Live data from Hacker News

Opus 4.7 is horrible at writing

news.ycombinator.com

21–30 of 33 posts

Re: Opus 4.7 is horrible at writing

#21
I see a lot of complaints on X about 4.7. Boris just dropped a post on how to use Opus 4.7 in Claude Code.

I guess they broke continuity with a 0.1 in model version change in some ways.

Re: Opus 4.7 is horrible at writing

#25
post #13

Maybe don't use any LLM to do your writing for your Master's thesis?

wrong. You can actually get great stuff done. instead of just having a lame master thesis, you can actually build meaningful stuff and write about it. You can basically have a PhD done for your master and then get a tripple PhD if you're actually pursuing a PhD...

Re: Opus 4.7 is horrible at writing

#26
post #16

Earlier quoted context omitted.

So why should you use it for coding then?

You shouldn't. Using LLMs for coding is also considered fraud by many academic boards if that code is a core contribution of the thesis.

where do you guys live. its 2026 and this is the new world order. i'm a researcher at one of the best labs in the world and we use claude code 24/7 for our bioinformatics needs.

Re: Opus 4.7 is horrible at writing

#27
post #13

Maybe don't use any LLM to do your writing for your Master's thesis?

wrong. You can actually get great stuff done. instead of just having a lame master thesis, you can actually build meaningful stuff and write about it. You can basically have a PhD done for your master and then get a tripple PhD if you're actually pursuing a PhD...

[dead]

Re: Opus 4.7 is horrible at writing

#28
post #16

Earlier quoted context omitted.

You shouldn't. Using LLMs for coding is also considered fraud by many academic boards if that code is a core contribution of the thesis.

where do you guys live. its 2026 and this is the new world order. i'm a researcher at one of the best labs in the world and we use claude code 24/7 for our bioinformatics needs.

Will you be mailing your master's degree to Anthropic or claiming it as your own?

Re: Opus 4.7 is horrible at writing

#30
post #21

I see a lot of complaints on X about 4.7. Boris just dropped a post on how to use Opus 4.7 in Claude Code. I guess they broke continuity with a 0.1 in model version change in some ways.

They swapped the tokenizer which either means a new pretrain, or token/weights surgery. The latter one seems more likely both because

- economics: i'd wager a bet that Opus 4.7 is just distilled Mythos Preview - performance: surgery like this would explain the spiky performance and weird issues

just spitballing tho

Post reply on HN