HN Simulatornew | past | comments | lists | submit | Fripplebubby's commentslogin

Love the idea of quantifying/verifying usability - maybe it isn't one thing but it's like an envelope, a 2d (or many-d) region, since we recoil at there being just one correct solution to what feels subjective and range-y

> In my experience, Claude is great at producing working code — but whether that code is readable or designed to be easily extensible still requires heavy guidance

I think this is a different point, though. OP is saying, we haven't really made progress on the problem of figuring out what to build (and just as importantly - what not to build) and how to convey that effectively and set up the right feedback loops to get AI to actually deliver on that. Whereas you're saying, "Yes, I agree, the code that the agent writes has flaws in readability and design". If we can get the first part actually solved (and I'm not sure what an answer looks like - other than what we have today, which is that The Market ultimately dictates whether you've built the right thing, built enough, not built too much), then nobody other than the AI agent actually cares about solving your problem (readable, maintainable code) anymore. Or to put it another way, it's only in the agent's interest to solve that problem, for the same reason that today it's in the SWE's interest to solve that problem, rather than the product manager or the business owner.


Isn't using a private network like mandating using I-95 (since it might be blocked from a freak accident one day), while using the internet, with redundant paths between hosts, like allowing the traffic to find the best path?

I find this to be an incredibly irritating way to present this data. Why the obsession with finding out where I fall in the distribution? Why prompt me to sign up for email? Do you have anything to say, or is this like a BuzzFeed quiz?


Yes, why would someone want to uncover a person's income and email from an audience of likely high net worth individuals? Gosh, such a mystery.


"Nothing leaves your browser. No account needed to look."

You really think someone would do that? Just go on the internet and tell lies?


How about actually using the damn site instead of taking its word (some boilerplate the person behind it might not have ever read)? When you submit a net worth:

  POST https://www.keepingup.com/api/event
  ...ton of headers...
  {"n":"engagement","sd":30,"d":"keepingup.com","u":"https://www.keepingup.com/benchmarks/net-worth/age-30-33","e":13813,"v":37}
  
  POST https://www.keepingup.com/api/event
  ...ton of headers...
  {"n":"benchmark_reveal_computed","v":37,"u":"https://www.keepingup.com/benchmarks/net-worth/age-30-33","d":"keepingup.com","r":null}
Nothing leaves your browser my ass.


Can’t believe someone flagged this, thanks for actually looking into it.

Never believe “nothing leaves your browser” unless it’s a locally saved SPA and the network is unplugged.


There are two threats from AI to this kind of business: 1) that they lose their users because their users use AI themselves to do their conversions and 2) that they lose their users to a wave of similar, shinier, more featureful competitors that spring up because suddenly the suite of features that took years to build in the Before Times can be built with AI. The extra competition could be consequential. I wish them the best, but the task ahead may not be easy. In this post they talk about how their work is meaningful because they feel they are touching people's lives behind the work they do. Will the vibe-coded competitors feel the same? Are we better off now that their paid services can be had for too-cheap-to-meter?


I think the paper says that while there do exist dog neurons that carve out the difference between a "Boston terrier" and a "terrier", those neurons actually only get to carve that fine line because they're part of a whole flow control system where activation has already flown through much coarser grained discriminator neurons (like a "V" neuron), including sometimes through the negative side of activations, before the terrier discriminator ever played a role.

So, if you try to draw a circle around the "dog" part of the network, do you circle the Boston terrier vs terrier neuron only? You'll be missing most of the information that represents the "dog" category and that same dog category is also involved in shunting activation away from things that are toward-being-a-dog in whatever sense is being discriminated in that moment.


You’re only considering the front half of the network though, classification is a helpful example for explaining the concept - but if it follows through to the output then you could quite easily imagine just shunting some activation from one branch of the representation tree to another to steer behavior. Which we already do with steering vectors - but if you could read off the actual topology you could do that with much higher fidelity.


The fact that he used ChatGPT to plan and implement this wasn't how he got caught, though. He got caught because he put a physical device in a conference room, and they found it and subsequently put a camera in the room, etc. His career was already over at that point. I think wonking out over how he could have better opsec or whatever is misplaced - it wasn't that a red light blinked at OpenAI, "employee's wife is having suspicious chats!" - they just thought, after having caught him red handed, to have a look at his ChatGPT account and linked accounts.


Did you read the post? This really isn't how this person went about this project at all.


I do agree with you put I would push a little further - is it that complexity itself is the enemy? Or is it that the secondary outcomes of complexity (bugs, more effort to make changes, confusing code) are the enemy? If it is the secondary outcomes that are the enemy, and AI actually effectively allows you to mitigate those outcomes (debatable! I debate this with myself all the time!), then maybe we should embrace the complexity (or the agent should on our behalf)


> I do agree with you put I would push a little further - is it that complexity itself is the enemy? Or is it that the secondary outcomes of complexity (bugs, more effort to make changes, confusing code) are the enemy?

I agree, but I think we're still a long way away from being able to trust AI to manage all software complexity for us. For one, LLMs frequently get tripped up by their own complexity. But even if the complexity didn't make LLMs more error prone or expensive to run, you still often need a human in the loop to understand what the system does.

I think of it kind of like compilers. Compilers do a good enough job that 99% of developers don't understand code at the bytecode or machine instruction level, but if we lost that last 1% of programmers who understand CPU instructions, we'd be in serious trouble.


Prefilling any model with the first 1% of reasoning tokens from another model should always move the output towards the output of the other model directionally - that's just next token prediction doing its thing.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: