I had the same exact issue. Turns out I have depression. Currently I'm on meds, therapy and functioning on 80% my ability but it's still better than near 0
I used AI, namely Claude Opus 4.8, to create native client for Mattermost using Slint with Rust bindings. To create such application you don't need AI, it's GUI and cross-platform and for exercise purposes it was perfect.
I manage to create functionally correct application but it fall very short in performance and required constant logic validation. GUI components also required constant oversight and manual fixing since AI was simply unable to fix it and gone in circles doing exactly the same thing 2 steps back. Sometimes it even broke perfectly correct component for no apparent reason
Lastly it tends to load data from server in very suboptimal way, ex. for each comment it loaded user profile often repeating request.
After a week I had GUI app which lacks in UI and no-one would use.
The same application to develop manually would take considerable longer but application would be useable and fixing above issues most likely would take longer due to accumulated tech debt
In which world throwing business logic to "optimize memory usage" isn't crappy solution? Or making 1T from 4G? It didn't returned decent solution, contrary it was worst than terrible
There's other word for such code and it sounds similar. It's crappy. I was tempted to use Claude couple of days ago to help me with optimization and from 4Gb RAM and 21s it managed to squeeze 1Tb and 41s. When I asked to fixed it AI reduced it to 1Gb and 10s. It just removed most heavy business logic. So yeah. 3 days and this was only total waste of time and cash
Cheap excuses. It's like saying Europe started without asphalt streets so we don't have it. Building/upgrading asphalt street is order of magnitude more costly than cable yet majority of streets aren't dirt and stones
The issue lies in willingness to upgrade and cutting costs on end users
Yes, because US CIA and NSA was hacking half of the World. You can learn about this in Cybersecurity books.
This is different situation. Cybersecurity specialists (at least those I found and read) don't consider Mythos as something really powerful. Good tool but not groundbreaking.
Anthropic was playing terror game and burned by it
I don't know any company with below good enough soft which became multibillion-dollar entity. Contrary M$ has started to have issues after recent "improvements"
But it's not good enough. We can see this all over industry where even M$ is producing software so bad even calculator is electron app. Slow, poor quality and for any engineer below acceptable
I still remember times when my work was mostly cleaning and refactoring old code written just to have higher LOC by devs from India paid in cents per line.
They were creating fast like AI and so mich like AI. But they were merely humans. Cleaning this slow and buggy mess took years killed many companies.
Today we have AI and industrial quantity of code without any engineering skill involved in the process
I had a taste of the other side for 2 weeks. I was building wood construction with my uncle who is doing this for 20 years.
I was digging 6 holes in full sun, 26 degree C, transporting 100 kg of dirt in 8 rounds, moved 600 kg of ceramic roof times, worked on rain and had sun burns which I treated 2 days with pain even sleeping