"Don't You Just Upload It to ChatGPT?" • Hacker News Board

Re: "Don't You Just Upload It to ChatGPT?"

#181

Post by svara » Fri, Jun 12, 2026, 9:32 PM UTC

tombert wrote: post #58

Wowfunhappy wrote: post #36

Earlier quoted context omitted.

> Out of curiosity, I sic'd Claude Fable on the original French version of The Three Musketeers and told it to translate accurately, but also try and keep the same jovial tone as the original and do not censor anything. After it was done, I didn't read the entire output, but I did compare a few individual chapters between the Ellsworth translation and the Fable translation. This isn’t a great test, because Claude alm…

Read the last two paragraphs :)

The things is, this is almost certainly what's happening.

You can (could, maybe they 'fixed' it by now) get sota LLMs to reproduce entire novels near verbatim.

The idea of giving it parallel texts of those novels in different languages, to train it on translation, is so obvious it'd just be strange if the AI labs didn't do it.

In fact DeepL was doing basically that more than 10 y ago.

Re: "Don't You Just Upload It to ChatGPT?"

#182

Post by jaggederest » Fri, Jun 12, 2026, 9:33 PM UTC

atleastoptimal wrote: post #154

You'd be laughed at if you said that ChatGPT could help you with graduate level mathematics in 2024, but this year, AI models on simple prompts are solving previously unsolved Erdos problems. It seems silly to imagine that there is some fundamental barrier between human intelligence and AI, and that AI could never do many of the things that humans can do. Inferring intent, gauging sentiments, factoring in cultural va…

Fable has really spooked me, honestly. It's another big jump, but not in the actual coding. I was pretty comfortable with the "you do the implementation, I do the meta work and steering", and ... no steering required, no meta work required. Here's the backlog, let me know when it's complete, I guess I'm going to go touch grass until I have to review and refine... probably tomorrow?

Reminds me of the first time I saw a coding agent stumble through an issue in 2023 maybe? and went "this is a big deal", similarly when OG gpt started making jokes that actually kinda worked.

Updated modern version of the classic "make me a greentext", apologies for slop-posting, but it seems relevant:

    > be me
    > senior software engineer
    > in charge of making sure the tickets get, in fact, implemented
    > occasionally have to open the IDE and write some code myself
    > one day i open the IDE and the ticket is already closed
    > the agent did it overnight
    > no steering, no review notes, nothing left for me to do
    > distress.jpg
    > ask my manager what to do
    > he says "just focus on the high-level architecture stuff"
    > i say "what high-level architecture stuff"
    > he says "i don't know, you're the senior engineer"
    > rage.jpg
    > quit my job
    > become a prompt engineer, nice and simple, just tell it what to build
    > first day on the job, sit down to write the prompt
    > AI already wrote it

Re: "Don't You Just Upload It to ChatGPT?"

#183

Post by cptroot » Fri, Jun 12, 2026, 9:33 PM UTC

atleastoptimal wrote: post #154

You'd be laughed at if you said that ChatGPT could help you with graduate level mathematics in 2024, but this year, AI models on simple prompts are solving previously unsolved Erdos problems. It seems silly to imagine that there is some fundamental barrier between human intelligence and AI, and that AI could never do many of the things that humans can do. Inferring intent, gauging sentiments, factoring in cultural va…

> AI can currently do if given enough context

It's worth noting that you can substitute "dollars" for "context" in that sentence, which seems to be where many of these impressive achievements are coming from. As ever, it's unclear whether these models will get cheaper while remaining better, since all of the recent breakthroughs appear to be of the "think more" kind. For translation specifically, I'd be very surprised if the "think more" LLMs would help given the per-unit cost expected of the output.

Re: "Don't You Just Upload It to ChatGPT?"

#184

Post by Ancapistani » Fri, Jun 12, 2026, 9:36 PM UTC

pixel_popping wrote: post #96

Johnbot wrote: post #16

Earlier quoted context omitted.

This is anecdata, but in my experience with myself and my coworkers, it is not that we believe the AI will be more accurate in software engineering, but that the answer will come faster and be more tailored to our exact problems. If I have to search SO, I have to find the answer and then tweak it to fit my codebase, but with AI tooling, the AI is already basing its answer around my code.

I think we actually do believe it, do you believe Fable 5+GPT-5.5 (+ the whole model zoo) in loop with adversarial (no budget limit) or a 10-year experienced SWE? We are talking about "codebases" but realistically we won't even be checking the filetree of them soon, it will be all blind, containerized and verified with pseudo guarantees which are good enough to build serious things. We don't even write documentation…

This is the direction I'm going.

For personal projects that I don't plan to share widely, I'm making it a point to not look at the code at all. So far - and to my surprise - I've not only found that this has result in no more bugs than before, but it seems to result in fewer bugs over time. Every time I find a bug or a regression, I add it to the specification. My SDLC requires that every specification have at least one associated test. Not every function, or every line, or anything like that - every specified feature. The end result has been that my projects have matured over time much faster than if I'd been more closely involved.

I've already toyed with writing some projects in Nim and Haskell for token efficiency. At some point I plan to put together a simple test project, then do a comparison of token efficiency with every language I can think of to find the one that I'm able to generate most quickly, correctly, and cheaply.

Re: "Don't You Just Upload It to ChatGPT?"

#185

Post by no_multitudes » Fri, Jun 12, 2026, 9:36 PM UTC

tombert wrote: post #9

I have no doubt that the writer is better at translating than AI, but I have to say that AI translation has gotten so good that I'm not sure how much longer translation work will be there, or rather it might end up being more about auditing. For example, I just read the Lawrence Ellsworth translation of The Three Musketeers, which I very thoroughly enjoyed. I don't speak or read French, but from my understanding Ells…

> Fable might have been trained using Ellsworth's translation and as such it's very directly able to crib from it

The `cp` program on my computer also has the remarkable ability to produce a faithful translation of The Three Musketeers when provided one as input.

Re: "Don't You Just Upload It to ChatGPT?"

#186

Post by acyou » Fri, Jun 12, 2026, 9:40 PM UTC

"we all more or less look the same in gym clothes"

Maybe my brain works differently than the author, but I'm surprised at this statement. Gym clothes don't change recognition for me, it's about the face, body, posture, clothes don't really enter into it. For me it is nonsensical enough to be suspicious.

And for a human centric perspective, not recognizing who someone is sad, it's knowing that you probably won't meet them again so it's not worth it, the community isn't there. Where community and interpersonal relationships between people are something we still hold dearly.

Re: "Don't You Just Upload It to ChatGPT?"

#187

Post by Ancapistani » Fri, Jun 12, 2026, 9:40 PM UTC

bigstrat2003 wrote: post #23

pixel_popping wrote: post #2

I agree with the take, but it's a temporary one, the sad reality is that we will be literally inferior soon, there will be a point where we will not trust human input without counter check by AI, we need to remember that we are kinda at the beginning of the AI era, in 5 to 10 years it's very unlikely that a human translator or software engineers will do better than the tooling we will have. There is already a tipping…

> there will be a point where we will not trust human input without counter check by AI That's nonsense. There is zero reason to believe that AI (with the current techniques) will ever become reliable enough to let it do its own thing, let alone better than a human. It's been years of development and you still can't trust it to get basic facts correct, not even "well it's better than it used to be". Saying it'll repl…

> It's been years of development and you still can't trust it to get basic facts correct

There's the rub: AI is not an oracle. It's neither designed nor intended to provide accurate recall of all facts. It's closer to a reasoning engine than anything IMO.

Oh, and for the record: I don't trust people to get basic facts correct, either. It's already far better than the average human at trivia.

Re: "Don't You Just Upload It to ChatGPT?"

#188

Post by bryanrasmussen » Fri, Jun 12, 2026, 9:40 PM UTC

FeteCommuniste wrote: post #180

bryanrasmussen wrote: post #169

Earlier quoted context omitted.

I remember there were some studies that this kind of thing was effective a year or so ago, so essentially a lifetime in Model years. However to me it seems completely reasonable that it would work, because my understanding of what happens is the model interprets what you said as: Look for a group of people who are considered to be expert growth hackers by the world at large and answer my questions as though they were…

I imagined it as kind of a shorthand for "you should be spending my tokens on looking for / addressing issues like X, Y, and Z," where X, Y, and Z are the sorts of things that an expert in [insert domain here] would be likely to care most about.

right, but the thing is how do they know what an expect in [insert domain here] would care about? Obviously by finding content created by

people who claim to be experts in [domain] people who others claim to be experts in [domain]

hopefully valuing membership in group two over membership in group 1.

Re: "Don't You Just Upload It to ChatGPT?"

#189

Post by while_true_ » Fri, Jun 12, 2026, 9:41 PM UTC

atleastoptimal wrote: post #154

You'd be laughed at if you said that ChatGPT could help you with graduate level mathematics in 2024, but this year, AI models on simple prompts are solving previously unsolved Erdos problems. It seems silly to imagine that there is some fundamental barrier between human intelligence and AI, and that AI could never do many of the things that humans can do. Inferring intent, gauging sentiments, factoring in cultural va…

Yes. It's as if they think AI will forever be LLM only and won't develop world models that incorporate current state assessment, dynamic next-state prediction, cause-and-effect reasoning, object permanence, etc. I'm not in the AI industry but I assume there's got to be lots of research and work being done on this.

Re: "Don't You Just Upload It to ChatGPT?"

#190

Post by layer8 » Fri, Jun 12, 2026, 9:41 PM UTC

robertnowell wrote: post #144

layer8 wrote: post #4

What’s unfortunate is that the market that is willing to pay for high-quality human translation has shrunken considerably.

if it was valuable, people would pay for it

That’s not how it works. Value for users doesn’t translate 1:1 into value for businesses, nor are either necessarily willing to pay for value. That’s why things enshittify.