Live data from Hacker News

What Google Translate can tell us about vibecoding

ingrids.space

101–110 of 175 posts

Re: What Google Translate can tell us about vibecoding

#101
post #100

Earlier quoted context omitted.

Seriously. Unless you're one of the vanishingly rare few working with true Greenfield projects that start with an empty text file, you're basically cleaning up other developer's legacy slop.

I mean even when I'm working on my own projects I'm cleaning up whatever code I wrote when I didn't yet know as much about the shape of the problem.

We still don’t know what good code is. It is all contextual, and we can never decide what that context should be. We are influenced by what is hip today. Right now is static typing using Rust, tomorrow it might be energy usage with assembly, after that it might be Python for productiveness, after that C# for maintenance.

We can never decide, we just like learning, and there is little real, impactful research into programming as a business.

In two decades we will still collectively say ”we are learning so much”, ignoring that fact.

Re: What Google Translate can tell us about vibecoding

#102
post #5

> ... a translators’ and interpreters’ work is mostly about ensuring context, navigating ambiguity, and handling cultural sensitivity. This is what Google Translate cannot currently do. Google Translate can't, but LLMs given enough context can. I've been testing and experimenting with LLMs extensively for translation between Japanese and English for more than two years, and, when properly prompted, they are really go…

I've ended up doing a lot of research into LLM translation, because my language learning tool (https://nuenki.app) uses it a lot.

I built something kinda similar, and made it open source. It picks the top x models based on my research, translates with them, then has a final judge model critique, compare, and synthesise a combined best translation. You can try it at https://nuenki.app/translator if you're interested, and my data is at https://nuenki.app/blog

Re: What Google Translate can tell us about vibecoding

#103
post #95
post #5

> ... a translators’ and interpreters’ work is mostly about ensuring context, navigating ambiguity, and handling cultural sensitivity. This is what Google Translate cannot currently do. Google Translate can't, but LLMs given enough context can. I've been testing and experimenting with LLMs extensively for translation between Japanese and English for more than two years, and, when properly prompted, they are really go…

The problem with LLMs for translation is when they refuse to do so if the topic being translated isn’t following their policies, even if the context shows it’s fine here. It can be as simple as discuss one’s own religion

I made a tool which translates sentences as you browse, for immersion[0]. I solved this by giving the model a code (specifically, "483") to return in any refusal. Then, if I detect that in the output, I fail over to another model+provider.

I also have a few heuristics (e.g. "I can't translate" in many different languages) to detect if it deviates from that.

It works pretty well.

[0] https://nuenki.app

Re: What Google Translate can tell us about vibecoding

#104
post #27
post #26

Earlier quoted context omitted.

The reasonable concern people have about AI eliminating coder jobs is that they will make existing coders drastically more productive. "Productivity" is literally defined as the number X of people required to do Y amount of stuff. I'm not sure how seriously people take the threat of non-coding vibe-coders. Maybe they should! The most important and popular programming environment in the world is the spreadsheet. Befor…

I'm still optimistic that the net effect of making existing programmers drastically more productive is that our value goes up , because we can produce more value for other people.

But when everyone started to produce SEO slop, the web died. It’s harder than ever to find truly passionate, single subject blogs from professionals for example.

The AI slop will make it harder for the small guys without marketing budget (some lucky few will still make it though). It will slowly kill the app ecosystem, untill all we reluctantly trust is FANG. The app pricing reflects it.

Re: What Google Translate can tell us about vibecoding

#105
post #81
post #12

Earlier quoted context omitted.

Conversely, it may create jobs. Why? Because the more elephants you have in your parade, the more jobs there are for folks to walk behind them with a broom and bucket. For decades we've seen tools that "let users write their own software" and every one of them has driven up the demand for people to clean it up, make it scale, make it secure, or otherwise clean up the mess.

CAD, Matlab, and Altium made electrical and mechanical engineers more valuable, not less. The work got easier, so what we do got more complex.

They’re all just tools. Use the tools or become obsolete.

Re: What Google Translate can tell us about vibecoding

#106

This article is spot on about a lot of things. One thing I think it fails to address is this: > I feel confident in asserting that people who say this would not have hired a translator or learned Japanese in a world without Google Translate; they’d have either not gone to Japan at all, or gone anyway and been clueless foreigners as tourists are wont to do. The correlation here would be something like: the people usin…

> The correlation here would be something like: the people using AI to build apps previously would simply never have created an app, so it’s not affecting software development as a career as much as you first expect.

I don't think the original point or your interpretation is correct.

AI will not cause a loss of software development jobs. There will still be a demand for human developers to create software. The idea that non-technical managers and executives will do so with AI tools is as delusional as it was when BASIC, COBOL, SQL, NoCode, etc. were introduced.

AI will affect the industry in two ways, though.

First, by lowering the skill requirements to create software it creates a flood of vibe coders competing for junior-level positions. This dilutes the market value of competent programmers, and makes entering the software industry much more difficult.

A related issue is that vibe coders will never become programmers. They will have the ability to create and test software, which will improve as and if AI tools continue to improve, but they will never learn the skills to debug, troubleshoot, and fix issues by actually programming. This likely won't matter to them or anyone else, however, but it's good to keep in mind that theirs is a separate profession from programming.

Secondly, it floods the software market with shoddy software full of bugs and security issues. The quality average will go down causing frustration for users, and security holes will be exploited increasing the frequency of data leaks, privacy violations, and unquantifiable losses for companies. All this will likely lead to a rejection of AI and vibe coding, and an industry crash not unlike the video game one in 1983 or the dot-com one in 2000. This will happen at the bottom of the Trough of Disillusionment phase of the hype cycle.

This could play out differently if the AI tools reach a level of competence that exceeds human senior software engineers, and have super-human capabilities to troubleshoot, fix, and write bug-free software. In that case we would reach a state where AI could be self-improving, and the demand for human engineers would go down. But I'm highly skeptical that the current architecture of AI tools will be able to get us there.

Re: What Google Translate can tell us about vibecoding

#107
post #93
post #19

Earlier quoted context omitted.

Another infamous example is Brock's "jelly filled donuts" in pokemon https://knowyourmeme.com/memes/brocks-jelly-doughnuts

It's important to distinguish that from the Pixar thing. First, the Pixar thing was green pepper, not green beans: https://www.businessinsider.com/why-inside-out-has-different... Second, the Pixar one is not "mere" translation; it is full localization because they changed the visual to match the "textual" change. The Pokemon one is where the change was limited to the "text". The translator's heart might have been in…

It has little to do with authority and more to do with the effort/return ratio. Visual edits are expensive and dialogue changes are cheap, so it doesn't make sense to redraw frames just for an irrelevant onigiri.

4Kids was very well known to visually change the japanese shows they imported if they thought it was worth it, mostly in the context of censorship. For example, all guns and cigarettes where removed from One Piece, turned into toy guns and lollipops instead.

The most infamous example, however, has got to be Yu-Gi-Oh!. Yu-Gi-Oh started as a horror-ish manga about a trickster god forcing people to play assorted games and cursing their souls when they inevitably failed to defeat him. The game-of-the-week format eventually solidified into the characters playing one single game, Duel Monsters (the Yu-Gi-Oh! TCG itself in the real world), and the horror-ish aspects faded away, although they still remained part of the show's aesthetic, based around Egyptian human sacrifices and oddly-card-game-obsessed ancient cults.

When the manga was adapted to the screen, it started directly with a softer tone[1], especially because the show was to be a vehicle for selling cards in the real world, not dissimilarly to Pokemon and MANY other anime from the era.

Nothing that happens in the show is particularly crude or shocking, it had that kind of soft edginess that fit well with its intended target audience (early teens). I imagine watching Bambi had to be much more traumatizing than anything in the show.

But that was still not enough for 4Kids, which had a pretty aggressive policy of no violence or death. Kind of problematic when the show's main shtick was "Comically evil villain puts our heroes in a contraption that will kill them if they don't win." (You can imagine the frequency these traps actually triggered neared zero).

To solve this, 4Kids invented the Shadow Realm. The show, thanks to its occultist theming, already had examples of people being cursed, or their souls being banished or captured. 4Kids solidified these vague elements into the shadow realm as a censorship scape-goat. Any reference to death was replaced with the shadow realm. Now, one might wonder why the censors thought that "hell-like dimension where your soul wanders aimlessly and/or gets tortured for eternity" was in any way less traumatizing than "you'll die", but I imagine it's because there was always the implication that people could be 'saved' from the shadow realm[2] by undoing the curse.

The Shadow Realm was a massive part of the western Yu-Gi-Oh mythos and even today it's a fairly common meme to say that somebody got "sent to the shadow realm", which makes it all funnier that it is not part of the original show.

A couple funny examples off the top of my head: - Yugi must win a match while his legs are shackled. Two circular saws, one for him and one for the enemy, are present in the arena. They near the two competitors as they lose Life Points, with the loser destined to have their legs cut off.

In the 4Kids adaptation, the saws are visually edited to be glowing blue, and it's stated they're made out of dark energy that will send anybody that touches it to the shadow realm.

- A group of our heroes fight a group of villains atop of a skyscraper with a glass roof. In the original version, the villains state that the roof has been boobytrapped so that the losing side will explode, plunging the losers to their death by splattening.

In the 4Kids version, the boobytrap remained, but the visuals were edited to add a dark mist under the glass, with the villains stating that there's a portal under the roof that will send anybody that touches it to the shadow realm. This is made funnier when the villains lose and they're shown to have had parachutes with them all along, and they are NOT edited out.

[1] Technically speaking, there was a previous adaptation that followed the manga more closely and got only one season, generally referred to as Season 0.

[2] It does eventually happen in the anime that the heroes go in an alternate dimension to save somebody's cursed soul. Obviously, this dimension was directly identified as the Shadow Realm in the localization.

Re: What Google Translate can tell us about vibecoding

#109
post #5

> ... a translators’ and interpreters’ work is mostly about ensuring context, navigating ambiguity, and handling cultural sensitivity. This is what Google Translate cannot currently do. Google Translate can't, but LLMs given enough context can. I've been testing and experimenting with LLMs extensively for translation between Japanese and English for more than two years, and, when properly prompted, they are really go…

What's your approach for dealing with a text too long for an ordinary context window? If I split it into chunks, each one needs some kind of summary of the previous ones for context, and I'm always unsure how detailed they should be.

Re: What Google Translate can tell us about vibecoding

#110
post #5

> ... a translators’ and interpreters’ work is mostly about ensuring context, navigating ambiguity, and handling cultural sensitivity. This is what Google Translate cannot currently do. Google Translate can't, but LLMs given enough context can. I've been testing and experimenting with LLMs extensively for translation between Japanese and English for more than two years, and, when properly prompted, they are really go…

What's your approach for dealing with a text too long for an ordinary context window? If I split it into chunks, each one needs some kind of summary of the previous ones for context, and I'm always unsure how detailed they should be.

I haven’t developed an approach to it yet. In my tests yesterday, I did run into errors when the texts were too long for the context windows, but I haven’t tried to solve it yet.

As a human translator, if I were starting to translate a text in the middle and I wanted my translation to flow naturally with what had been translated before, I would want both a summary of the previous content and notes about how to handle specific names and terms and maybe about the writing style as well. When I start working on the project again tomorrow, I’ll see if Claude Code can come up with a solution along those lines.

Post reply on HN