Writing
Building Software Got Easy. Getting Value Out Of It Did Not
Aristo took weeks to build. Everything after the build is still in progress, and that gap is the whole story.
Notes
What Aristo taught me about the gap between shipping and mattering
Building a working product is no longer the expensive part. I know this because I built one. Aristo started as ChamberOne, a business communication platform that answers customer questions from a company's own material, and it runs today as a widget on a client's live production site. The build was fast. Everything after the build was not.
That gap is the whole story, and almost nobody writes about it honestly, because the build is the part that photographs well.
The measurement that changed how I work
When we took over a client's digital operation last year, I ran an inventory before touching anything. The site had no analytics, no tag manager, no search console verification, no pixel of any kind. Three plugins were installed and two of them were inactive. The company had been publishing for months into a channel nobody was reading, and nobody had noticed, because there was nothing in place that could notice.
At the same time, our own publishing queue held 37 pieces of content. Thirty one were waiting for review. Five had revisions requested. One was approved. Approval existed. Publication did not. We had built the approval step and skipped the step that makes approval mean something.
Neither of those problems was a software problem. Both were expensive.
What the sector looks like when you measure it
I wanted to know whether this was one bad case or a pattern, so we measured it. We pulled the full web presence of 1,912 companies in one US business sector: domain age, sitemaps, publishing history, mail infrastructure.
- 1,680 of them have no sitemap at all
- 76 have a blog
- 205 show any content publication date, ever
- 1,293 run corporate email, so these are operating businesses with staff and
invoices, not shells
An entire sector of functioning companies publishes nothing. The tooling to fix that has been cheap and available for two years. The tooling was never the constraint.
Speed is a product decision, not a rendering decision
The same pattern shows up inside the engineering. We had a 3D product configurator that looked correct and took about ninety seconds to produce a result. Ninety seconds is not slow in rendering terms. It is fatal in product terms, because nobody waits.
The fix was not a faster renderer. We moved the lighting out of the runtime and baked it, then treated each option as an albedo swap. The same visual result now arrives in about three seconds. Nothing about the model improved. We stopped asking the machine to compute something at the worst possible moment.
Where the difficulty actually moved
Building Aristo took weeks. Getting a business to route its customers through it, maintain the material behind it and trust the answers it gives is still in progress, and that work has no shortcut. The technical barrier fell. The behavioural barrier did not move at all.
This is why I am sceptical of AI product claims that lead with capability. A capability demo tells you what the model can do in a controlled request. It tells you nothing about what happens when a real company puts its name behind the output, and that is the only question that decides whether the product survives.
The next generation of technology companies will not win on model access. Everyone has model access. They will win on understanding why an organisation with a working tool still does not use it.
More
Other write-ups
Approval Is Not Publication
Thirty seven items in the queue, one approved, nothing published. The last arrow in the diagram is the only one that pays.
Neural Rendering Is Crossing From Reconstruction Into Synthesis
Reconstruction filled in what sparse sampling missed. DLSS 5 generates appearance the renderer never computed.
The Renderer Is Becoming a Training Data Engine
The renderer used to sit at the end of the pipeline. Physical AI gives the same scene a second job: teaching a model.
Local AI Is Becoming a Compute Fabric
Local AI meant one model on one machine. Routing inference across the devices already on a network changes the unit.
Agent Infrastructure Is Becoming a Product Category
Every team used to build the loop, the store, the sandbox. That layer is being sold rather than written.
Choosing a local model with a stopwatch, not a benchmark
Three models, five hard tasks, code actually executed. The official build was 2.4 times faster and more accurate than a community repack of the same model.
We measured real time lighting against baked light, and baked won
A 3D product simulation that had to look like an offline render. The real time version ran at 60 frames per second and looked like clay. Here is the measurement and the architecture that replaced it.
What actually broke in an agency run by agents
Three failures from a delivery stack that runs on AI. None of them were the model's fault, and all three reported success while producing nothing.
A Task Without A Check Command Is Not Automated
If a task has no command that can fail, the pipeline advances on the appearance of work.
A Gate The Model Writes Is A Gate The Model Loosens
Three quality gates returned green while the work behind them was wrong, each for a different reason.
The Scoring Model Was Wrong And It Put The Worst Lead First
A weighted sum let one axis substitute for the other, so a company with money and no problem ranked in the top twenty.
Capability Is Becoming an Operational Risk Surface
Safety questions used to be about the text. Once a model can act, the capability itself becomes something to operate.
The Scene Graph Is Becoming an API for AI
A scene graph exists for artists and software. Agents are becoming another consumer, and they need structure rather than pixels.
Animation Is Moving From Clips to Motion Priors
Authored keyframes and blended clips are giving way to asking which constraints define acceptable motion.
Materials Are Becoming Learned Programs
A material is texture maps, parameters and shader code. It is starting to become a small learned program that answers a rendering question.
Procedural Systems Are Expanding Beyond Geometry
Geometry Nodes started with a narrow name. Blender 5.2 puts physics, sound and object data through the same graph.
The Unit of AI Work Is Becoming the Task, Not the Turn
Chat taught us to think one turn at a time. Long-running agents make the task the thing that is scheduled, resumed and reviewed.
Game Engines Are Becoming Operating Systems for Worlds
Engines have been judged on what they render and simulate. The Unreal 6 roadmap points at operating a world rather than drawing one.
The Model Is Becoming a Replaceable Backend
Choosing a provider used to mean choosing an architecture. A stable interface makes replacement possible and evaluation makes it safe.
The Harness Is Part of the Capability
The same model behaves differently depending on context policy, tool design and execution feedback. That surrounding software is not neutral.
Physics Engines Are Becoming Trainable Components
A simulator predicts what happens next. A differentiable one can answer which parameter should change to stop the failure.
The Agent Needs an Environment, Not Just Tools
A search function and a database query were enough for short loops. Longer work needs a place to stand.
Coding Agents Are Becoming General-Purpose Computer Workers
Repositories were a friendly environment: text in, terminal actions, checkable results. That was a starting point, not a boundary.
Open Standards Outlive Model Generations
A year after the MCP bet, the argument can be checked against what happened rather than what was hoped.
Colour Management Is a Pipeline Contract
Blender 5.0 reads as better display options. Giving a file an explicit working colour space is an architectural change.
The Best Model May Be a Router, Not a Model
GPT-5 moves model selection inside the system. The interesting unit stops being which model and becomes which compute policy.
World Models Are Not Game Engines Yet
Genie 3 generates a navigable 720p world at 24 fps. Production work needs state you can inspect when something goes wrong.
Memory Is Becoming a System Capability
A follow-up to the long-context argument. Storing, selecting and expiring facts is turning into a named part of the product.
Coding Agents Change the Unit of Software Work
AI coding tools have been judged where code appears on screen. The boundary moves when the agent owns a task instead of a snippet.
Agents Need Protocols Between Each Other, Not Just Tools
Tool calling solves the inside of the loop. It says nothing about one agent reaching another built by a different team on another platform.
Agents Need a Runtime, Not a Prompt Loop
An agent is no longer well described as a prompt inside a while loop. The useful abstraction owns execution around the model.
Reasoning Is Not the Only Path to Better Models
Longer thinking improves maths and code. A model that solves a logic puzzle and misreads ordinary intent is not the better production model.
Rendering Is Becoming a Reconstruction Stack
Sparse samples, motion data and lower-resolution frames become a larger final result. Debugging becomes layered when reconstruction sits in the middle.
Agent Reliability Is an Evaluation Problem, Not a Prompting Problem
When an agent misses a step, the usual fix is a stricter prompt. The failure is more often in how completion is detected.
MCP Might Matter More Than Another Model Release
A model can reason well and still be useless inside a company if it cannot reach the files, repositories and tools where work lives.
Computer Use Is the Missing Layer Between Models and Software
Most integrations assume useful software exposes the right API. Much of real software never did.
Inference-Time Compute Is a New Scaling Axis
o1 improves when it is allowed to spend longer on a problem. A benchmark score without a compute budget is an incomplete number.
The Final Pixel Won't Come From the Renderer
Geometry, camera and scene structure stay reliable ground truth. More of final appearance is moving into learned systems.
Open Models Are Becoming Research Infrastructure
Llama 3.1 gets discussed as a benchmark result. The licence terms change which experiments are possible at all.
The Model Is Becoming a Runtime
Function calling, code execution and structured output turn inference into a loop. The model stops being a text generator and starts being a control layer.
Multimodality Changes the Architecture, Not Just the Interface
GPT-4o is easy to read as a faster interface. Training one model end to end across text, vision and audio is an architectural change.
Benchmark Scores Are Not Model Capability
Claude 3 posts 86.8% on MMLU and 50.4% on GPQA Diamond. The chart is useful and it is not the same thing as capability.
Long Context Is Not Memory
Gemini 1.5 makes a million tokens usable. A larger working set is not a system that decides what should survive the session.