I am curious why OP has not explained why they dislike LLM. The conversation will be much different if it’s is a question of environment, IP theft, or slop (or all of the above)
3 Likes
This reminds me of an old thread from the pre-LLM age.
The circle is complete now. 
(It’s just a joke, don’t take it too seriously)
https://forums.online-go.com/t/im-leaving-from-ogs-community-this-is-why/
2 Likes
As with most power tools, if you don’t understand its proper use first, when you push the button you are likely to lose a finger. 
2 Likes
Nah this platitude is just wrong when it comes to more complicated technologies. Even setting aside Uberdude’s argument about the materials used in construction, some technologies are simply built in a way that systemically produces effects that can be evaluated morally.
The value of motor vehicles, for example, can be evaluated in terms of emissions, inherent hazards, and necessary infrastructure.
Tanks are one type of motor vehicle which are designed for armies creating a show of force in potentially hostile territory. Though ethical uses of tanks may exist, you cannot drive around without communicating some amount of violent and hostile intent.
More to the point, how about the Eleventh Edition of the Newspeak Dictionary that the Ministry of Truth is compiling in Orwell’s Nineteen Eighty-Four? Yes it does have the practical use of possibly defining a word for you, when you read something you need to look up. (After all if no one actually needed to use it, it wouldn’t serve its purpose very well, as no one would read it.) However it is engineered to have a certain, arguably malicious, social impact.
Description of the goals of the Newspeak Dictionary
‘The Eleventh Edition is the definitive edition,’ he said. ‘We’re getting the language into its final shape—the shape it’s going to have when nobody speaks anything else. When we’ve finished with it, people like you will have to learn it all over again. You think, I dare say, that our chief job is inventing new words. But not a bit of it! We’re destroying words—scores of them, hundreds of them, every day. We’re cutting the language down to the bone. The Eleventh Edition won’t contain a single word that will become obsolete before the year 2050.’
He bit hungrily into his bread and swallowed a couple of mouthfuls, then continued speaking, with a sort of pedant’s passion. His thin dark face had become animated, his eyes had lost their mocking expression and grown almost dreamy.
'It’s a beautiful thing, the destruction of words. Of course the great wastage is in the verbs and adjectives, but there are hundreds of nouns that can be got rid of as well. It isn’t only the synonyms; there are also the antonyms. After all, what justification is there for a word which is simply the opposite of some other word? A word contains its opposite in itself. Take “good”, for instance. If you have a word like “good”, what need is there for a word like “bad”? “Ungood” will do just as well—better, because it’s an exact opposite, which the other is not. Or again, if you want a stronger version of “good”, what sense is there in having a whole string of vague useless words like “excellent” and “splendid” and all the rest of them? “Plusgood” covers the meaning, or “doubleplusgood” if you want something stronger still. Of course we use those forms already. but in the final version of Newspeak there’ll be nothing else. In the end the whole notion of goodness and badness will be covered by only six words—in reality, only one word. Don’t you see the beauty of that, Winston? It was B.B.‘s idea originally, of course,’ he added as an afterthought.
A sort of vapid eagerness flitted across Winston’s face at the mention of Big Brother. Nevertheless Syme immediately detected a certain lack of enthusiasm.
‘You haven’t a real appreciation of Newspeak, Winston,’ he said almost sadly. ‘Even when you write it you’re still thinking in Oldspeak. I’ve read some of those pieces that you write in “The Times” occasionally. They’re good enough, but they’re translations. In your heart you’d prefer to stick to Oldspeak, with all its vagueness and its useless shades of meaning. You don’t grasp the beauty of the destruction of words. Do you know that Newspeak is the only language in the world whose vocabulary gets smaller every year?’
Winston did know that, of course. He smiled, sympathetically he hoped, not trusting himself to speak. Syme bit off another fragment of the dark-coloured bread, chewed it briefly, and went on:
‘Don’t you see that the whole aim of Newspeak is to narrow the range of thought? In the end we shall make thoughtcrime literally impossible, because there will be no words in which to express it. Every concept that can ever be needed, will be expressed by exactly one word, with its meaning rigidly defined and all its subsidiary meanings rubbed out and forgotten. Already, in the Eleventh Edition, we’re not far from that point. But the process will still be continuing long after you and I are dead. Every year fewer and fewer words, and the range of consciousness always a little smaller. Even now, of course, there’s no reason or excuse for committing thoughtcrime. It’s merely a question of self-discipline, reality-control. But in the end there won’t be any need even for that. The Revolution will be complete when the language is perfect. Newspeak is Ingsoc and Ingsoc is Newspeak,’ he added with a sort of mystical satisfaction. ‘Has it ever occurred to you, Winston, that by the year 2050, at the very latest, not a single human being will be alive who could understand such a conversation as we are having now?’
A huge amount of careful planning by teams of experts goes into engineering the experience of the casual end user. Consequently while the user is just trying to meet everyday needs, the designers of the technology are able to create unknown and perhaps unwanted effects.
The same phenomenon is obviously in evidence with the technology known as the “smartphone”, which we all now know is spying on people and stealing their information against their will. Their ability to defend against these unwanted effects is limited. Therefore this technology is not ethically neutral.
10 Likes
Apparently we’ll have to simply disagree here. Tools have no morality.
Sure, the more complicated the technology, the more challenging it may become to tease out both the benevolent vs malign use cases.
But a tool of any complexity is inherently inert - and therefore lifeless.
Absent life, there is no morality.
[If you’d prefer, perhaps we can pull and tug on the concept of free will? 
]
2 Likes
Considering tools without humans is the only way it makes sense to think they have no mortality. Yes, a gun without a human may be morally neutral. However, that’s neither interesting nor useful to discuss.
What’s interesting is the tool-use. The combination of tool and human. Think of the difference of a person with a hammer and a gunman. Which one is more likely to kill someone?
The affordances that different tools create are radically different. Absent of humans there’s no morality. But dismissing the moral implications of different tools (when wielded!) by saying that the tool itself has no will is a lazy, thought-terminating argument.
6 Likes
I feel this is a (not so) subtle semantic deviation that seeks to avoid having any real discussion.
Sure, an atomic bomb will not open a philosophy book on the meaning of human life as it hurdles down from the sky. It is “neutral”, as it merely follows the laws of physics. The tool itself is incapable of pondering on the effects it will have once it lands on its target.
What is being argued instead is about the morality of creating them. I’m using an atomic bomb as an extreme example, obviously. Most people agree they are a terrible and cruel idea. I myself find it so distasteful that I’m incapable of praising the genius of the people who participated in the Manhattan Project (at least their villainy cannot be overshadowed).
An LLM, for sure, may have many practical uses. I agree it could have the potential to make our lives at least a little better. Yet, so far, it’s not what we are seeing, especially when the CEOs of the companies that created them seem to be handpicked to be the most immoral people on the planet.
PS: by the way, I use LLMs all the time. I hardly have a choice: at my place of work it’s either that or be fired.
6 Likes
The neutrality of tools debate is off topic imo. Regardless of whether you believe tools have morality, the topic is about the way humans create and use the tools.
6 Likes
Again, I beg to differ.
I feel neither “lazy” nor “thought-terminating” in offering my comments above.
Bad people reach for tools to do bad things. Good people can use these same tools to do good things.
Even violence on a large scale is moral and entirely just when applied in protection of the innocent or in the course of self-defense.
Tools enhance our capabilities - not our ethics.
Simply put: I find the urge to imbue our tools with core human principles - like ethics, virtue and morality - is to undermine the very essence of humanity while deflecting responsibility, accountability, authority and control whence it belongs. Squarely on humans.
2 Likes
It would be helpful if OP returned to elucidate on his point, but my comment was based on a presumption he was making an objection such as LLMs being built on a foundation of massive copyright violation. So even if you use the LLM tool for good, you are complicit in that original bad action. Much like using a product created by slave labour, or blood diamonds.
3 Likes
You mean this is comparable to the “I don’t like [insert your atrocity of choice], therefore OTHER people should stop doing that” attitude?
I that case I could agree.
If, OTOH, it is about tastes in music, or about ice-cream flavours, etc., such an attitude is nonsense, OF COURSE.
The difference, as I understand it, is whether someone does something to others (including non.-human animals) or to oneself.
If someone wants to drill a hole into their kneecap, well … 
… some people do even weirder things. But drilling a hole into the kneecap of others is a very different thing.
Typo: mortality ≠ morality
+1
This!
PLUS there’s the environmental issue.
BTW, even though some people seem to believe that I am generally against AI, this is not the case, as I have explained several times, I can’t remember all places where I did so.
3 Likes
We have outlawed most atrocities, so most of them are no longer up to debate nor in the hands of random people to campaign against. The law exists and there are, depending on the case, the appropriate civil services that most people need to call in case of most actual atrocities.
In the cases of things that are emergent or new or “high tech” or currently beyond the scope of the law, it should be pointed out that there are many ways to inform the public about the issue at hand and attempt to swing the public opinion in favour of taking action against whatever is going on, atrocious or otherwise.
It could be argued, but I will spare you all the ordeal, that the “I don’t like [insert your atrocity of choice], therefore OTHER people should stop doing that” or, even worse, “therefore OTHER people should do something about it”, is probably the worst and least efficient and most likely to fail, among the multitude of “ways to convince/sway the public”.
You might notice that it is rarely used by demagogues, politicians or marketing campaigns (those use the positive re-inforcement version which goes “this famous person likes this things and if you want to be like that famous person you should do that as well”… the “Grant Hill drinks Sprite” advertisement in the 90ies comes to mind, for example) exactly because they know that it is a bad/inefficient idea.
To showcase how vegans became a meme due to the over-application of that bad idea, only 4-6% of Greeks identify as vegans or vegeterians. It is considered a fringe and cult-like idea here and severely frowned upon by a culture whose religion, the Greek Orthodox Church, already mandates around 180 to 200 days of fasting per year. These people are more than half-way there, but still think that the practice of vegeterianism is insane.
That’s how much “bad marketing” is that kind of practice.
The same trend now applies to the “I hate AI for X-Y-Z reasons and I want other people to stop using it”… even if those reasons are very valid for that person, they might not be for those that already use the AI or, as it has been pointed out by many people, the AI usage is mandatory at their jobs and they have no choice in AI usage, at all.
Sanctimoniusly declaring yourself above those people’s problems and judgement and showcasing how elevated your ideals and practices are, compared to the other people that have yet to realise how bad this practice is, rarely generates any sympathy for that cause.
Very few people like AI. Most people dislike it. The vast majority of people share a lot of worries about the future it will create, the cost it will have in resources and the ethics of the companies that trained the AI and how they took all that training data from other humans.
It should be easy to convince them to dislike AI or be as keen as possible to reduce its usage or advocate for it being regulated and so forth. In fact one of the few ways to not convince people of those things is the “I do not like it, so you shouldn’t do it either” attitude. There are far better arguments and ways.
Smoking is a problem which you probably understand far better than most of us and most people now understand that it is a hazardous thing. Have you ever tried the “I quit smoking, so you should quit too” strategy to anyone and has it ever worked? 
Or did you use a totally different approach and told them how many problems smoking caused you and how much your health improved when you stopped smoking or how much that other person could also benefit in health and longevity and in better ability to enjoy life and its experiences like walking/running etc? Isn’t the “I found many benefits in doing this, so maybe others could also benefit by doing something similar?” obviously better than the “I don’t like this, so you should stop doing it?”
What about an experienced programmer vibe coding?
3 Likes
that’s what I said
1 Like
There is a lot of room between vibe coding and coding without AI assistance.
3 Likes
yeah, an experienced programmer vibe coding is like, the ai is vibing with the programmer, instead of the opposite with random vibe coders
The other day, I was complimenting my electrician, as he finished up with his apprentice at my house, on his communication throughout the process. After calling them, I had received a quote I could approve by email and clicking on a link, followed by a booking and confirmation and SMS that they were on their way at the designated time.
He said “Oh, yeah, we used to use a service for that, paid monthly, but it was crummy, so I built this one, now it’s perfect for what I want”.
“Oh, did you have someone code it up for you?”, I asked.
“No, I built it. You know, you can do that with with AI these days - no worries”.
My electrician built his own quotation and booking service, to avoid paying “some big SASS company for a service”… he was able to do this because… AI.
I think the time is behind us when you can avoid AI-built-software.
And the time is behind us where you can even determine the pros and cons of doing so: was it better that my local electrician got my money, or should I have given it to some other bigger company that uses the SASS that was built by AI anyway?
As anoek mentioned, using AI to code has allowed folk to contribute that otherwise could not. How do you measure the benefits of this against boycotting LLM companies (ineffectually).
I started doing this because I needed to learn about what was possible, to stay abrest of my professional field in my day job.
Being able to learn about coding-with-AI here at OGS, ahead of the curve, literally shaped my career in the last year.
Then, I continued doing it here because it gets things done easier and quicker.
We have, at OGS, Community Moderation now: a whole new sophisticated chunk of code that has made it possible to scale moderation - literally changing whether or not we can keep up with the load - thanks to being able to describe it to AI and get it coded. It’s not like “if I chose to code it by hand it would come a bit later”. It simply wouldn’t get done: being able to generate massive features like this totally changes whether they are possible or not.
Asking “why it would seem like a good idea” is like asking why industrial manufacture seemed like a good idea, at the time that machine looms and production lines were emerging. It was (is?) awful in so many ways, but its obvious why it is a “good idea” if you are in the business of manufacture - and the same is true now if you are in the business of making software.
A few people still wear only hand-woven clothes, but it’s not really a sustainable business to be in on the whole.
I guess in this respect it is a “good idea for OGS” because we want OGS to be the site that people chose when they look for Go online … and a “hand woven OGS” won’t cut it to meet this goal - other than for the few who are specifically looking for that.
14 Likes
I totally agree that this is most certainly true, however I have to point out two worries:
a) Your electrician might have managed to prompt-generate his own quotation and booking service, which I find awesome, but is he ready to provide support for it, if things go pear-shaped? I am worried that one day he might encounter a client who, upon noticing that the app is “AI generated”, they might get the idea to “AI reverse engineer it” and hack into it probably by pretending to be the “app owner that forgot his admin password access” to bypass any “ethican constraints” imposed by the AI companies.
b) In a similar RL discussion I had with some people in the field that had come here for vacation, it seems to be that even “senior programmers” now are just participating in an “AI hand-off” between them where one AI is reading the emails, makes the summary, then the programmer just feeds the summary to another “programming oriented AI” and orders it to apply those changes, repeats the process a few times with some “improvement tools and metrics”, then re-summarise the results, send them to the “emailing AI” which makes them more presentable and off they go to the “project manager” who doesn’t read them, has his own AI and so on and so on and on…
Someone told me that he has people impressed at work because the company uses “AI Z” while he also uses “AI Y” to parse the code and it finds different kinds of mistakes so everyone at work is looking forward to his “astute observations”.
None of the contributors have looked at the code manually, for months. Noone knows what’s in there. They only know that it works… somehow.
Replace “garbage” with “tech debt” and the future of coding will probably look like the intro of Wall-E. 
Now, I am willing to bet that this “noone knows what’s in the code” level will never be reached in OGS (since it is a project with a realistic scope run by people that actually care for it), however it is going full steam ahead elsewhere and just like the “AI tide” rose and brought us to a level where coding with AI cannot be denied, it might also rise to the level where “knowing what your own service’s code contains is an utopic fantasy”.
2 Likes
This doesn’t sound like a very plausible scenario. AI companies are generally not storing account credentials, and they definitely aren’t handing them out to other customers.
If this were the case, then our problems go far deeper than one contractor’s quoting app.
Excellent point. It illustrates the larger challenge regarding AI as a tool.
Ultimately, it’s the human user who makes AI either powerfully constructive or powerfully destructive.
Whether one is optimistic or pessimistic about the future of humanity generally determines whether one has a positive vs negative view of the future of LLMs and AI in general. 