Does anyone else hate AI?
-
Everyone I know thinks AI is totally awesome and uses chatgpt for everything (e.g. finding a restaurant to eat at, or plan their childrens birthday).
It gives me such bad vibes and I can envision a future where people can't do even the most simple tasks without consulting chatgpt beforehand.
Does anyone else feel the same?
-
@Korven said in Does anyone else hate AI?:
Does anyone else feel the same?
I don't trust AI since it gives you an answer programmed by mainstream knowledge.
And when I use it, it's to make it easier for other people to understand a complex concept. Except, there're often biases / errors.
If you "tell AI" you don't agree and give an science source, it well say "sorry" and admit in some cases, it could be different. Than the explanation becomes more acceptable. But always mention at the end of your question: "Please, give a developed answer. Take time to do so".
Otherwise, the first caught "official info" will be presented. -
@Korven You need to be extremely vigilant when using AI. Always double check the information if it is important. But it is very useful for getting information that is usually heavily gated by academic language, like biochemistry.
Always ask very specific questions, never general ones.
-
@Korven about a year ago it was whack as fuck, I re-tried a different ai and now its very useful if you set it to dial in on academic papers . i like perplexity ai it bundles the popular models to optimize answers
but yeah downside i can see, it could amplify typical mistakes in thinking because thats whats shared often in general sources (specific ais targeted at academic papers can help solve that)
or could be some manual fuckery going on biasing answers in some way , people of a certain mind would see that and boycott for better ones but many would just go with the most popular currentlyAn upside ive already seen in a few forum posts is people who havent got understanding on health much using it to gain insights on their health problems their doctors couldnt give them
@pittybitty yeah i found if you bombard it with specific questions in an area you dont understand well , can stack the highlights to build understanding very quick without so much digging having to scan papers for the next part u want to know.
sometimes it takes asking the same thing in a different way, and asking it to highlight specific sections of sources to back up its answer, to make sure its accurate or see its logic if its making assumptions instead. sometimes i ask it to forget the sources used and use whole new sources to give its answer -
AI removes the critical thinking we employ as humans to collate information/data and base a decision on it. It IS amazing, it IS capable of elevating mankind to a new level of technological and societal innovation, it will most likely devolve a huge part of our population just as smart-phones have/do to the point where nothing but an empty husk is left where the soul used to be.
-
@DKJoeAgain Nah, nothing like that. It's just a fancy search engine that allows some people to find the information they need more easily.
-
@Korven it's just a search engine mostly. it makes researching topics and finding answers way easier than going through primary texts. it also has to be double checked often so it isn't that good. when people use it to make forum posts it makes forums pointless to read btw. the point of using forums is to interact with humans not chatgpt
-
"AI" or rather LLMs known also as stochastic parrots, can be useful in some specific applications, but in general, it may make humans more stupid and less creative. Also, they consume a lot of energy... etc.
There is a nice manifesto against it https://copin43.hashnode.dev/the-nuremberg-defense-of-ai
A little learning is a dangerous thing ;
Drink deep, or taste not the Pierian spring :
There shallow draughts intoxicate the brain,
And drinking largely sobers us again.
~Alexander Pope, An Essay on Criticism -
Yes, I hate it. I try to look away when I'm shown any AI content like google's AI search result or AI art. I don't want it to pollute my mind. The more accurate and powerful it gets, the more I hate it. It is against nature and inhuman.
-
hate...
I understand but you've got to budge because it's out there and so are you. It's not "AI" anyway, it's a statistical inference from a curated dataset.
Everyone I know thinks AI is totally awesome and uses chatgpt for everything (e.g. finding a restaurant to eat at, or plan their childrens birthday).
Another reveal of communication problems between people that were already there. Think forward of them.
There are interests that benefit from that but it's no good to rest on blaming them really. If I get between you and a pal you could blame me for being a dickhead, and yourself for not going around me.
It gives me such bad vibes and I can envision a future where people can't do even the most simple tasks without consulting chatgpt beforehand.
Don't. Why would you. Are you going somewhere Korven.
-
@Kvirion said in Does anyone else hate AI?:
it may make humans more stupid and less creative
The contrast is going up.
-
Did roads make humans more stupid? Or did it just help people get there sooner? I mean you can still walk wherever the road isn’t.
We can just not use ai and then when your trying to go far you use ai. -
Okay you bring up some fair points:
-
You can prompt AIs in such a way so as to not get spoon-fed mainstream stuff
-
Chatgpt and other LLMs are just better search engines
-
Someone used them to get insights into their health problems (AI smarter than doctors? Most likely).
-
AI is not actaully AI, AI is a statistical model trained on large data sets.
-
AI is here to stay so stop being bigoted towards AI you fool
I guess the main concern is what @Kvirion wrote about making humans less creative.
I admit I have used it in my work occasionally but it's a guilt ridden experience where I feel like I am implicitly contributing to the downfall of humankind... so you're telling me I can use chatgpt without feeling guilt/shame?
-
-
@Insr That meme is hilarious
This is totally me -
@dapose said in Does anyone else hate AI?:
Did roads make humans more stupid?
LOL! It's a terrible analogy...
Roads don't bullshit i.e. make a fake things up...
When you use a road, you still use your legs...When a human mind looks for a piece of information and checks some sources, discusses them with others, then synthesizes it, this can: 1. train a (social) mind, 2. allow for a creative insight, i.e., to spot something novel.
LLMs can't do it... - They look only for existing connections...
For a human, not using the mind means atrophy of it...And I don't even need to go further and discuss the meaning of non-ergodicity, power laws, and fitness landscapes in creativity... LLMs (and their midwit fans) simply can't comprehend this as they just follow Gaussian models...
-
@dapose Very bad example because automotives made cities unbearable to live in. Much of the charm of cities was clever use of space so everything you could dream of is in walking distance. Now that cleverness and all it's positive side effects is lost. Cars, and the laws of locality they circumvent, are genuinely a thing that made the world a stupider place.
-
Hating AI in general would be reactionary. As others have pointed out it is "nothing more" than an advanced word association database-machine. But if this is already sufficient to do a passable job of replacing some low-tier knowledge professionals like doctors and lawyers, (or perhaps nurses and paralegals) that is great.
I also welcome the fact that AI signals the end of intellectual property rights. The contradictions are beginning to sharpen there.
-
Don’t be naive; or worse- ignorant.
Are you familiar with Huxley’s “Brave New World”?
-
@Peatful It is not surprising that a British person enjoying the fruits of world empire would write a weepy treatise about "the importance of true individualist adventure". Code: the importance of backbreaking labor for people halfway around the world that I ideally never see, to fund my trips to the cafe for African coffee.
This is what Orwell was as well. He writes his truest fear as that of being objectively understood, which is very telling.
I've 1984 about 5 times, Brave New World probably twice. They were my sauna books when I was a teenager. I wanted something to read that I didn't value very much and could happily sweat on.
Doctors, especially the mediocre ones, "may as well be" AI machines already. The fact that these medical association devices are implanted in a human brain does not make them special. In fact, their subjectivity in part works to keep them closed to the complete and scathing critique necessary of the medical ideology. When the ideology is directly machine-generated, it is easier to critique in this way. Shelve it along version #221 as the broken tool that it is.
The fact that DeepSeek was able to more or less from-scratch develop a state of the art LLM gives me great hope that this technology can develop along objective lines instead of those brought down by the western elites.
-
yes, it isn't good to rely on.
However, it is very helpful when researching things in this Peat sphere and sciences. One reason being if you're paying attention you must question the ai about its assumptions. Since the Peatsphere is so often contrarian or just "empirical," this attitude has to be brought to the ai. It must always be questioned, and sometimes it is still wrong.I was never a good chemistry or biology student, that's for sure. And I understand there are dogmas that need to be questioned. But before getting into assumptions and dogmas, there are some basic facts that the AIs will communicate, just things like the names of enzymes or proteins or hormones. All of this is very technical so it's nice to have the ai on hand more as a dictionary or reference. But you have to be very specific sometimes.
Obviously I'm not going to automatically believe an ai pulling sources from wiki, or random articles that may not have strong evidence or sources. But it helps to create a general picture, and in the case of drug or vitamin interactions and all of this biological stuff, it's very interesting and helpful.
Again, it cannot tell us anything beyond what is presently available (if its even looking at all of the papers and books that are, which they aren't). We don't know what we don't know. But what we know might be nonsense or half-true at best, as Peat often showed. That's the crux of the matter. It can provide information, but we basically have to construct a narrative or picture of what we're trying to learn. The ai does not understand any story / picture beyond what is often "the case." There's probably some Pareto distribution principle guiding the ai's ability to affirm platitudes and ignore problems and questions present in the knowledge relevant to the query.
