The best engineers aren’t writing code anymore

Oct 7, 2025 · 11:02 AM UTC

What hasn't changed since Sonnet 3.5: - It can do small functions ok - It can't do anything else Give it a small problem with ~3 constraints and it will write code that looks sort of ok but upon inspection the constraints are not properly maintained throughout.
10
85
Replying to @Hasen_Judi
No improvement since sonnet 3.5?????? You outta your mind
7
288
No improvement
8
54
Replying to @Hasen_Judi
yes, and are you betting that it's not an S curve?
1
2
I'm not betting. This is not a prediction about the future. It's a description of the present. A new breakthrough is possible.
25
Replying to @Hasen_Judi
What does this say? Claude 4.5 is the same as claude 3.5? Does anyone actually agree to that?
1
47
I haven't seen any improvement from sonnet 3.5 to sonnet 4.5 If anything, the model got worse. Does not respect my system prompt as much. Really likes to babble even when I tell it to keep it short.
5
35
Replying to @Hasen_Judi
All I know is that GPT-5-codex in Codex cli and I have been able to solve and complete development of projects that stranded from hitting the capability limits of the tools available just a few months ago.
1
21
What kind of project? What did the previous models stall on?
2
3
Replying to @Hasen_Judi
This was created by someone typing a paragraph into an app on their cell phone. Tell me again how AI isn’t rapidly advancing faster than any of us can even begin to realize?
My favorite new trend in the Sora app is putting Pikachu in every movie. This is “Saving Private Pikachu” 👇
7
8
I was talking about LLMs My guess is that video generation is also going to platue similarly, if it hasn't already. The clip you posted is literally slop, and with current techniques it probably will not turn from slop to not-slop
2
1
28
Replying to @Hasen_Judi
Ngl, they likely aren't. If I know a system very well I can ask ChatGPT to do the tedious work of putting my thoughts into code. I then review it, accept/reject it, and stitch it with the code it made it write earlier. Continue until completed.
1
5
> the tedious work of putting my thoughts into code You hate programming
1
32
Replying to @Hasen_Judi
What are the axes?
3
4
Ask ChatGPT 🤣
1
2
Replying to @Hasen_Judi
No way that’s true. Maybe for you use case. 3.5 would hallucinate, was almost unusable by current standards
1
1
all the new models hallucinate still, even when writing code.
2
4
Replying to @Hasen_Judi
not even accounting for the energy costs involved, I would put it negative , we are actually regressing
1
1
I have a feeling that Sonnet 3.5 was much more useful than 4.5 but I can't really prove it because they don't provide access to 3.5 anymore .. otherwise I would have like to compare responses
2
3