{"slug": "i-built-one-project-in-a-year-and-another-in-two-months-what-changed-with-ai", "title": "I Built One Project in a Year and Another in Two Months: What Changed With AI-Assisted Development", "summary": "A developer reports that building the AlbumMap video platform took roughly two months with GPT-5.5, compared with about a year for the earlier EngHub project using GPT-5, though the developer cautions the two projects were not controlled experiments. The developer argues that evaluating AI coding assistants on generated code quality alone is misleading and that defining expected end-to-end behavior before prompting produces better results.", "body_md": "[EngHub](https://enghub.network/) took me roughly a year to develop, with GPT-5 being part of my development workflow.\n\nMy newer project, [AlbumMap](https://albummap.com), took about two months to build while I was using GPT-5.5.\n\nThat difference was difficult for me to ignore.\n\nThe newer model felt noticeably more efficient during my own development process. I could move from an idea to an implementation faster, iterate on problems more quickly, and spend less time getting stuck on smaller technical issues.\n\nBut I do not think the right conclusion is:\n\nGPT-5.5 made me six times faster.\n\nThese were two different projects, built at different times, with different requirements. They were not controlled experiments.\n\nWhat the experience did show me was something more useful: AI coding tools are becoming much better at reducing the distance between an idea and a working implementation.\n\nThere are many factors that affect how long a software project takes.\n\nScope matters. Experience matters. The technologies you already know matter. The amount of time you can dedicate to the project matters.\n\nEven your definition of “finished” can completely change the timeline.\n\nBy the time I started AlbumMap, I also had knowledge from previous projects that I did not have when I started [EngHub](https://enghub.network/).\n\nSo I cannot attribute the difference in development time entirely to GPT-5.5.\n\nWhat I can say is that working with the newer model felt significantly more efficient in my own development workflow.\n\nThat made me think about a more interesting question:\n\nHow should we evaluate AI coding assistants when we are using them to build real products?\n\nOne thing I have learned is that evaluating an AI coding assistant based only on the code it generates can be misleading.\n\nA piece of code can look clean.\n\nThe UI can render correctly.\n\nButtons can respond.\n\nAnd the feature can still be wrong.\n\n[AlbumMap](https://albummap.com/) is a map-based video platform that lets people combine photos, videos, messages, and locations into a visual story.\n\nUsers can also invite other people to contribute media, review those contributions, organize them, and generate the final video.\n\nThat creates a good example of why I think end-to-end behaviour matters more than generated code quality alone.\n\nImagine an organizer receives two video contributions.\n\nThey approve the first contribution and reject the second.\n\nAt the UI level, everything might appear to work perfectly:\n\nBut the real acceptance test is different.\n\nWhen the final video is generated, does the approved contribution appear?\n\nIs the rejected contribution actually excluded?\n\nThat is the behaviour the user cares about.\n\nNow make the scenario slightly more complicated.\n\nThe organizer has five contributions.\n\nThey approve three.\n\nThen they change their order.\n\nThey refresh the browser.\n\nThey return to the project later.\n\nFinally, they generate the video.\n\nA useful test would check:\n\nThese are simple questions, but they test the actual product rather than the appearance of the implementation.\n\nThat distinction becomes even more important when working with AI-generated code.\n\nAI can produce an implementation extremely quickly.\n\nBut speed increases the importance of having a clear way to determine whether the implementation is actually correct.\n\nI have also found that the quality of the result improves significantly when the expected behaviour is defined before asking an AI coding assistant to build the feature.\n\nFor example, instead of writing:\n\n“Make the contribution editor work.”\n\nA much better specification would be:\n\n“Build the contribution review flow so that only approved contributions appear in the preview and final export. Preserve the user’s selected order after refreshing the page. If saving fails, show a clear error and do not display the change as successfully saved.”\n\nNow there are measurable conditions.\n\nYou can test them.\n\nThe AI can reason around them.\n\nAnd if the implementation fails, it is much easier to identify where the problem is.\n\nThis has become one of the biggest changes in how I work with AI-assisted development.\n\nI try to spend more time describing what must be true when the feature is finished rather than simply describing what code I want generated.\n\nAlbumMap gave me a useful environment for testing this approach because the product involves several connected systems.\n\nUsers can collect media from different people, associate memories with locations, arrange content, change how the map appears, and eventually turn everything into a video.\n\nFor example, someone creating a [group birthday video](https://albummap.com/video-maker/birthday-video) might collect messages and videos from friends or family living in different cities or countries.\n\nThe interesting part is not simply uploading those files.\n\nThe system has to preserve the relationship between:\n\nAn AI coding assistant can help implement those systems faster.\n\nBut it still does not decide what the correct product behaviour should be.\n\nThat remains the developer’s responsibility.\n\nThis may be the biggest difference I noticed between my earlier project and building AlbumMap.\n\nAs coding assistance improves, writing the implementation is becoming less of the bottleneck.\n\nThe harder questions increasingly become:\n\nWhat exactly should this feature do?\n\nWhat happens when something fails?\n\nWhat should persist?\n\nWhat does the user expect to happen next?\n\nHow do I know the implementation is actually correct?\n\nThose are product and engineering questions rather than code-generation questions.\n\nAnd faster code generation makes them more important, not less.\n\nIf an assistant can generate a feature in minutes, it is very easy to move quickly in the wrong direction.\n\nThe contrast between these two projects made the improvement in AI-assisted development very noticeable to me.\n\n[EngHub](https://enghub.network/) took roughly a year.\n\nAlbumMap took roughly two months.\n\nAgain, I would not use those numbers as a benchmark for comparing GPT-5 and GPT-5.5.\n\nThere are too many other variables involved.\n\nBut from my perspective as the person building both projects, the development experience has changed substantially.\n\nI am spending less time fighting through implementation details and more time thinking about behaviour, product decisions, edge cases, and how the complete system should work.\n\nThat is where I currently see the biggest value of modern AI coding assistants.\n\nThey do not remove the developer from the process.\n\nThey reduce the distance between:\n\nidea → implementation → testing → iteration\n\nAnd when that distance becomes shorter, you can test more ideas, reject bad ones earlier, and spend more time improving the parts of the product that actually matter to users.\n\nFor me, that is much more interesting than simply measuring how many lines of code an AI model can generate.", "url": "https://wpnews.pro/news/i-built-one-project-in-a-year-and-another-in-two-months-what-changed-with-ai", "canonical_source": "https://dev.to/albummap/i-built-one-project-in-a-year-and-another-in-two-months-what-changed-with-ai-assisted-development-54io", "published_at": "2026-09-12 02:41:41+00:00", "updated_at": "2026-09-12 03:26:45.261779+00:00", "lang": "en", "topics": ["ai-tools", "developer-tools", "ai-products", "generative-ai"], "entities": ["EngHub", "AlbumMap", "GPT-5", "GPT-5.5"], "alternates": {"html": "https://wpnews.pro/news/i-built-one-project-in-a-year-and-another-in-two-months-what-changed-with-ai", "markdown": "https://wpnews.pro/news/i-built-one-project-in-a-year-and-another-in-two-months-what-changed-with-ai.md", "text": "https://wpnews.pro/news/i-built-one-project-in-a-year-and-another-in-two-months-what-changed-with-ai.txt", "jsonld": "https://wpnews.pro/news/i-built-one-project-in-a-year-and-another-in-two-months-what-changed-with-ai.jsonld"}}