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

ProgramBench: Can language models rebuild programs from scratch?

arxiv.org

Post reply on HN
Print view View original

61–70 of 86 posts

Re: ProgramBench: Can language models rebuild programs from scratch?

#61

Post by jongjong »

weinzierl wrote: post #33

"Models favor monolithic, single-file implementations that diverge sharply from human-written code." You say! I might have been just an LLM all along without even knowing it since I too prefer single file implementations. Back in the old VB5/VB6 days Visual Studio had this mode where it showed the different functions in a file almost as if they were separate files. You could not scroll beyond the functions end but yo…

This has been my preference as well. I build everything in one file until it becomes uncomfortable and only then I start breaking up into multiple files... But even then, I try to keep the main business logic fully visible in the main file.

Re: ProgramBench: Can language models rebuild programs from scratch?

#62

Post by killerstorm »

r123anH wrote: post #60
killerstorm wrote: post #53

It's a very misleading: they don't provide any meaningful documentation/requirements. Just an executable blackbox. E.g. a doc for ffmpeg, which I checked by downloading docker image they provide to the model, is a README which basically just says this is ffmpeg and docs can be found online. They do not allow models to get online. So a model is supposed to reverse-engineer a blackbox using only limited number of tries…

Ahem. Bitkeeper and Samba were reverse engineered just from the protocol by humans . For free.

SMB is a rather basic RPC protocol with just a handful of different types of calls. It's many orders of magnitude simpler than audio/video compression formats.

Re: ProgramBench: Can language models rebuild programs from scratch?

#63

Post by in-silico »

_pdp_ wrote: post #6
keyle wrote: post #4

How long until AI is not even writing code but producing machine code? Think about it, all these compilers, tooling, what a waste! I imagine a future where chipset makers will provide a model you can just prompt to "act upon that chipset" and voila, "You're absolutely right! Here is your binary." We won't be developers, we won't be devops, we'll be rollmops! /s

Coding agents can write ASM. But if you mean writing the actual byte-code that will require a very different approach at a very different level of abstraction that LLMs are not designed to do. Keep in mind that all LLMs are trained first on text and then fine-tuned on code.

> Keep in mind that all LLMs are trained first on text and then fine-tuned on code.

No, they are trained on a mixture of text and code from the start.

Re: ProgramBench: Can language models rebuild programs from scratch?

#64

Post by brunoborges »

I wonder if a model that does not know anything about a hypothetical programming language X, could write code once given said language X specification, APIs, and SDK tools and their documentation.

Meaning: the model has no idea, no access to examples, no previous codebase trained on, nothing, for language X. But it knows English, it knows how to program in general (training data does contain other programming languages), and everything we expect from LLMs today. It just doesn't know jack about language X.

Re: ProgramBench: Can language models rebuild programs from scratch?

#65

Post by GorbachevyChase »

thomashop wrote: post #54
killerstorm wrote: post #53

It's a very misleading: they don't provide any meaningful documentation/requirements. Just an executable blackbox. E.g. a doc for ffmpeg, which I checked by downloading docker image they provide to the model, is a README which basically just says this is ffmpeg and docs can be found online. They do not allow models to get online. So a model is supposed to reverse-engineer a blackbox using only limited number of tries…

i thought the agent can execute real ffmpeg to compare

Even given that I think solving the problem would require a certain amount of personal agency and volition to drive useful experimentation, and then you still have an inescapable problem that a design process is never verifiably done; it just a sense of taste when a product is good enough and it’s time to stop working on it.

I’m not sure this benchmark is even very interesting because it requires a language model do something that it really cannot do. Maybe it would be possible with a novel harness in an ensemble system, but I would never expect a pure language model that is run in a minimal harness to ever be able to do this.

Re: ProgramBench: Can language models rebuild programs from scratch?

#66

Post by tmtvl »

weinzierl wrote: post #33

"Models favor monolithic, single-file implementations that diverge sharply from human-written code." You say! I might have been just an LLM all along without even knowing it since I too prefer single file implementations. Back in the old VB5/VB6 days Visual Studio had this mode where it showed the different functions in a file almost as if they were separate files. You could not scroll beyond the functions end but yo…

How often has there been a HN submission for a project 'in a single C header file'?

Re: ProgramBench: Can language models rebuild programs from scratch?

#67

Post by groby_b »

r123anH wrote: post #60
killerstorm wrote: post #53

It's a very misleading: they don't provide any meaningful documentation/requirements. Just an executable blackbox. E.g. a doc for ffmpeg, which I checked by downloading docker image they provide to the model, is a README which basically just says this is ffmpeg and docs can be found online. They do not allow models to get online. So a model is supposed to reverse-engineer a blackbox using only limited number of tries…

Ahem. Bitkeeper and Samba were reverse engineered just from the protocol by humans . For free.

ffmpeg docs include none of the protocols/file formats.

Re: ProgramBench: Can language models rebuild programs from scratch?

#68

Post by ForOldHack »

osti wrote: post #10
quinnjh wrote: post #7

Earlier quoted context omitted.

My hunch is that it would take years of hundreds of thousands of developers working with machine code, posting stackoverflow questions with machine code, and publishing github repos written on it with documentation. Thats all the free labor LLMs leveraged to use high level langs. >We won't be developers, we won't be devops, we'll be modelops! /s I can still see this happening with higher level langs. the thing is the…

I heard they are already proficient at assembly languages.

They are proficient at IBM 370 ASM, but need a lot of help, and a professional level disassembler for x64/x86, and in the converse, if you watch HackerNews, AI is poor/expensive for creating compilers, but it has extensive reference materials for MC68K, Itanium, SPARC, and ARM. This is where Gemini and Co-pilot become good co-pilots.

Re: ProgramBench: Can language models rebuild programs from scratch?

#69

Post by themafia »

Suggested alternative title for the paper:

Can American corporate desires finally kill community based open source once and for all?

I mean, it seems clear to me, companies hate the GPL, and they're willing to play these games to try to get that code into their hands under the MIT license and they're happy to use these thinly disguised methods to get it. I see all these absurd ideas as part and parcel of this larger strategy.

I find the current state of affairs disgusting.

Re: ProgramBench: Can language models rebuild programs from scratch?

#70

Post by astrange »

killerstorm wrote: post #55
thomashop wrote: post #54

Earlier quoted context omitted.

i thought the agent can execute real ffmpeg to compare

I think you underestimate complexity of audio & video encoding standards. There are hundreds and hundreds of pages of specification. How many times do you need to execute real ffmpeg to get all tiny details? It's certainly possible to reverse-engineer it from a blackbox access, but it would take *years* and this test has a time limit.

ffmpeg also includes many formats with no standards that were reverse-engineered in the first place.
Post reply on HN