. Productivity gains do not automatically flow to workers. Sometimes they flow entirely to margins, and the "augmented" worker just gets a heavier load.
• The benefits and the costs are unevenly distributed. The people who own the AI capture most of the upside. The people whose tasks are automated absorb most of the downside. Without deliberate choices, AI is a wealth concentrator.
• Skill erosion is a genuine risk. If AI does the hard parts, people may stop developing the underlying competence. The first generation augmented by a tool understands what is happening underneath. The danger is the next generation that only knows the tool.

We name these because they are true, and because the companies that pretend AI is pure upside are exactly the companies you should not trust. The right response to a real tension is not denial. It is design.

Why we say "augmentation can become replacement" out loud

This is the tension closest to home, so we will be specific about it. When you build a tool that makes someone three times more productive, you have handed their employer a choice. The employer can keep the team the same size and triple the output. Or they can cut the team to a third and keep the output flat. The tool does not decide. The tool just creates the option.

We cannot control what every company does with productivity. But we can control what we build, and we can be honest that the gift of leverage always comes with that fork in the road. The companies that use AI leverage to do more — better products, faster iteration, new things that were previously impossible — will pull ahead of the companies that use it only to do the same with fewer people. That is our bet, and we are building the company that way on purpose.

Why ETAPX Builds Tools That Empower People

The reason we keep coming back to augmentation is not sentiment. It is a conviction about where durable value actually comes from. A product is only as good as the judgment, taste, and care poured into it. Strip those out in the name of efficiency and you get something cheaper that nobody loves. We are not interested in the race to the bottom on cost. We are interested in the race to the top on craft.

Building everything in-house is part of this. ETAPX builds its own engine, its own platforms, its own connective tissue. That is a deliberately harder path than assembling someone else's pieces. We take it because craft requires control over the details, and control over the details requires people who care about them. An organization optimized purely for automation has no reason to develop that depth. An organization optimized for craft cannot survive without it.

This is also why the user comes first in our framing rather than the cost structure. When you start from "what does the person using this actually need," you naturally land on augmentation, because users do not want to be removed from their own creative work — they want to be better at it. The replacement story only makes sense if you start from the balance sheet and work backward. We start from the human and work forward.

"The companies that win the next decade will not be the ones that fired the most people the fastest. They will be the ones whose people, armed with AI, did things their competitors could not even attempt."

— AJ, Founder & CEO, ETAPX

A Closer Look at the Agentic Shift

Most of the recent anxiety about AI and jobs is really anxiety about agents. A chatbot that answers questions is a tool you operate. An agent that takes actions, makes decisions across multiple steps, and completes whole workflows feels like something closer to a worker. That is the leap that has people genuinely unsettled, and it deserves a careful, non-hysterical look — especially from a company that ships agentic systems.

GLSRM tracks Agents as one of its core categories precisely because this is where the field is moving fastest. We watch the agentic frontier daily, and the honest picture is more nuanced than either the boosters or the doomers admit. Agents are genuinely capable of stringing together long sequences of work. They are also genuinely unreliable in ways that matter, and that unreliability is not a temporary bug to be patched away next quarter — it is structural, because agents accumulate small errors over many steps and have no intrinsic sense of when they have gone wrong.

What this means in practice is that the most useful agentic systems are not autonomous replacements. They are powerful collaborators that need a human positioned to set the goal, check the trajectory, and own the outcome. The agent does the multi-step grind; the human supplies the direction and the accountability. This is not a limitation we are waiting to engineer away. It is the correct division of labor, and designing around it is a feature, not a stopgap.

How Ocsidian's agents stay in their lane

Inside Ocsidian, agentic workflows operate against a backdrop of total creative control for the human. The agent can help compose a scene, wire up logic, or accelerate asset work — but the creator remains the author. The platform even includes self-healing project repair that automatically patches import paths, engine package links, and physics binaries, which is a perfect small example of the philosophy: automate the tedious, error-prone plumbing that nobody enjoys, so the human never has to think about it, while leaving every creative and design decision firmly in the creator's hands. That is what good agentic design looks like — aggressive automation of the mechanical, zero automation of the meaningful.

What AI Is Genuinely Good At — and What It Is Not

Clear thinking about jobs requires clear thinking about capability. The hype cycle blurs this on purpose. From building agentic systems daily, here is our practical read on the line between what AI handles well and what it does not.

Where AI is strong today

• Compression of effort: Tasks that used to take hours of mechanical work — scaffolding, boilerplate, first drafts, routine transformations — collapse to minutes. This is real and it is large.
• Breadth over depth: AI is astonishingly broad. It can attempt a wider range of tasks than any individual specialist, which makes it a superb generalist assistant.
• Pattern-heavy work: Anything that is mostly recognizing and reproducing patterns — formatting, summarizing, translating between structures — plays to its strengths.
• Tireless iteration: It does not get bored on the hundredth variation. For exploration and brainstorming, that endurance is genuinely useful.

Where AI is weak, and likely to stay weak for a while

• Accountability: AI cannot be responsible for an outcome. It cannot be fired, sued, or trusted with a reputation. Someone human has to own the result, and ownership is the heart of most real jobs.
• Taste and judgment under ambiguity: Knowing which of ten good options is the right one for this audience, this moment, this brand — that is taste, and taste is built from lived context AI does not have.
• Genuine novelty: AI is excellent at remixing what exists and weaker at the leap into what has never existed. The truly new idea still comes from a person.
• Stakes and care: When it matters that someone gives a damn, the human presence is the product. You cannot automate caring.

Notice that the list of weaknesses maps almost exactly onto the most valuable, most human parts of any job. That is not a coincidence. It is the clue to where work goes next.

What Comes Next for Workers

If AI absorbs the mechanical layer of work and humans keep the judgment layer, then the shape of valuable work changes in predictable ways. We are not predicting the end of work. We are predicting a shift in what work consists of.

• The premium moves to judgment. When execution gets cheap, deciding what to execute becomes the scarce skill. Knowing what is worth making is worth more than knowing how to make it.
• Generalists who direct AI gain leverage. The person who can orchestrate AI across several domains — describe intent precisely, evaluate output critically, and stitch the pieces together — becomes extraordinarily productive. Directing the work becomes a core competency.
• Taste becomes a hard skill. When everyone can produce competent output instantly, the differentiator is the ability to produce output that is actually good. Taste graduates from "nice to have" to "the whole job."
• Verification becomes a job in itself. Someone has to check whether the AI is right. As AI does more, the value of skilled review goes up, not down.
• The ceiling rises for the ambitious. The most exciting outcome is not that work gets easier. It is that the scope of what one motivated person can attempt expands dramatically. The solo creator can now think like a studio.

What this means for creators specifically

Creators are the clearest winners in the augmentation story, which is exactly why our ecosystem is built around them. A creator's bottleneck was never imagination — it was the technical and logistical overhead between the idea and the audience. AI eats that overhead. The creator who used to spend eighty percent of their time on production grind and twenty percent on actual creative decisions can flip that ratio.

This is the connective logic of the ETAPX ecosystem. Ocsidian lowers the barrier to making. Whistlr lowers the barrier to connecting and being discovered. GLSRM lowers the barrier to staying current in a field that moves faster than any individual can track. One account connects a person across all three. The throughline is the same: take the overhead off the human so the human can do the part only a human can do.

What Companies Owe Their People in This Transition

If displacement is a real cost, then companies deploying AI have real obligations. We hold ourselves to these and we think the rest of the industry should too.

1. Retrain before you replace. The first move when a task becomes automatable should be to move the person up the value chain, not out the door. The expertise they already have is an asset, not a liability.
2. Share the productivity gains. If AI makes a team three times more effective, some of that upside should reach the team. Capturing all of it as margin is a choice, and a short-sighted one.
3. Be honest about timelines. Workers can adapt to change they can see coming. Surprise is what destroys trust. Tell people what is changing and give them runway.
4. Preserve the learning pathways. Do not automate away the entry-level rungs that people climb to become experts. If you do, you win this quarter and lose the next decade.
5. Keep humans accountable for human outcomes. Anything that affects people's lives, livelihoods, or safety needs a human who owns the call. "The model decided" is not an acceptable answer.

The Economics Nobody Wants to State Plainly

Underneath the moral debate about AI and jobs sits a colder economic reality, and it is worth stating plainly because most coverage dances around it. When the cost of producing something drops dramatically, two things can happen, and they are not mutually exclusive. Demand can stay flat, in which case the same output now requires fewer people — that is the layoffs scenario. Or demand can expand to meet the lower cost, in which case more gets produced and the total work can hold steady or even grow — that is the expansion scenario.

Which one dominates depends on something economists call elasticity: how much the demand for a thing grows when its price falls. For commodities with fixed demand, cheaper production means fewer workers. But for creative and knowledge work, demand has historically proven remarkably elastic. When making games got cheaper, the world did not decide it wanted the same number of games at lower cost. It wanted vastly more games, more varied games, games for niches that were never viable before. The same is true of video, writing, software, and design.

This is the strongest economic case for optimism, and it is why we build for creators. Creative demand is close to bottomless. There is no ceiling on how many stories people want, how many worlds they want to explore, how many tools they want for their specific situation. When you lower the cost of creation in a market with near-infinite latent demand, you do not shrink the labor pool. You unlock a wave of creation that was previously economically impossible — and that wave needs people to ride it.

But elasticity is not guaranteed, and that is the catch

We want to be intellectually honest here, because this is where optimists usually stop and declare victory prematurely. Elasticity is an empirical fact about specific markets, not a universal law. There are kinds of work where demand really is roughly fixed, and in those areas, cheaper production genuinely does mean fewer jobs. Pretending otherwise to preserve a comforting narrative helps no one. The right move is to be honest that the expansion story is strong where demand is elastic — which includes most of what we build for — and weaker elsewhere, and to direct effort and support accordingly.

The Specific Danger of Building AI Without Values

We track the AI industry closely through GLSRM, which means we watch how the field talks about itself. The Wire, our live newsfeed, runs across News, Models, Releases, Research, and Agents. Data Pulse tracks hundreds of models and dozens of labs daily. When you watch the frontier that closely, a pattern becomes obvious: capability is racing ahead of consensus on how to use it well.

That gap is the actual risk. Not that AI becomes too powerful in some abstract sci-fi sense, but that powerful tools get deployed by organizations that never stopped to ask what they were optimizing for. A model that can write code, generate art, draft documents, and run agentic workflows is a profoundly neutral thing. Aimed at empowering people, it is one of the best tools ever built. Aimed at extracting maximum cost savings with no regard for the humans involved, it is a wrecking ball.

This is why values are not a soft add-on to AI development. They are the steering wheel. A company without a clear position on automation versus augmentation will default to whatever the spreadsheet suggests, and the spreadsheet always suggests subtraction. Having an explicit, defended philosophy — the kind we have tried to lay out here — is what keeps a powerful tool pointed at the right target.

Why We Are Optimistic Without Being Naive

Optimism is easy to mistake for naivety, so let us be precise about the difference. Naive optimism says everything will be fine and ignores the costs. Grounded optimism acknowledges the costs, takes them seriously, and still concludes that the trajectory bends toward more human capability rather than less.

Our optimism rests on a simple observation from inside the work: the more capable AI gets, the more obvious it becomes how much the human contributes. Every time we hand a harder task to an agent, the value of the human judgment around it goes up, not down. The agent can do the thing — but deciding whether the thing is good, whether it serves the user, whether it is the right thing at all, remains stubbornly human. AI has not shrunk the human role in our work. It has clarified it.

That is the future we are building toward. Not a world with fewer people doing less, but a world where more people can do more, where the barrier to making something great keeps falling, and where the parts of work that machines cannot touch — taste, care, judgment, accountability, the genuine new idea — become the most valued things in the economy. That is a better world for workers and creators, not a worse one. But it is not automatic. It has to be built on purpose, by people who decided it mattered.

Frequently Asked Questions

Will AI cause mass layoffs?

AI will change which tasks are valuable and will displace specific roles, and that disruption is real for the people it affects. But "mass layoffs" frames AI as the actor, when the real actors are the companies deciding how to deploy it. The same capability that justifies cutting a team can be used to make that team dramatically more effective. Whether a given company chooses subtraction or expansion is a business decision, not an inevitability of the technology. Historically, tools that lowered barriers expanded the number of people who could participate in a field rather than shrinking it — but the transition was hard on individuals, which is exactly why how companies handle it matters so much.

What is the difference between automation and augmentation?

Automation hands a task to a machine so a human never touches it again — it removes the person from the loop. Augmentation gives a human a tool that makes them better at a task they still own — it keeps the person in the loop and raises their ceiling. ETAPX defaults to augmentation. We automate genuine drudgery, but for anything involving judgment, taste, or accountability, we build tools that empower the person rather than replace them.

How does ETAPX use AI in its products?

AI runs through all three ETAPX products differently. GLSRM uses it to track and distill the AI industry into a clean, continuously updated intelligence surface. Ocsidian uses agentic workflows to help creators build playable 3D worlds without an entire studio — handling technical complexity while keeping the creator in full creative control. Whistlr is the social layer that connects people across the ecosystem. In every case, AI is pointed at removing overhead from the human, not removing the human.

Does AI augmentation just lead to replacement anyway?

It can, and we say so openly. A tool that makes someone three times more productive hands their employer a choice: do more with the same team, or do the same with fewer people. The tool does not decide — the organization does. Our bet, and the way we run our own company, is that the firms using AI leverage to do more and better work will outcompete the ones using it only to cut costs. But we will not pretend the risk does not exist, because it does.

What jobs are safest as AI advances?

The work that is hardest to automate clusters around the things AI is weakest at: accountability for outcomes, judgment under ambiguity, taste, genuine novelty, and care where it matters that a human is present. As execution gets cheaper, the premium shifts to deciding what is worth making, directing AI effectively across domains, and verifying that the output is actually good. Generalists who can orchestrate AI and people with strong taste are particularly well positioned.

What do companies owe workers during the AI transition?

We believe companies deploying AI have real obligations: retrain people before replacing them, share productivity gains rather than capturing all of them as margin, be honest about timelines so workers can adapt, preserve the entry-level pathways through which people become experts, and keep humans accountable for any outcome that affects people's lives. Aggregate economic optimism does not pay an individual's rent, so the transition has to be managed deliberately and humanely.

Is ETAPX optimistic or worried about AI and jobs?

Both, in the right proportions. We are worried about the honest tensions — displacement, brutal timelines, uneven distribution of benefits, and skill erosion — and we refuse to wave them away. But we are ultimately optimistic, because from inside the work we see that the more capable AI becomes, the clearer and more valuable the human contribution gets. Our optimism is grounded, not naive: the better future is achievable, but only if it is built on purpose by people who care.

How can creators prepare for an AI-augmented future?

Lean into the parts AI cannot do. Develop your taste and judgment, because when everyone can produce competent output instantly, the ability to produce genuinely good work becomes the whole job. Learn to direct AI precisely and to evaluate its output critically. And treat AI as the way to remove overhead from your craft so you can spend more time on the creative decisions that only you can make. Tools like Ocsidian exist precisely to collapse the gap between a creator's vision and their technical ceiling — the creators who thrive will be the ones who use that leverage to attempt bigger things, not smaller ones.

What Comes Next — and How We Intend to Build It

The layoffs conversation will keep getting louder as AI gets more capable, and most of the noise will continue to treat the technology as a force of nature that simply happens to people. We reject that framing entirely. AI is built by people, deployed by people, and pointed at targets that people choose. The question was never whether AI is good or bad for jobs. The question is what we decide to build, and who we decide to build it for.

ETAPX's answer is to keep building tools that make people more powerful, to keep the human accountable for the things that matter, and to keep betting that craft, taste, and care are the most durable forms of value in any economy AI touches. We will keep watching the frontier move through GLSRM, keep lowering the barrier to creation through Ocsidian, and keep connecting people across the ecosystem through Whistlr — one account, one connected set of tools, all aimed in the same direction.

The future of work is not something that will be done to us. It is something we are building, decision by decision, tool by tool. We intend to build the version where more people get to do more of the work that only humans can do — and we intend to be honest, the whole way, about how hard and how important that is to get right.