They are all far from being able to do a proper high research level analysis. What you were doing was fairly exploratory. Next time try asking the AI to perform a multivariate analysis or explore multiple generalized regression models. None will be able to do it quicker than a well trained analyst.
@yellscooking8975 сағат бұрын
Elsevier Jounal Finder is so good 👍 to find THE HOME!! Thank you 👍
@margaretzaitoun90078 сағат бұрын
cant work with WPS? which microsoft can i use for it
@5pm_Hazyblue8 сағат бұрын
Dont listen to the quitters. They quit anyway. All they do is just make excuses.
@VizlyLabs9 сағат бұрын
Hi Andy! Thanks a lot for making a video about Vizly. I believe the wrong logo was used during the video when describing Vizly, if anyone is interested in trying the tool, please look on Google for Vizly!
@TrevorRockwell9 сағат бұрын
The logo for Visily is incorrect, the correct name for the tool seems to be Vizly
@therese_paula10 сағат бұрын
Which research/notes organizing tool do you use, good 'ol Word? Can you let us have a glimpse of your actual writing process using that tool please? Thank you :)
@gopika1310 сағат бұрын
I have a useless PhD degree. And I failed in my career and life.
@mackss946810 сағат бұрын
How about working for nonprofits?
@jamalnuman11 сағат бұрын
Does the literature review need to be divided into sections to answer the research questions, say 3?
@YvonneCapeles11 сағат бұрын
Hello, your videos have helped me start my literature review but being bilingual and putting everything in writing and editing is a struggle. Where or how would I look for a good editor to help me with that? Any recommendations will be greatly appreciated.
@shawnb474519 сағат бұрын
I do this during my lit reviews for papers.
@dmitry531919 сағат бұрын
Fully subscribe to what was said here. I would also add a couple of things 1. Considering how tough it is and how little it pays, there are, essentially, no reasons to do research except proving that you are the cleverest person ever lived. As a result academia is full of people with some sort of narcissistic disorder. Everyone is trying to prove that they are better than others, find excuses why they are not so successful etc. Narcissism also implies no empathy. 2. Academia is very unfair. Besides things mentioned in the video, connections of your supervisor and prestige of you university matter a lot. You may well do exactly the same stuff as other group from a more prestigious place and be, essentially, ghosted: not invited to conferences, not cited, etc. Just because to them you are no one. They would not even bother answering your emails. 3. Strong hierarchy. A lot depends just on your age. Even with the same productivity and quality of research, the older you get the more papers you have written and the more time they had to be cited/known. As a result, older people have by far more recognition and power. Even at the age of 40 nobody takes you seriously even if you outperform everyone of your age. You still have to rely on the support from more senior researchers and they often treat you in condescending manner, even if they haven't produced anything within last 20 years.
@erickgomez777520 сағат бұрын
What about a phd on machine learning and statistics with a focus in solving physics? Its like the combination of cs, maths, statistics, and physics
@carlosmspk22 сағат бұрын
I only went as far as a master's thesis, but one thing that I was frustrated about was that my supervisors effectively prohibited me from including "bad" results. That is to say, results that either disprove or exhibit shortcomings on the conclusion you hoped for. To begin with, you shouldn't even "hope" for a conclusion, you should just come up with a thesis (duh) and see how it holds. If the results show that the thesis doesn't hold, why isn't that an acceptable outcome?
@wagnernunesribeiro9559Күн бұрын
Thank u
@lenhatquangkhoi5819Күн бұрын
Wow, didn't expect to see my go-to channel for improving my analysis skills, also did some research on the same field and materials as I am too. Can we discuss more via email?
@KLAUZVКүн бұрын
complete video pls
@ekwanzalamdКүн бұрын
I fully agree with most of the video, except the part about not reading the method section. One of the main reasons we publish is to allow others to replicate or build on our work. I think you should advise PhDs to take extra time in reading this section. With this advice in mind, it is no wonder why most experiments cannot be replicated.
@zimmejoc15 сағат бұрын
Being a methods guy, I love it when people skip the methods. That means when they have their own data to analyze, they come to me and write the methods and results. They do the hard part, I crunch the numbers. I get a pub. You're right though, you should absolutely read the methods. I read it right after then abstract. If the methods suck, the findings are immediately suspicious.
@gonzoz1Күн бұрын
Jullius has the option in its paid model of using ChatGPT, Claude and/or Opus. So why have the middleman ? Just stick with ChatGPT.
@dear.momo786Күн бұрын
Dude telling you what not to do by showing you what exactly you need to do ;) Lit!
@lukejay5460Күн бұрын
Pretty easy to read your own paper just saying. BUT you should do a paper that you would mark up and need to know as well as who ever did the paper but one that isn't your own.
@VeyedeoКүн бұрын
This is fantastic. I guess Elsevier thinks we are just going to figure this out, but it is so quirky and buggy, as you point out, that it is hard to know what it is doing half the time. Thanks very much for explaining it. I thought my system config was all wrong. It turns out it is very useful, but bad software that requires patience and a persistence to see what it loaded correctly and what it didn't.
@LeifOfWilsonCreekКүн бұрын
You're shirt is so loud I can't hear your abstract! 😜
@firstname4337Күн бұрын
no need to read a paper -- just tell ChatGPT to summarize it for me
@grahamashe9715Күн бұрын
Unless you’re from one of the top ten universities in the world, nobody (especially journalists who report on science) gives a toss about what you write anyway. It’s automatically considered lacking in credibility or to be taken with a grain of salt. If more academics realised this, fewer people would waste their life and earning potential there. Not to mention sanity.
@boredscientist5756Күн бұрын
This is so wrong, it only depends on the field and the team. MIT or Stanford would have no chance in my field, they are barely peasants. I now work in a top 8 Univ, and trust me, they are very far from my 2 previous teams...VERY far. You can have "low" rank universities who are world leaders in a specific field, I have many examples! The ranking is MEANINGLESS. I really doubt that you know how researchers work together. Even NASA worked with my 2 previous teams (THEY came to us, because we were just the best in the world)... Same goes for Cambridge or MIT. To get my current position, it was uberly easy considering my background in a "low" ranked university, go figure. People should stop taking QS rankings too seriously, and realize that good researchers do not give a damn.
@grahamashe9715Күн бұрын
@@boredscientist5756 The public only cares about the names of institutions they recognise (and keep hearing) and journalists know this too. Don’t even get me started on non-Western institutions which have basically ZERO credibility even in the minds of their own people.
@souravdutta8813Күн бұрын
I couldn't find a way to upload data in Visily. Did you mean any other AI tool instead of Visily?
@souravdutta8813Күн бұрын
Found the correct one. It is Vizly. Thanks!
@flingnflyКүн бұрын
Sorry if I missed it - but what pdf markup software are you using here? I will re-watch to try figure it out
@flingnflyКүн бұрын
Foxit PDF reader
@steverundus2878Күн бұрын
Fantastic tips. Thanks for posting. What software do you typically use to markup PDFs and save them for future reference? Not everyone has adobe acrobat pro. Do you use "snip and sketch" to cut and paste the figures into PPT file as mentioned at the end of your video?
@keerthanrrao579Күн бұрын
What’s really annoying is the acronyms. Most of them are unnecessary.
@boredscientist5756Күн бұрын
????? All acronyms are introduced once.... It makes the reading painful otherwise.... It is the norm....
@keerthanrrao579Күн бұрын
@@boredscientist5756 yes. But as I mentioned, most of them are not necessary. Also, maybe a table would be better so there’s one place for it.
@boredscientist5756Күн бұрын
@@keerthanrrao579 specialists know the acronyms by heart .... I don't see any problem
@boredscientist5756Күн бұрын
@@keerthanrrao579 most of them ARE necessary, are you still a student? For us researchers, SEM, EDS, TEM, XRD, TOC, DSC, TMA, NMR, FTIR, DCDC etc makes total sense, and it would be totally ridiculous to write the full name more than once. You just make no sense......
@keerthanrrao579Күн бұрын
Well, some papers you need to dig deeper than the abstract. As most abstracts are exaggerated to a point It it’s almost unethical.
@olukunleolawole7012Күн бұрын
Awesome
@AumiNadimКүн бұрын
Can you do a video on business or marketing related paper?
@patrickokafor-yt9oyКүн бұрын
I like!
@davidmacfarlane1763Күн бұрын
There are alternatives to academia for a Phd, depending on your discipline. I entered Pharmaceutical manufacturing with my PhD in chemistry (and post grad diploma in Bioinformatics). Its rewarding and there is a lot of work
@fn6679Күн бұрын
Thanks sooooooooooo much for sharing your experience. And for your advice
@nthmostКүн бұрын
Just posting data doesn't solve the reproducibility problem though. Large and very-large datasets require hosting (at best it's a cost to bear; at worst, there are many situations, outside the scope of a YT comment, where institutions restrict access to that data). Worse still, datasets have to be processed, which means code needs to be run. Sure, you can post the code on github, but there are innumerable situations where the code as published no longer works due to its underlying libraries shifting or even going defunct. (Also, as a software engineer, I can tell you honestly, code produced by academics is unfortunately rather stinky.) So even just re-running the data under the same conditions in order to see the experimental results can be nearly impossible. And it will only grow more difficult with time as software changes. I was part of the NIH Data Consortium that sought to try to do something about this. As a 600-person consortium, of course, it failed in the Pilot stage. 😀
@deepthisdk7562Күн бұрын
important. Thank u so much
@247AI-sp5zzКүн бұрын
I think a PhD will become a Bachelor's Degree in the coming 10-20 years, it won't be worth like today but still worth it for a Career 😂
@MeTravel247Күн бұрын
Hi Andy, can you please recommend some Universities for doing a PhD on scholarship in the Data Science area? If a person completes their MSc IT in Business Data Analytics from IBS Budapest Hungary. Thank You 🙏
@MeTravel247Күн бұрын
Hi Andy, can you please share some information if someone is planning to do a PhD in the Computer Science area after completing their Masters from Budapest Hungary? Please recommend Universities that are really great 👍 and they offer the best scholarships.
@screew708Күн бұрын
7:30 is so accurate I had to chuckle thinking about my supervisor posing for pictures next to my experiment.
@jerichodejesus3047Күн бұрын
I normally paraphrase my content to Undetectable AI to make it more undetected
@GokhanYusuf-gk1bg2 күн бұрын
Awesome stuff
@SerkanElif2 күн бұрын
I’ve been struggling with this for weeks, and this video finally cleared things up. Thank you!
@AhmetMelda2 күн бұрын
Does anyone know if these techniques work on older versions?
@SirinUfuk2 күн бұрын
This was super helpful, thanks! Subscribed for more.
@NezaketMubeccel2 күн бұрын
I tried using AI humanizers/bypassers, but Winston AI can detect those tools effectively. Better stick to genuine methods!
@Yeldaibrahim2 күн бұрын
Amazing content as always. You never disappoint!
@SelahattinDilara2 күн бұрын
Great tutorial! This is exactly what I needed to get started.
Пікірлер
They are all far from being able to do a proper high research level analysis. What you were doing was fairly exploratory. Next time try asking the AI to perform a multivariate analysis or explore multiple generalized regression models. None will be able to do it quicker than a well trained analyst.
Elsevier Jounal Finder is so good 👍 to find THE HOME!! Thank you 👍
cant work with WPS? which microsoft can i use for it
Dont listen to the quitters. They quit anyway. All they do is just make excuses.
Hi Andy! Thanks a lot for making a video about Vizly. I believe the wrong logo was used during the video when describing Vizly, if anyone is interested in trying the tool, please look on Google for Vizly!
The logo for Visily is incorrect, the correct name for the tool seems to be Vizly
Which research/notes organizing tool do you use, good 'ol Word? Can you let us have a glimpse of your actual writing process using that tool please? Thank you :)
I have a useless PhD degree. And I failed in my career and life.
How about working for nonprofits?
Does the literature review need to be divided into sections to answer the research questions, say 3?
Hello, your videos have helped me start my literature review but being bilingual and putting everything in writing and editing is a struggle. Where or how would I look for a good editor to help me with that? Any recommendations will be greatly appreciated.
I do this during my lit reviews for papers.
Fully subscribe to what was said here. I would also add a couple of things 1. Considering how tough it is and how little it pays, there are, essentially, no reasons to do research except proving that you are the cleverest person ever lived. As a result academia is full of people with some sort of narcissistic disorder. Everyone is trying to prove that they are better than others, find excuses why they are not so successful etc. Narcissism also implies no empathy. 2. Academia is very unfair. Besides things mentioned in the video, connections of your supervisor and prestige of you university matter a lot. You may well do exactly the same stuff as other group from a more prestigious place and be, essentially, ghosted: not invited to conferences, not cited, etc. Just because to them you are no one. They would not even bother answering your emails. 3. Strong hierarchy. A lot depends just on your age. Even with the same productivity and quality of research, the older you get the more papers you have written and the more time they had to be cited/known. As a result, older people have by far more recognition and power. Even at the age of 40 nobody takes you seriously even if you outperform everyone of your age. You still have to rely on the support from more senior researchers and they often treat you in condescending manner, even if they haven't produced anything within last 20 years.
What about a phd on machine learning and statistics with a focus in solving physics? Its like the combination of cs, maths, statistics, and physics
I only went as far as a master's thesis, but one thing that I was frustrated about was that my supervisors effectively prohibited me from including "bad" results. That is to say, results that either disprove or exhibit shortcomings on the conclusion you hoped for. To begin with, you shouldn't even "hope" for a conclusion, you should just come up with a thesis (duh) and see how it holds. If the results show that the thesis doesn't hold, why isn't that an acceptable outcome?
Thank u
Wow, didn't expect to see my go-to channel for improving my analysis skills, also did some research on the same field and materials as I am too. Can we discuss more via email?
complete video pls
I fully agree with most of the video, except the part about not reading the method section. One of the main reasons we publish is to allow others to replicate or build on our work. I think you should advise PhDs to take extra time in reading this section. With this advice in mind, it is no wonder why most experiments cannot be replicated.
Being a methods guy, I love it when people skip the methods. That means when they have their own data to analyze, they come to me and write the methods and results. They do the hard part, I crunch the numbers. I get a pub. You're right though, you should absolutely read the methods. I read it right after then abstract. If the methods suck, the findings are immediately suspicious.
Jullius has the option in its paid model of using ChatGPT, Claude and/or Opus. So why have the middleman ? Just stick with ChatGPT.
Dude telling you what not to do by showing you what exactly you need to do ;) Lit!
Pretty easy to read your own paper just saying. BUT you should do a paper that you would mark up and need to know as well as who ever did the paper but one that isn't your own.
This is fantastic. I guess Elsevier thinks we are just going to figure this out, but it is so quirky and buggy, as you point out, that it is hard to know what it is doing half the time. Thanks very much for explaining it. I thought my system config was all wrong. It turns out it is very useful, but bad software that requires patience and a persistence to see what it loaded correctly and what it didn't.
You're shirt is so loud I can't hear your abstract! 😜
no need to read a paper -- just tell ChatGPT to summarize it for me
Unless you’re from one of the top ten universities in the world, nobody (especially journalists who report on science) gives a toss about what you write anyway. It’s automatically considered lacking in credibility or to be taken with a grain of salt. If more academics realised this, fewer people would waste their life and earning potential there. Not to mention sanity.
This is so wrong, it only depends on the field and the team. MIT or Stanford would have no chance in my field, they are barely peasants. I now work in a top 8 Univ, and trust me, they are very far from my 2 previous teams...VERY far. You can have "low" rank universities who are world leaders in a specific field, I have many examples! The ranking is MEANINGLESS. I really doubt that you know how researchers work together. Even NASA worked with my 2 previous teams (THEY came to us, because we were just the best in the world)... Same goes for Cambridge or MIT. To get my current position, it was uberly easy considering my background in a "low" ranked university, go figure. People should stop taking QS rankings too seriously, and realize that good researchers do not give a damn.
@@boredscientist5756 The public only cares about the names of institutions they recognise (and keep hearing) and journalists know this too. Don’t even get me started on non-Western institutions which have basically ZERO credibility even in the minds of their own people.
I couldn't find a way to upload data in Visily. Did you mean any other AI tool instead of Visily?
Found the correct one. It is Vizly. Thanks!
Sorry if I missed it - but what pdf markup software are you using here? I will re-watch to try figure it out
Foxit PDF reader
Fantastic tips. Thanks for posting. What software do you typically use to markup PDFs and save them for future reference? Not everyone has adobe acrobat pro. Do you use "snip and sketch" to cut and paste the figures into PPT file as mentioned at the end of your video?
What’s really annoying is the acronyms. Most of them are unnecessary.
????? All acronyms are introduced once.... It makes the reading painful otherwise.... It is the norm....
@@boredscientist5756 yes. But as I mentioned, most of them are not necessary. Also, maybe a table would be better so there’s one place for it.
@@keerthanrrao579 specialists know the acronyms by heart .... I don't see any problem
@@keerthanrrao579 most of them ARE necessary, are you still a student? For us researchers, SEM, EDS, TEM, XRD, TOC, DSC, TMA, NMR, FTIR, DCDC etc makes total sense, and it would be totally ridiculous to write the full name more than once. You just make no sense......
Well, some papers you need to dig deeper than the abstract. As most abstracts are exaggerated to a point It it’s almost unethical.
Awesome
Can you do a video on business or marketing related paper?
I like!
There are alternatives to academia for a Phd, depending on your discipline. I entered Pharmaceutical manufacturing with my PhD in chemistry (and post grad diploma in Bioinformatics). Its rewarding and there is a lot of work
Thanks sooooooooooo much for sharing your experience. And for your advice
Just posting data doesn't solve the reproducibility problem though. Large and very-large datasets require hosting (at best it's a cost to bear; at worst, there are many situations, outside the scope of a YT comment, where institutions restrict access to that data). Worse still, datasets have to be processed, which means code needs to be run. Sure, you can post the code on github, but there are innumerable situations where the code as published no longer works due to its underlying libraries shifting or even going defunct. (Also, as a software engineer, I can tell you honestly, code produced by academics is unfortunately rather stinky.) So even just re-running the data under the same conditions in order to see the experimental results can be nearly impossible. And it will only grow more difficult with time as software changes. I was part of the NIH Data Consortium that sought to try to do something about this. As a 600-person consortium, of course, it failed in the Pilot stage. 😀
important. Thank u so much
I think a PhD will become a Bachelor's Degree in the coming 10-20 years, it won't be worth like today but still worth it for a Career 😂
Hi Andy, can you please recommend some Universities for doing a PhD on scholarship in the Data Science area? If a person completes their MSc IT in Business Data Analytics from IBS Budapest Hungary. Thank You 🙏
Hi Andy, can you please share some information if someone is planning to do a PhD in the Computer Science area after completing their Masters from Budapest Hungary? Please recommend Universities that are really great 👍 and they offer the best scholarships.
7:30 is so accurate I had to chuckle thinking about my supervisor posing for pictures next to my experiment.
I normally paraphrase my content to Undetectable AI to make it more undetected
Awesome stuff
I’ve been struggling with this for weeks, and this video finally cleared things up. Thank you!
Does anyone know if these techniques work on older versions?
This was super helpful, thanks! Subscribed for more.
I tried using AI humanizers/bypassers, but Winston AI can detect those tools effectively. Better stick to genuine methods!
Amazing content as always. You never disappoint!
Great tutorial! This is exactly what I needed to get started.