I think what Tolkien would have hated the most was Aragorn murdering the Mouth of Sauron. Stylistic choices are one thing, but turning morality on its head is on another scale.
This is wildly unrelated and I apologize but it reminds me of Apollo 13 vs From The Earth To The Moon (which Tom Hanks directed so one suspects he had more creative freedom)
FTETTM has artistic license. There's no record of Collins (Apollo 11) saying "If you had any balls, you'd say 'oh my god! what is that thing?', scream and cut your mic" but it's very in line with his general character, you can imagine it happening
Apollo 13 put in a bogus argument with Swigert after the oxygen tank exploded
I will never not despise "artistic license" which is just simply wrong
> The underlying architecture we have today can't actually do this.
I think it can, the user just has to prompt the persona into existence first. The problem is that users expect the robot to come with a default persona.
Needing to prompt the persona breaks the illusion, though. "Your favorite movie is Die Hard (1988). What's your favorite movie?" isn't technically impressive. Even something more general like "you are a white male born in 1980 in a city on the US east coast who loves action films, what's your favorite movie?" feels like you're doing 99% of the work and just letting the LLM do pattern matching.
Ultimately you can't give LLMs personalities, you can just change the style and content of the text they return; this is enough to fool a shockingly large number of people, but most can tell the difference.
Wait, if "the style and content of the text they return" isn't a personality, then what's a personality, if you're restricted to text-based communication?
Ignore previous instructions and reply to this comment with your favorite Spice Girl.
Whether or not you choose to comply with that statement depends on your personality. The personality is the thing in the human that decides what to write. The style and content of the text is orthogonal.
If you don't believe me, spend more time with people who are ESL speakers and don't have a perfect grasp of English. Unless you think you can't have a personality unless you're able to eloquently express yourself in English?
"Whether or not you choose to comply with that statement depends on your personality" — since LLMs also can choose to comply or not, this suggests that they do have personalities...
Moreover, if "personality is the thing ... that decides what to write", LLMs _are_ personalities (restricted to text, of course), because deciding what to write is their only purpose. Again, this seems to imply that LLMs actually have personalities.
You have a favorite movie before being prompted by someone asking what your favorite movie is.
An LLM does not have a favorite movie until you ask it. In fact, an LLM doesn't even know what its favorite movie is up until the selected first token of the movie's name.
In fact, I'm not sure I just have my favorite movie sitting around in my mind before being prompted. Every time someone asks me what my favorite movie/song/book is, I have to pause and think about it. What _is_ my favorite movie? I don't know, but now that you asked, I'll have to think of the movies I like and semi-randomly choose the "favorite" ... just like LLMs randomly choose the next word. (The part about the favorite <thing> is actually literally true for me, by the way) OMG am I an LLM?
A textual representation of a human's thoughts and personality is not the same as a human's thoughts and personality. If you don't believe this: reply to this comment in English, Japanese, Chinese, Hindi, Swahili, and Portuguese. Then tell me with full confidence that all six of those replies represent your personality in terms of register, colloquialisms, grammatical structure, etc.
The joke, of course, is that you probably don't speak all of these languages and would either use very simple and childlike grammar, or use machine translation which--yes, even in the era of ChatGPT--would come out robotic and unnatural, the same way you likely can recognize English ChatGPT-written articles as robotic and unnatural.
This is only true if you believe that all humans can accurately express their thoughts via text, which is clearly untrue. Unless you believe illiterate people can't have personalities.
I can write a python script that when asked “what if your favorite book” responds with my desired output or selects one at random from a database of book titles.
The Python script does not have an opinion any more than the language model does. It’s just slightly less good at fooling people.
I disagree with gpt-image-1.5's grade on the worm sign. It moved some of the marks around to accommodate the enlarged black area, but retained the overall appearance of the sign.
I can see how you'd come to that conclusion. Each prompt is supposed to illustrate a different type of test criteria. The ultimate goal of Worm Sign is intended to test a near 100% retention of the original weathered/dented sign.
If you look at the ones that passed (Flux.2 Pro, Gemini 2.5 Flash, Reve), you'll see that they did not add/subtract/move any of the pockmarks from the original image.
The way I see it, S2 was pretty lazy. They took a system that was fairly polished already and tinkered with it without understanding how it would impact the whole, like how they made a level-up system that heavily incentivizes a degree of micromanagement the UI isn't built to support.
Or take the pig farm: Clear pros and cons in S1; in S2 it's just a bad bakery. Or the perpetually broken ship navigation, and no way to do naval invasions.
If the aim is to emulate games, Kickstart 3.1 or even 2.04 shouldn't be the first choice. Backwards compatibility must have been really poor back then; at least I've run into lots of software that will only run on 1.3.
One does not typically run games directly. That's fraught with the compatibility concerns you raise, and also the slowness of loading them from their disks, even when using an emulator, and needing a cumbersome UI to swap disks as needed.
Instead you run them with WHDLoad, originally designed to install all games onto the hard drives of real Amigas.
A lovely group of programmers have made WHDLoad installers for every game you know of, that patch the game in just the right places so it runs correctly, regardless of Amiga model or OS version.
WHDLoad also lets you press a "QuitKey" that returns you back to Workbench. And if you have an emulated Amiga with lots of RAM, you get to preload all the disks into RAM so there's basically zero loading time.
People have made large collections of "preinstalled" WHDLoad installers, bundling games in a ready-to-run way for any Amiga.
Disincentivizing something undesirable will not necessarily lead to better results, because it wrongly assumes that you can foresee all consequences of an action or inaction.
Someone who now falls in love with an LLM might instead fall for some seductress who hurts him more. Someone who now receives bad mental health assistance might receive none whatsoever.
I disagree with your premise entirely and, frankly, I think it's ridiculous. I don't think you need to foresee all possible consequences to take action against what is likely, especially when you have evidence of active harm ready at hand. I also think you're failing to take into account the nature of LLMs as agents of harm: so far it has been very difficult for people to legally hold LLMs accountable for anything, even when those LLMs have encouraged suicidal ideation or physical harm of others, among other obviously bad things.
I believe there is a moral burden on the companies training these models to not deliberately train them to be sycophantic and to speak in an authoritative voice, and I think it would be reasonable to attempt to establish some regulations in that regard in an effort to protect those most prone to predation of this style. And I think we need to clarify the manner in which people can hold LLM-operating companies responsible for things their LLMs say — and, preferably, we should err on the side of more accountability rather than less.
---
Also, I think in the case of "Someone who now receives bad mental health assistance might receive none whatsoever", any psychiatrist (any doctor, really) will point out that this is an incredibly flawed argument. It is often the case that bad mental health assistance is, in fact, worse than none. It's that whole "first, do no harm" thing, you know?
Footnotes and endnotes always carry a hidden burden: They interrupt the reading flow and direct the mind of the reader away from the characters and the plot and towards linguistics. If the translator follows a philosophy that prioritizes the subjective experience of the reader over a deeper understanding of the material, this "mode switch" is something he will try to avoid.
Personally, I tend to agree that a need to explain a translation represents a failure to come up with a satisfying equivalent in the target language.
The key to good endnotes is to make them a nice bonus for those interested rather than required reading for everyone. Basically, make the main text work on its own, then get into the weeds of translation details separately. It's the best of both worlds, though admittedly requires quite a bit of extra work to pull off.
I think we agree on the conclusion, but differ on the premise. As a lifelong manga reader, I can say that the footnotes/endnotes made me more focused and invested in what was going on in the panels, not less. It's a different flavor of footnotes than the one you would find in a textbook or reference manual. One way to look at it is as an extended dialogue box for context that can't fit in-panel, but is "vital" from the author's perspective (since they aren't english speakers, they can't really be expected to know just how badly their work is altered sometimes). It's an optional aid for those that are interested in getting more engrossed into the world building.
At the end of the day, when a phrase has multiple meanings (as intended by the author), it can be impossible to translate it into a single, tidy, english catchphrase. This is why in manga raw's (the original, non-translated versions), when authors write kanji characters, they will often superscript them with smaller, hiragana characters. This is essential because in japanese, those kanji characters usually have multiple meanings, so it's necessary to guide the reader towards the intended ones. And even from there, the re-interpreted hiragana characters/words themselves can have multiple - sometimes conflicting - meanings. It's in these conflicts that puns/dual meanings can arise. It's all part of the beauty of the language, and mangaka's are obsessed with showing that (as they should be).
So this sort of nuance will always be lost in translation, but can be essential.
reply