Any alternatives to Perplexity? I've been using it for my stem degree, where I'm required to reference sources of pretty much everything I write in essays.

Perplexity was helpful at finding the right articles efficiently and summurizing them in bullet points, while placing inline citations to the sources it grabbed info from.

However, I had to constantly check for mistakes and redundant information in the output text, which I've been okay with, for the most part, because it's already doing half of the work for me and I'd rather rewrite things myself.

Since I've been using it, there has been crazy amounts of AI models constantly being released trying to top eachother off.

I wanna know from the tech experts here in /g/ if there's a better alternative to Perplexity in fetching science articles (like Pubmed, ScienceDirect, etc) and providing more accurate and less repetitive information about them. I don't really care about how quick or slow it outputs an answer, so long as the quality is there.

I've also just started using ai for my thesis and notebooklm has been helpful so far but you have to upload the file yourself, it can't search for you as far as I know. It gives me better info than most though. I just copy its citations and elaborate using my own words, no need to browse through super long texts anymore. Might be worth trying for you if you're unsure about perplexity's output

>>106137097
Is there a limit to the amount of uploading you're able to do with notebooklm?

>>106136971 (OP)
KYS zoomer and then do it yourself without AI

>>106137214
who hurt you lmao

>>106137214
>dont use tools that massively help your goal

Sure, anon why dont you grow your own food instead of getting it McDonalds

>>106136971 (OP)
ChatGPT release a new Study and Learn feature, maybe try that

>>106136971 (OP)
felo ai

>>106137773
I think this might be what I'm looking for, thank you!

>>106137495
>Comparing it to food
You're paying to learn dummy. That's like paying a private tutor to teach you math and then doing all the practice problems with a calculator. The point of the problems was for you to learn the process. The point of writing essays is for you to learn how to research a topic, formulate the opinion and then express it persuasively. When you get AI to write it for you, you rob yourself of the experience. You also don't develop your own voice. If you were just using it for your job I wouldn't blame you, but using it for your education is just robbing yourself.

>>106138071
OP here. Personally I don't use AI as a calculator nor do I let it do shit for me to the point where I copy-paste the answers. I just use it as a search browser that gets me quick and general information in a neat text with the source material attached. If I like the brief text provided by the AI, I'll read the article myself. That's pretty much it. Searching for articles in Pubmed and other databases of science articles is extremely tedious because I want to find an answer to a specific question that would take hours to find by looking through the top 50 articles, when AI can just answer me in seconds, most importantly with the answer backed up with the citation - which is the most important thing for me.

>>106138506
Fair enough I thought you were one of those copy paste retards.

>>106138642
Totally understandable, I have colleagues that do copy-paste, and you can imediately notice from miles away when it was written by AI. They always get away with it because teachers cba to verify if AI wrote it, they're just focused on if things are correct.

If anything it lets you know which people in class you can rely on vs ChatGPTards. They'll always get hurt in the end because you'll eventually have to do presentations and if you don't know beyond what AI wrote for your script, you're fucked.

>>106138071
problem is that "going through the process" is mostly worthless for writing essays or "scientific" articles.

>>106138071
How is learning how to answer the problem the way people actually do it a bad thing? There's nothing wrong with using a calculator to solve math problems, you're only cheating yourself if you copied the answer from somewhere.

>>106140616
>Going through the process is worthless
Its literally the only thing that makes your paper worth reading. If an AI wrote your paper I have no clue whether or not anything it says is true. I don't know if anything you are saying holds weight. The point of going through the process is to show the results of your work and synthesize the results into a conclusion that is actionable. Ironically morons like you will be the death of AI as it consumes your generated slop paper that doesn't have any substance.
>>106140681
>How is answering the problem the way people actually do things a bad thing
Because you didn't master the skill. The act of doing the work stimulates your brain and makes the knowledge stick with you. If you just use a calculator then you'll have a harder time retaining the information. If you don't retain the information you'll have a harder time building off that information. Work that is valuable tends to build off preestablished things, learning the basics is the first step to get there.

>>106140835
The skill is learning how to do the work in the real world. If real world engineers did all their math with pen and paper, you'd be stupid to use a calculator. But they don't, so you're just a boomer retard.

>>106140835
>Its literally the only thing that makes your paper worth reading.
nobody is gonna read your paper. you're not going to learn anything by doing it, except for how to conform to ideological bullshit and agenda pushing. the "work" is worthless.
I never use AI for scientific articles I write in uni. I don't use AI at all in fact. But I'm not so blind and naive either to the fact that the "scientific" process is completely broken and that universities have become nothing more than virtue-signalling propaganda machines.

>>106140872
You missed the nuance of what I was saying. Nothing is wrong with taking a shortcut when you already know how to do something its just important you understand the process before taking it, because its important to have a strong and adaptable mental model of the subject you wish to work with. If you just use calculators you'll never gain the insights to truly be great.
>>106140902
>Science is equal to politics
Science didn't break down the people doing it just got lazy. The answer is not to become a bad apple, but instead to keep doing things the right way and let the bad apples rot. People still read the papers as they are extremely valuable for solving specific problems. You need to stop generalizing outliers.

>>106140978
science has broken down. people didn't get lazy (well, they did, but that's not the reason why science has broken down). politics and bad incentives have corrupted the "scientific process". the selection has become such that it's only possible to be or stay in "academia" as a bad apple. if you're interested in truth, you won't get subsidies, you won't get job or project opportunities, and you'll get fired for "unrelated" reasons. this is the norm now. the few good articles are the outliers.
sounds like you're in a politically irrelevant (niche) field that hasn't been corrupted by politics too much yet, but the social sciences and humanities are completely gone and increasingly the same is happening in the more technical fields.

>>106141223
>Science broke down
Science can only "break down" if there was something inherently wrong with the scientific method. People doing science wrong and calling it science is just political opportunist larping as scientist.
>Only possible to stay in academia as a bad apple
Depends on the field and what you are researching. If you are studying differences in races you are absolutely fucked because of dogmatic beliefs that need to be asserted true, but if your studying something unrelated to politics you are fine. Social science is fucked because social science is a great justification for policy change. So there is a lot of money in social science for finding conclusions that support policy positions politicians want. I remember seeing a grant to develop a system that could spray plants with just the right amount of water to allow them to grow without wasting water. There were no politics involved at all it was just there to increase agricultural efficiency.

>>106140835
If you learned to do it the wrong way, you didn't master the skill. You faggots are why college degrees don't mean anything anymore. In your obsession with doing things the "right" way, you forgot to teach anyone how to actually do the work.

>>106141750
>People doing science wrong and calling it science is just political opportunist larping as scientist.
that's 95% of "science" nowadays
>Depends on the field and what you are researching.
yes, but pretty much every field is affected to some degree. if I look at the bachelor programs of the universities in my country, every single one has some woke bullshit about "sustainability", "equity", "welfare", "global warming", "harm", or some other buzzword as their description.
>There were no politics involved at all it
kek, that's naive and blind. there's a huge push to eradicate farmers and meat, and make everyone dependent on trillion $ megacorps for "food".

>>106137097
NotebookLM is really good, it helped me a lot to speedup some parts of my thesis. You can use it to check if you understood things correctly too

>>106136971 (OP)
LLMs are essentially word shufflers. They can't verify sources or fact-check. There is no way around it.

>>106141807
>Wrong way
Doing mental math is not the "wrong way" its just a different way. It has value. Do you really want to be the guy that has to pull out a calculator to figure out what 64 + 32 is? A strong mental model is the key to innovation. Everyone I knew who thought like you ended up being a loser because they refused to understand the value of learning something that wasn't immediately useful to their lives. When it started to matter later on in life they all said the same things.
>I wish I tried harder
>I wish I had take my studies more seriously
>I sure was dumb back in the day.
You can take the advice or leave it, but you'll be the one suffering the consequences.
>>106141843
>naive
No they were genuinely seeking truth without letting politics get in the way. What they used the truth for is a different matter. They were in fact doing science correctly.

>>106142038
it's impossible for politics to not get in the way. if every researcher who is interested in pursuing the "wrong" truths gets filtered out, you can still have researchers interested in the "right" truth. there are genuine woke vegan idiot scientists. they are not the problem per se. the real problem is every scientist interested in other things getting banned from pursuing that.

>>106142094
>Banned from pursuing it
They aren't they just aren't funded. Which I agree is a problem, but doesn't mean science has failed. Science is a tool that is used to uncover truths, if someone prevents you from using the tool it doesn't mean the tool itself is broken.

>>106142153
in many western countries they're now actually banned from pursuing it. e.g. the UK, Canada, Germany, Australia, New Zealand, etc. in other they're not getting any funding/opportunities, which almost as big of a problem as bans.
what matters is how "science" is being done in practice, not that the scientific method is sound in theory. and "science" in practice is nothing more than a political and corporate tool nowadays.

>>106142185
>Nowadays
Unfortunately this has always been the truth of science. The only things that get funding are things that are useful for those in power. In the US we still are allowed to pursue and talk about it, but goof luck getting funding. That you'll have to do on your own.

>>106142231
>Unfortunately this has always been the truth of science.
to some degree, yes, but it's gotten way worse since about 2010

>>106136971 (OP)

>>106141860
Dude same, it's helping me quite a lot as far as checking secondary literature goes which is typically my least favorite part about academic writing.
The ui is awesome, I like the concept of having a short summary of relevant quotes in the middle and then the side bar for full source context, makes everything so much more reliable and precise than the typical chatbot nonsense that tends to hallucinate and give me irrelevant info I did not ask for. My prompt is typically to just formulate a specific chapter of my thesis into a question.
Then all I do is copy paste the most important quotes into my thesis and elaborate using my own words. I'd definitely not let AI write my actual thesis though. They're just not good enough in understanding the actual material in my experience. But they definitely make research easier.