For the past few months, I have been continuously building Toket. What started as a collection of AI tools has gradually become an independent AI product lab.
Today, Toket includes AI cost analysis, prompt optimization, model knowledge, Workspace, News, Knowledge, and new experiments around GEO and SEO.
But yesterday was different. I did not launch a major feature. I did not ship a big update. I simply took some time to rest. ⸻ In traditional software development, independent builders were often limited by coding ability and execution speed.
AI is changing that. With tools like: ChatGPT for product thinking; Codex for engineering analysis; Cursor for implementation; a single builder can now move much faster.
But when execution becomes easier, a new challenge appears: Decision making. ⸻ The hardest part of building a product is no longer only writing code.
It is deciding: What should be built? What should be removed? What actually creates value? ⸻ The Human + AI Collaboration Experiment behind Toket is about exploring this question.
Over the next month, I will document: How AI participates in product decisions; Which ideas become real features; Which directions are abandoned;
How existing capabilities can be connected into something meaningful. The goal is not to prove AI replaces builders. The goal is to understand what a builder and AI collaborators can create together.
⸻ Yesterday’s rest was also part of the experiment. Because building a product is not a sprint. The most valuable resource is the ability to keep making good decisions.
Toket continues. Building Toket.
Estimate task cost in the AI Cost Analysis or refine prompts in the Prompt Optimizer.