Is anyone else tired of the snobbery, condescension and down right hate around this topic? People are writing articles like this saying how AI is useless (also, the sarcasm is so spicy), yet I feel this underlying sense of fear towards this. Can they please make up their mind? And I know the argument is that investors don't understand quality and this takes away resources from quality writers, but quality writers have lost the war anyways, apparent from the decline in the percentage of quality content. Are they fighting for the right to produce low quality content?
I was not old enough when internet was coming up, but I wonder if there was a similar tone towards internet when it came along. I've seen some quotes in places complaining about how early internet destroys the human part of booking a holiday. That was interesting to read. I wonder if there is a book about this.
> People are writing articles like this saying how AI is useless (also, the sarcasm is so spicy), yet I feel this underlying sense of fear towards this. Can they please make up their mind?
I don't see a contradiction here. "LLMs produce low-quality content at extremely low prices" is something that can reasonably inspire both fear and ridicule simultaneously. (A good word for it might be "dystopian".)
> Are they fighting for the right to produce low quality content?
Paying writers is a great way to produce higher-quality content. But even producing low-quality content is better than being a Walmart greeter.
> I was not old enough when internet was coming up, but I wonder if there was a similar tone towards internet when it came along.
I was, and I do not remember such a tone. (There certainly were moral panics over things like children getting access to porn, though.) The difference (to my mind) is that in the early days of the internet, new technology made clear and dramatic improvements in human well-being. Things like email, mapping software, and online shopping let people do things that were difficult, expensive, or impossible before. (Search for "long tail" to find discussion of the latter.) If anything, there was too much derision in the other direction -- the phrase "buggy whip manufacturers" was thrown around quite a bit if you want to hunt down examples.
Over the last 10-15 years, the big new internet-related tech has been much less beneficial. Instead, we've gotten things like the "gig economy", cryptocurrencies, walled-garden app stores, and algorithmic social media feeds. In terms of human well-being, these changes range from ambiguous to terrible. And the whole time, tech companies and VCs have continued crowing about how they're advancing humanity toward the future, with (apparently) very little introspection about failures like "creating a giant get-rich-quick scam that boosts carbon emissions" or "accidentally funneling millions of young men toward white supremacist propaganda", alongside lesser failures like "SEO winning the arms race against search engines" or "creating user-tracking systems that rival the East German secret police for the purpose of ad sales".
Given that context, a new technology that seems like it's designed to drive writers into unemployment while flooding the rest of us with spam is going to get a fair bit of skepticism.
> Things like email, mapping software, and online shopping let people do things that were difficult, expensive, or impossible before.
The only thing I'll say to this is that the LLM tech is pretty new and we don't know what it's applications are yet. It's more akin to lasers than the internet. Lasers were theorized in physics and even when the first ones were invented I don't think anyone would have imagined how ubiquitous they will become in manufacturing.
I recall those days too, they sadly seem like from another, very different universe compared to 2023. Everybody was optimistic, internet had endless possibilities simply not available before. Of course all highly functioning sociopaths that now form the thick layer of founders/owners of richest companies saw another potential there, and there was nobody to stop them. Thus here we are.
> creating user-tracking systems that rival the East German secret police for the purpose of ad sales
East german police were amateurs compared to what few big companies can (and do) offer to CIA or NSA within few clicks. What we have now is definitely beyond their wildest dreams. I recall not so long ago 1984 was mentioned to scare people. Not so much anymore
Thing is, you can't trust that the take over of a technological innovation will actually improve people's lives, the economic incentives for their use lining up can lead to a local maximum where some balance sheets look better but in every other way people's lives got a little worse, or much worse.
We already saw the web drastically go down in quality when guided by the invisible hand of PageRank and ad networks, content was optimized to appear on the frontpage of Google searches and bring in ad revenue, while being of low quality, extremely sparse in useful content, and melting your computer as your browser renders the ungodly amount of ads that are bundled with it.
That's a very direct precedent already, no need even to make analogies to historical developments happening in entirely different contexts or to other industries.
Any sufficiently large system will have measurable incentives and disincentives, ceilings and floors, in place of personal loyalties and trust as you would in small groups of people.
The unfortunate outcomes of them of course would be different with different political and economic systems, we know the ones that socialism created in the XX century, we are still getting slapped by the ones created by capitalism and representative democracy now, with new ones coming up hot.
> Is anyone else tired of the snobbery, condescension and down right hate around this topic
Isn't it the logical response to the fanaticism, ubiquitousness and down right obsession that seems to dominate forums like HN and Reddit?
I can't think of anything that could make a topic less interesting than fanboys making every other topic be about that. Every conversation has a "here's what chatgpt said" comment on it.
Aren't you being a bit unfair? Shouldn't enthusiasm around new technologies be encouraged in a technology forum like HN?I get that it can be a bit overbearing but come on it's exciting and people want to express that. It literally came out like a couple months ago, whats so wrong with being excited by something cool? I would have that over this negativity any day.
I hears ya: the eternal September of discovery. Perhaps, like me, you have a cynically high threshold for novelty plus a low regard for repetition? But https://xkcd.com/1053/ seems relevant - trying to empathise with the joy in others and maintain our own childish joy in the wonders.
Low value, yes. However I would give a strong negative value to your very public cynicism. Pissing on people’s excitement is not exactly helping. ChatGPT is revolutionary, even if I personally find the AGI discussions turgid.
It is especially egregious to complain about it when the link is about ChatGPT - why don’t you just avoid commenting on those links?
> linking to a google search
I started doing that a year ago (with duckduckgo links), and I think this is often a genuinely useful thing to do and it saves other people time. Firstly I ensure the search terms are specific, saving the reader the effort to discover what search terms work. Secondly I do it for acronyms - to try and prevent that one person asking what the acronyms are.
Well, there was Clifford Stoll's "Silicon Snake Oil" from 1995 in which Stoll (who had been using computers and the Net for quite some time and had famously caught a hacker spy as described in his "The Cuckoo's Egg"), although he was mostly claiming that the early consumer Internet was overhyped. He was right and wrong. Yes, there was the dot-com crash of 2000, so it was somewhat overhyped, but Stoll also claimed that nobody would want to shop online when they could just go to a store, which clearly wasn't the case.
I really enjoyed The Cuckoo's Egg, I loved his writing. I will not let go of the chance to encourage people to read that book. It's amazing and worth your time.
Will check this out. Thanks for the recommendation.
I think of current AI as the worlds best bullshitter.
It often sounds good enough compared to what most high school students can write in essays. As the #1 criteria for an essay is length, students gets trained in adding bullshit, as 5 pages of bullshits is acceptable but 1 page of concise text is a fail.
A lot of people don't feel they have enough control over their live, especially if a single accident (or losing your job) can make you homeless.
So I think the underlying emotion is fear. In in dogs fear often leads to attacks, so most dogs that bites are fearful dogs, not aggressive ones.
Snobbery, condescension, and downright hate seems fairly appropriate. To paraphrase someone on social media: "If ChatGPT is the solution, the problem couldn't have been very interesting." Despite this, we're constantly inundated with heavy-handed prophesy about the world-changing effects this technology will have... any day now. The hype cycle is naturally met with resistance.
I think what's missing from this is the ROI calculation. They see AI as having degraded quality compared to human authors, but the difference in quality doesn't produce enough profit to offset the cost of hiring a human staff.
> And I know the argument is that investors don't understand quality and this takes away resources from quality writers, but quality writers have lost the war anyways, apparent from the decline in the percentage of quality content.
For some use cases you don't need quality. E.g. where I live there's a free newspaper, the Metro, given out on buses/trains. The articles only exist to get people to read the ads. Most of the articles could probably just as well be written by AI. For all I know, some already are.
But some use cases do need quality, and quality will remain.
When content farms were on the rise, people feared reputable sources like NYT would lose out because they cannot compete with the quantity of low-quality articles being churned out.
It seems, while NYT is still around, NYT did lose the bar of quality they had along the way, so it was a net loss for good writing.
Funnily enough, if you analyze these articles less from a "is AI crap?" or "AI is gonna take our jobs!" perspective, you get a higher level one. You'll find "AI is a hot topic and I can capitalize by stirring the pot and making click bait about it for profit!".
Many editorial execs don't really care if AI is or isn't crap, gonna take our jobs, or changes the world. They care they can write about it in a way that will draw your attention while filling ad slots on their site.
You are not meant to say that out loud, you are meant to show that it doesn't phase you in silent approval. The culture curators have gone to great lengths to leave an opening for your social advancement at the watercooler, it'd be unwise to refuse their offer.
People are tired of the tech industry crapifying everything. We built the current situation where nothing we do can be trusted and most of it is scams like Uber.
I was not old enough when internet was coming up, but I wonder if there was a similar tone towards internet when it came along. I've seen some quotes in places complaining about how early internet destroys the human part of booking a holiday. That was interesting to read. I wonder if there is a book about this.