1. X
  2. Michael Hönnig
Log inSign up
Michael Hönnig
5,179 posts
user avatar
Michael Hönnig
@javagil
agile cross stack software craftsman: backend: Java / Kotlin / JEE / Spring / PostgreSQL; frontend: TypeScript / JavaScript / Angular. Voluntarist.
Hage (Ostfriesland)
michael.hoennig.de
Joined August 2009
964
Following
536
Followers
RepliesRepliesMediaMedia
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Don't miss what's happening
People on X are the first to know.
Log inSign up

New to X?

Sign up now to get your own personalized timeline!

Create account

By signing up, you agree to the Terms of Service and Privacy Policy, including Cookie Use.

  • user avatar
    Michael Hönnig
    @javagil
    4h
    That was a close thing!
    18
  • user avatar
    Michael Hönnig
    @javagil
    10h
    He just nailed PoC → MVP → Production-ready in the age of software development with agentic AI:
    user avatar
    Tristan Rhodes
    @tristanbob
    12h
    It only takes 10 messages to your agentic coding tool to prove an idea will work. It takes 100 messages to get your MVP ready. It takes 1000 messages to make your app production ready.
    46
  • user avatar
    Michael Hönnig
    @javagil
    11h
    Great idea:
    user avatar
    Rod Johnson
    @springrod
    22h
    Ran Opus 5 in a loop overnight with instructions to optimise performance across 50 tests, some very difficult. Also requested regular review by another model to check for overfitting Woke up to find it had gamed it. Deleted 18 tests on flimsy grounds and stopped running reviews.
    58
  • user avatar
    Michael Hönnig
    @javagil
    Jul 25
    Interesting agenetic AI benchmark for software development:
    user avatar
    Paweł Huryn
    @PawelHuryn
    Jul 24
    Opus 5 fixed 11 of the 45 bugs I hid in my own repo. Opus 4.8 fixed 2. Same repo, same prompt, same setup. Opus 5 just shipped. I ran it straight through Bug Hunt Bench: 45 bugs across my VS Code extension, ~28K lines, the tests still passing so nothing gives the bugs away. It
    59
  • user avatar
    Michael Hönnig
    @javagil
    Jul 25
    Great statement about why code reviews make little sense when software gets developed with agentic AI:
    user avatar
    Саша шитпостит
    @sudo_belousov
    Jul 25
    Дядя Боб выдал базу разработки в 2026 году. Новые ИИ модели пишут качественный код лучше чем 98% программистов, ошибаются реже и работают в 10 раз быстрее. Число ошибок которые они делают - мизерное. И на процесс разработки надо уже смотреть не с позиции разработчика а тим
    56