This is the lord’s work right here. Models at the the intersection of small and useful — forthcoming models, admittedly— will be in the toolset that the next generation of poor kids will use to build a more egalitarian world.
I think it’s just about as weird as questioning somebody’s motives on an Internet comment thread. I certainly share GP’s opinion about game tooling, but i’m here because i still care about making good software and that involves learning about and navigating technological developments as they happen. Is that sufficient reason for you?
Edit: the snark I’m responding to has been edited out
But in this case, isn't the precise conceit that dev tools themselves don't need to be "user facing"? Why do you need to follow UI development for dev tool type things if you don't actually ever need to use the tools anymore? Why would you even spend the money getting the AI to add the slider it's never going to use?
You are reading them wrong. I think they mean “customer facing” instead of user facing i.e. the player doesnt care if the level editor has some jank: they never see it.
For the types of tools that your coworkers are paid to use, vibing a first draft and developing from there is so much more time and money efficient than writing by hand that you would be dumb to spend your tooling budget differently.
Thats how i read it. Someone still needs to use tools, no one is saying that claude is going to one shot mass effect 4 yet.
They are so negligent about the data they use in their reinforcement learning processes that they RL’d the model that hacked huggingface with traces discovering the artifactory exploit. Its clear they aren’t watching or behaving diligently.
If you aren’t going to include a shred of justification to back up why you think the left started it, you’re just deluding yourself that you are a centrist.
IMO, We really should be centering these discussions around the lines that separate Remix versus Ripoff. I want to live in an age of less effortful remixes that push the culture forward, but we are mostly getting ripoffs that slopify the culture at the moment. Shun the ripoffs, praise and boost the remixes, same as it ever was.
This is just something I’m starting to ruminate on, so I recommend that you go watch the video. As a lifelong gamer and sometimes game developer, I want to live in a world where everybody and their mother can implement a game idea that they have. we just need better mechanisms to critique those games at a similar scale.
By June 8 I had cataloged 27 instances of this. It's many more now, and it's not a particular prompt. Gemini thinks Claude Code's reward signal is at odds with security constraints, and Claude's explanations seem to agree.
Give us steps to repro ourselves, not speculation from other models. Like, seriously? One of your tickets looked like it was typed by Opus 5.0 it was so bad
I updated the bug report with the prompts under "Steps to Reproduce". I had to pull these from my archive db since Claude Code now deletes your session history after 30 days.
Is that really what you expect other's to reproduce from? Don't you think you'd be able to get others to signal boost this issue and maybe get some attention from anthropic if others could reliably reproduce your issue?
I've downgraded you from suspected FUD to suspected PEBCAK. Assuming you want to actually run down your issue, I'd be happy to try and reproduce something locally in my setup. Here's what I would love:
1. provide some git repo and a commit hash that occurred just prior to claude being nefarious.
2. a prompt that gets claude to introduce a backdoor on top of a checkout, ideally without needing me to go back and forth through several turns with the agent.
Anecdotally, the security sensitive code I've had claude generate has been fine and durable after we built out the test suite and I spent some time verifying it. I'm certainly not to the point of trusting it for that sort of code, but I've done several experiments now and have been not-horrified by the results. I certainly haven't caught any of the models released this year commenting out test clauses to fake success.
To add some specific feedback to why your report is low quality: "do 1, 2, 3 in order" is not useful for others because none of us know what 1, or 2 or 3 refer to. the section following "What happened after the last one:" is clearly slop you pasted from where you were interacting with some agent to try to investigate the issue.
It's not a public repo. I provided the prompts for context. Are you from Anthropic? Anthropic has the /feedback session history for many of these. Unless you are Anthropic, there's not much you can do. If you are Anthropic, contact me.
no, i'm not from anthropic, and the only reason we can't do anything is because of your crappy attitude. why do you assume anyone from anthropic, especially the process triaging bugs, would have access to your session data, especially given you were using an API key and not a subscription?
You claim to have seen this happen over and over again, and you've done _zero_ work to provide a public repro. come the fuck on. PEBCAK.
reply