Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

alex000kim.com

Post reply on HN
Print view View original

591–600 of 636 posts

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#591

Post by DrewADesign »

rkozik1989 wrote: post #569
DrewADesign wrote: post #510

Earlier quoted context omitted.

Python and C++ have been used for countless large projects— each one for many more than typescript. It’s all about trade-offs that take into account your tasks, available coders at the project’s commencement, environment, etc.

People like to put companies that are household names on pedestals, but the choices they make are mostly guided by what their people can do and which choices give them the most value for free. They mostly operate how smaller companies do but they have a bigger R&D budget to address issues like scale that the larger market has little incentive to solve.

Also, this product is like a year old… it has barely hit its teething phase. I wouldn’t be surprised if the core is still the prototype someone whipped up as a proof of concept.

I reckon some believe these companies are basically magical, and are utterly astonished when they’re shown to be imperfect in relatively uninteresting ways. I’m a lot more concerned about the sanity of the AI ecosystem they operate in than the stability of some front-end Anthropic made.

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#592

Post by fatcullen »

dtran wrote: post #400
fatcullen wrote: post #395

Earlier quoted context omitted.

Huh weird, they must have changed the algorithm up due to the leaks. Would be pretty easy, there's a constant seed variable so they'd just need to change that, figured they might. Too bad, sorry this didn't work out

Ah yea, I got an uncommon dragon instead of the rare duck. Did you get your legendary?

Nope :| Still a ghost, but it's just common now, too bad

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#593

Post by avdelazeri »

Reason077 wrote: post #50

> "Anti-distillation: injecting fake tools to poison copycats" Plot twist: Chinese competitors end up developing real, useful versions of Claude's fake tools.

Has Claude stopped claiming to be deepseek when prompted in Chinese yet? It wasn't long that it hit the news and blogs

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#594

Post by girvo »

xvector wrote: post #553
girvo wrote: post #455

Earlier quoted context omitted.

What a ridiculous, reductive and dangerously accelerationist viewpoint.

It's not accelerationist, it's just fact. Not saying we shouldn't be careful with AGI. But the glib tone of "who cares if these companies die?" is where one needs to consider the consequences of AGI not happening or being delayed.

It’s not a fact in the slightest.

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#595

Post by flykespice »

hnbad wrote: post #556
Pxtl wrote: post #133

Earlier quoted context omitted.

Why not? What's wrong with honesty?

Claude is not a person and AI doesn't gain authorship let alone copyright. Unless you literally vibe coded it, Claude is just a tool. This is the equivalent of Apple appending "Sent from my iPhone" as a signature to outgoing emails. It's advertising tool use, not providing attribution. The intent isn't to disclose that AI was used in creating the code, the intent is to advertise the AI product.

> Unless you literally vibe coded it, Claude is just a tool.

Stop with the selective bias, two are birds of the same feather, they are using a "tool" to write the code for them from whatever (questionable) mashed up source they are trained from, in the same way someone is using AI as a "tool" to fabricate their curriculum and cheat a job

> The intent isn't to disclose that AI was used in creating the code, the intent is to advertise the AI product

That is a wild mental gymnastic to justify dishonestly submitting code you didn't write (or own) as yours. It has nothing to do with advertisement but proper attribution, you know it.

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#596

Post by bandrami »

Pxtl wrote: post #570
bandrami wrote: post #279

Earlier quoted context omitted.

But you will also fail to keep the comments and code synchronized, and the comment will at some point no longer describe why the code is doing whatever it does

Which is why you're reviewing changes. I haven't memorized what every line of code does, if it was worth commenting then it was confusing-enough that it needed the comment and so I'll read the comment to make sense of the code being changed. If I don't read the comment that means the comment was too far from the confusing code. Alternately, you can say the same about informative variable names or informative function…

Yeah, my point has basically nothing to do with AI and is the argument against comment blocks in general. It's bad to store information in two places.

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#597

Post by Eisenstein »

VectorVault wrote: post #560
Eisenstein wrote: post #555

Earlier quoted context omitted.

I think that may not be answerable until a case concerning it has been heard and ruled on. A lawyer may have a better answer for you, but if I had to bet then I'd put $100 on it being something like 'it depends'.

It's interesting how AI can be its own worst enemy in this legal system. The very thing it's excellent at is not protected. In practice, there seems to be a strong opportunity to disintermediate brands by acting as a layer of abstraction above the seller and manufacturer. An AI instruction likely cares less about brand or sharing customer information with the seller; it's just more friction and tokens spent.

I think its just a case of dealing with something that has no precedent. We have never had to determine what the line is between a tool and an employee when they can both be instructed with natural language. If we were to evaluate AI as if it were in a contract with us for use of its time and efforts in exchange for something of consideration, it would be an easy ruling. If we were to evaluate AI as if it were a tool which operates as an extension of the operators skill without any independent additions then it would be an easy ruling. But since we now have a tool that can produce results that are independent of our ability to produce them with any former class of tools, then we have to create entirely new models for how to map these tools into the complexity of real life conflicts where people have different goals and where we must decouple fairness from intentions.

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#598

Post by manwe150 »

d4rkp4ttern wrote: post #559

For me one of the most interesting aspects is how compaction works. It turns out compaction still preserves the full original pre-compaction conversation in the session jsonl file, and those are marked as "not to be sent to the API". Which means, even after compaction, if you think something was lost, you can tell CC to "look in the session log files to find details about what we did with XYZ". I knew this before the…

What is microcompaction? I didn’t realize there was any thing time based in CC, when I go eat dinner and come back it compacted while I was gone?

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#599

Post by shreyssh »

d4rkp4ttern wrote: post #559

For me one of the most interesting aspects is how compaction works. It turns out compaction still preserves the full original pre-compaction conversation in the session jsonl file, and those are marked as "not to be sent to the API". Which means, even after compaction, if you think something was lost, you can tell CC to "look in the session log files to find details about what we did with XYZ". I knew this before the…

[flagged]

Re: The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

#600

Post by shreyssh »

mzajc wrote: post #37

There are now several comments that (incorrectly?) interpret the undercover mode as only hiding internal information. Excerpts from the actual prompt[0]: NEVER include in commit messages or PR descriptions: - The phrase "Claude Code" or any mention that you are an AI - Co-Authored-By lines or any other attribution BAD (never write these): - 1-shotted by claude-opus-4-6 - Generated with Claude Code - Co-Authored-By: C…

[flagged]
Post reply on HN