> you can already see this with AWS, NVIDIA and Microsoft beginning to suffer the early stages of LLM blight in their outputs. Things break, they're inefficient and they don't work as expected.
I don't think it's possible to identify those things as caused by LLMs. Microsoft has been producing inefficient code for many years.
Also, why didn't they try out Claude Code? It's the leader in the space and the pro plan isn't too expensive to try out for 1 month. Other than that I agree with most of what was been written. Especially about the tendency to write a lot of code that isn't highly efficient and the tendency to go off the rails.
this is a horrible review because they used mediocre tools / models. you also do not learn how to get great code out of these tools in just a week. it takes a lot of time to build up those skills.
> you can already see this with AWS, NVIDIA and Microsoft beginning to suffer the early stages of LLM blight in their outputs. Things break, they're inefficient and they don't work as expected. I don't think it's possible to identify those things as caused by LLMs. Microsoft has been producing inefficient code for many years.
Also, why didn't they try out Claude Code? It's the leader in the space and the pro plan isn't too expensive to try out for 1 month. Other than that I agree with most of what was been written. Especially about the tendency to write a lot of code that isn't highly efficient and the tendency to go off the rails.
this is a horrible review because they used mediocre tools / models. you also do not learn how to get great code out of these tools in just a week. it takes a lot of time to build up those skills.