* REMOVED *

This comment was deleted on Apr 10, 2025, 12:23. Reason: Personal attacks (rule 1)

Burrito of Peace wrote on Apr 9, 2025, 10:04:

El Pit wrote on Apr 9, 2025, 09:41:
It will take time but AI will make things work quicker and more effective than we could. This will result in very bad things (AI used for crime and warfare) and very good things (AI used in medical procedures and administration).

This is future tense spoken with certainty when, while certainly a possibility, it's not a forgone conclusion. It's a good example of an unfounded belief based upon the hype. It's a lot like the Nuclear Age of the 50s when every home was going to have its own nuclear power plant, robots to do everything, and infinite leisure time. That did not happen but a lot of statements were made with future tense and certainty at the time.

I look at it like a calculator. A calculator can do long division faster than I can. So for certain things, AI will be faster, it remains to be seen what that is exactly yet.

I agree that any wariness that one might have for the potential for AI to be abused by unscrupulous powerful people is not hysteria. Which is why I didn't claim that. But neither am I going to consider the future from the bleakest, darkest perspective imaginable as people are wont to do. As you are prone to doing from time to time.

Donald Trump, Vladimir Putin, and Kim Jong Un all have demonstrated varying degrees of psychosis, narcissism, sociopathy, etc and yet have the sole decision-making capability to launch weapons that will annihilate mankind in 45 minutes. Yellowstone is 500,000 years overdue for an eruption that will affect the entire planet for generations. Asteroids and comets have potential of impacting the earth every single year. A solar burst from the sun (which already happened less than 300 years ago) could send mankind back to the stone age for years. Many viruses are one lucky mutation away from becoming global killers. And on and on and on and... Everyone can stop falling over themselves looking for new things to be petrified about; there's plenty already. So FFS stop worrying. Something is GOING to kill you. Whether that's tomorrow or 500 million years from now...who the fuck knows? Live your fucking life.

AI is a great tool for people who already know what they're doing but a real pain for those same people because they KNOW what they're looking for and what they're getting aren't necessarily the same thing. The danger is for people who aren't competent, they won't understand why, if the response is 99% right, the last 1% is causing their result to be undesirable. High-functioning coders for instance aren't born, they're bred, and if they don't have opportunities to gain experience, we'll lose them.

As Kirk told Saavik, "You have to learn why things work on a starship." Else you get command-coded to oblivion.

El Pit wrote on Apr 9, 2025, 09:41:
Imitation is the very basic principle and technique of human learning. And this is what is going on with this Quake 2 AI demo: it is a basic and still clumsy imitation. And the AI will learn and improve on it.

You're anthropomorphizing "AI" here. "AI" doesn't learn. Learning requires perception, inspiration, and intuition. Three things that the autocomplete chatbot has shown a decided lack of capability in doing on its own. It can ingest data, even novelize the data in different ways in a random pattern, but it cannot inherently learn.It cannot, at this time, create anything new.

El Pit wrote on Apr 9, 2025, 09:41:
It will take time but AI will make things work quicker and more effective than we could. This will result in very bad things (AI used for crime and warfare) and very good things (AI used in medical procedures and administration).

This is future tense spoken with certainty when, while certainly a possibility, it's not a forgone conclusion. It's a good example of an unfounded belief based upon the hype. It's a lot like the Nuclear Age of the 50s when every home was going to have its own nuclear power plant, robots to do everything, and infinite leisure time. That did not happen but a lot of statements were made with future tense and certainty at the time.

El Pit wrote on Apr 9, 2025, 09:41:
But one thing is sure: once the genie is out of the bottle, it can never be put in again. We will have to make sure that we stay in control because this is a bit like making contact with alien life: you don't really know how the other intelligence ticks.

Agreed. But what we can do is be cautious, regulate it, and not give techdouchebros and corporations free reign to jam it down the throats of every human possible. You aren't building nuclear power plants in your garage/basement/spare bedroom are you? Obviously not.

Of course, da Vinci and Lilienthal copied what they saw in nature. This is obvious in their models. But then their human intelligence and genius improved and changed the designs when they found out they don't have the mobility and muscular structure to fly like the birds. And this is the typical technical evolution: copy nature, improve the design to fit human requirements/limitations, and the result gets worked on over decades to (sort of) perfection.

Imitation is the very basic principle and technique of human learning. And this is what is going on with this Quake 2 AI demo: it is a basic and still clumsy imitation. And the AI will learn and improve on it. It will take time but AI will make things work quicker and more effective than we could. This will result in very bad things (AI used for crime and warfare) and very good things (AI used in medical procedures and administration). But one thing is sure: once the genie is out of the bottle, it can never be put in again. We will have to make sure that we stay in control because this is a bit like making contact with alien life: you don't really know how the other intelligence ticks.

El Pit wrote on Apr 9, 2025, 04:09:
Otto Lilienthal "plagiarized" bird wings. And he died when he tried out his wings. But he started something that ended in us having intercontinental flights. Things usually start small with failed experiments but give it some time and they might result in something that changes the world - for good and/or for bad.

Technically...no. While Lilienthal was absolutely inspired by bird wings, there were significant technical differences between the stork wings that fascinated him and his gliders. Further, his scientific and mechanical methods were extensively and rigorously documented and tested. And there's the argument that Michelangelo's Codice sul volo degli uccelli predated Lilienthal by a few centuries.

But beyond lift and duration, plagiarism implies a 1:1 copy and clearly that isn't the case with Lilienthal. We look, and have looked, to nature for inspiration in many ways in everything from flight to construction to nuclear power. But natural evolutionary occurrences are not intellectual property barring invisible sky wizard daddy fantasies. If I had to pin it down, the difference is in the act of perceiving something. From that perception comes inspiration. "AI" cannot perceive anything and therefore is functionally incapable of inspiration. At best, it can take what already exists, jam it all in to a box, and shake it up to see what sticks. That's just pure randomization. Whereas both Michelangelo and Lilienthal were inspired to go beyond and take leaps in to the unknown.

Prez wrote on Apr 9, 2025, 00:25:
The cringe-worthy, obnoxious corporate overuse of AI for its "buzzwordiness" is worthy of constant ridicule. The claims of "AI should be outlawed because it might one day do my job better than I do" is what makes luddite hysteria. Suck it up folks. I don't know how to be any more blunt.

Is it hysteria to be concerned about the autocomplete chatbot deciding or arguing death penalty cases? How about the outcome of CPS cases? Or the decision as to whether someone receives life sustaining medical care...or not? Should those examples be determined by a qualified human?

The consternation isn't solely that the spicy chatbot is replacing entry level code monkey jobs. It's what techdouchebros and their narcissistic Randian vision want further down the road that directly impacts the life, health, and safety of possibly millions of people. They've quite literally said as much.

I would not at all agree that it is all "hysteria". I would argue that is prudent caution and concern. Especially in the US where our own government is more than happy to trample the liberties and rights of the people at the behest, and for the enrichment, of corporations. Even moreso when we consider that the US fundamentally has one of the worst social safety nets among all first world nations.

Overon wrote on Apr 8, 2025, 23:05:
1 If I plagiarize Huckleberry Finn by Mark Twain because it is now free because it entered the public domain, does that mean it's not plagiarism?
2 Have you played it, it's slop.

Otto Lilienthal "plagiarized" bird wings. And he died when he tried out his wings. But he started something that ended in us having intercontinental flights. Things usually start small with failed experiments but give it some time and they might result in something that changes the world - for good and/or for bad.

The cringe-worthy, obnoxious corporate overuse of AI for its "buzzwordiness" is worthy of constant ridicule. The claims of "AI should be outlawed because it might one day do my job better than I do" is what makes luddite hysteria. Suck it up folks. I don't know how to be any more blunt.

Timmeh wrote on Apr 8, 2025, 16:04:

Overon wrote on Apr 8, 2025, 15:22:
Carmack should be calling it out for what it is slop and plagiarism.

um no it is neither one of those thing.

1. Quake code has been free for 20 years. So is Quake 2 and 3.

2. how is AI recreating Quake "slop"?

1 If I plagiarize Huckleberry Finn by Mark Twain because it is now free because it entered the public domain, does that mean it's not plagiarism?
2 Have you played it, it's slop.

It's pretty clear that I have a problem with "AI" but I think the issue for many people isn't with the technology, itself. It's with the blind corporate rush to overhype it as much as possible and shoehorn it in to everything possible with absolutely zero regard to actual usability, ethics, and whether it brings any value to either the product or the user. In addition to that, the average person doesn't have the ability to properly elucidate what their core issue with it is. What they do know is that suddenly "AI" is being crammed down their throat without their consent. Like jamming Co-Pilot in to Notepad or looking up a recipe that includes glue...when the recipe itself doesn't actually use glue* or "summarizing" their search results that misses crucial nuance and details.

There are always going to be people whom have an issue with technological progress but to label everyone who complains as a "luddite" is unfair, in my opinion. Because there are very real and valid complaints with the way corporations are behaving with regards to "AI". We haven't even addressed how wildly energy inefficient it is on a power grid that, in many places, is already burdened as it is. It's no better than cryptomining farms and yet no one had a problem mocking the shit out of that.

One of the things I hear most frequently is that the autocomplete chatbot handles what is, essentially, busywork. "Well, it writes my emails or TPS reports". If it's busywork, it needs to be cut away entirely. Optimizing a workflow by removing dead weight is always a much greater net gain than simply brute forcing your way through it with raw power. Because that brute force incurs both a real and technical debt that has to be calculated on the balance sheet. It's overall much better to cut away the dead weight and pivot. You don't incur additional costs and you can reallocate both funds and manpower to more productive ends.

I still stand by the notion that so many people latch on to "AI" because they're desperate for a real, progressive breakthrough. Either they lived through the truly miraculous and rapid breakthroughs we saw in the 60s through around the mid-2000s or they are at least the beneficiaries of it. We've kind of stagnated since then. I recently watched a video where some very earnest but historically illiterate individual was hyping EVs as a "21st century breakthrough". They were completely oblivious to the fact that we've had EVs since the late 19th century. They have undoubtedly gotten better, exponentially so, but they are not "new". But when you've several generations that lived through truly revolutionary technologies and now we're treading water with evolutionary technologies, it is easy to see why people are latching on to whatever they can out of hope that the hellscape dystopia around them is temporary and a new technology will come along that will save them.

*Someone is going to say "But they fixed that!" The point is, it never should have made it out in to the world in the first place. It is an example of a stunning lack of oversight by techdouchebros who are looking to grift and "disrupt" without any regards to safety, accuracy, or feasibility. The three things which killed off the Ford Nucleon, and with good reason.

Mystic95Z wrote on Apr 8, 2025, 16:21:
AI needs to be outlawed... Think about it, what happens when AI replaces millions of workers? Yes a huge savings for companies, but when those workers don't have the income to spend what happens...

Wheelcarts need to be outlawed because think about all the people carrying goods and their jobs. And assembly lines need to be outlawed because think about craftsmen losing their jobs.

Progress cannot be stopped, no matter the good or bad consequences for people and their jobs. AI is changing many working environments and many jobs in e.g. administrations and software development will be gone in the next decade. This is why the basis of the economy (everyone needs to earn enough money to financially support himself and the family) needs to be reconsidered. No, tariffs won't help that. The changing society because of the job market will be the biggest challenge for the next few governments and the panic fear of changes of the current government and its supporters are not helping.

Regardless, “don’t use power tools because they take people’s jobs” is not a winning strategy.

Preach JC. The luddite hysteria has become deafening; good that someone still speaks sense.

Mystic95Z wrote on Apr 8, 2025, 16:21:
AI needs to be outlawed... Think about it, what happens when AI replaces millions of workers? Yes a huge savings for companies, but when those workers don't have the income to spend what happens...

That's why there's pilots around the world on universal basic income. Because the writing is on the wall. It's coming for peoples jobs and society needs to find a place for it.
There's talks about taxing robots and corporations but I'm pretty sure it'll end up being more strife and a broken system put into place well after it's caused damage.

taryuken wrote on Apr 8, 2025, 13:45:
Not a fan of middle and upper management gauging analytics to find that I'm not using AI at work and then messaging me about it. I started using AI to respond to emails instead of mgmt's vision of its use and haven't heard a peep from them since. It's fine for boiler plating some code, but trying to compile and work entirely within AI is not currently ideal. I'd rather spend the time researching subjects and developing than fixing AI's errors that were based upon google searches and online knowledgebases.

My dev friend's workload has tripled, as his managers expect him to use AI codegen for development, and him just to tweak it. He says it's exhausting, and half the time he doesn't fully understand the code he's deploying due to the time pressures.

Bill Borre wrote on Apr 8, 2025, 12:08:

social media, where creative entrepreneurship has flourished at many different scales

Really? I wouldn't have expected such an opinion from Carmack. It sounds more like something shills for Zuckerberg would say.

It's simply an observation. Look at how sites like YouTube have democratized content creation. So much of that content would never be created by a studio, but is incredibly informative and entertaining while at the same time providing some compensation to incentivize creators (although, not enough).

Or something like OnlyFans with its niche content about the right way to handle a baseball bat, or how to filet a fish. A studio would never create content like that, but it can be an enterprise requiring only hundreds or thousands of supporters, as opposed to tens or hundreds of thousands that a traditional studio might require to be profitable. In addition, it offers personal interaction with content creators, something you can't usually get from the traditional system.

Mystic95Z wrote on Apr 8, 2025, 16:21:
but when those workers don't have the income to spend what happens...

I'm selling new bumper stickers, "Learn to prompt"

AI's ushering in a culture comprised of meaningless Content For Content's Sake. Just a whole lot of "stuff" that says nothing, means nothing, contributes nothing. Look what we're talking about here, a demo of something entirely derived from things that already exist. There's no purpose to it. Just like the Studio Ghibli filters, a pointless act of content creation that someone glances at for half a second and then scrolls past.

Carmack's power tools analogy only goes so far because with power tools you still have a human tradesperson directing their use. If power tools were erecting buildings automatically but randomly used the wrong joins or something else risking building collapse because they fucked up every so often, the construction industry would probably ban their use.

AI needs to be outlawed... Think about it, what happens when AI replaces millions of workers? Yes a huge savings for companies, but when those workers don't have the income to spend what happens...