[00:01.270 --> 00:02.130]  Hello, all.
[00:02.670 --> 00:09.710]  I'm pleased to introduce Mu, who will be talking about a sleuth story on detecting and revealing large-scale research fraud.
[00:10.290 --> 00:11.450]  We're excited to have you.
[00:18.390 --> 00:19.510]  Thank you, everyone.
[00:20.470 --> 00:21.850]  So, I'm a sleuth.
[00:22.110 --> 00:27.290]  I've been doing this for the... So, I had to put up a disclaimer slide just to cover my base.
[00:27.810 --> 00:30.410]  This is probably the most lame slide here.
[00:30.730 --> 00:33.390]  What it's saying is basically I have no skill of hacking.
[00:33.550 --> 00:34.490]  I didn't hack anywhere.
[00:34.490 --> 00:36.730]  That is basically what it says.
[00:38.090 --> 00:38.150]  Yeah.
[00:39.310 --> 00:43.130]  So, about me, I'm a scientist at Columbia University.
[00:43.130 --> 00:48.930]  I do the mouse preclinical assessment disease models.
[00:49.590 --> 01:00.150]  So, until about five years ago, I was your run-of-the-mill Ivy League scientist who was obsessed with publishing papers and being on papers and all that.
[01:02.010 --> 01:08.630]  So, since 2020, I have been working as a sleuth, as a side hustle.
[01:08.870 --> 01:15.750]  I have caused over 240 retractions, which are your peer-reviewed scientific research papers.
[01:15.750 --> 01:17.970]  And I also got a book retracted.
[01:18.070 --> 01:18.970]  We'll talk about that.
[01:19.490 --> 01:22.890]  And I am very active on this thing called Pug Peers.
[01:23.050 --> 01:30.230]  Pug Peers is basically our community where people flag fraudulent... No, potentially problematic research paper.
[01:30.350 --> 01:31.790]  I cannot use the word fraudulent.
[01:33.750 --> 01:36.910]  So, when we talk about peer-reviewed research, right?
[01:37.050 --> 01:37.150]  So, when we talk about peer-reviewed research, right?
[01:37.150 --> 01:43.470]  What we're talking about, those are typically papers published by some of these major publishers.
[01:43.790 --> 01:49.950]  So, Elsevier is the biggest one, followed by Springer, Nature, Wiley, and Taylor Francis, MDPI.
[01:50.050 --> 01:53.410]  So, these are the major publishers for scientific journals.
[01:54.150 --> 01:59.950]  And then, here is one of the studies by my sleuth friends.
[02:00.350 --> 02:01.930]  This just came out last week.
[02:02.050 --> 02:12.010]  And so, what it is showing is that the rate of total number of scientific publication doubles at the rate about 15... every 15 years.
[02:12.010 --> 02:14.110]  Paper mill products.
[02:14.310 --> 02:14.310]  Paper mill products.
[02:14.310 --> 02:22.290]  So, these are industrial products, man-made, that they look like scientific research paper, but they are not.
[02:22.510 --> 02:27.290]  They're basically... there is a cartel network that trade...
[02:27.290 --> 02:28.130]  It is a business.
[02:28.590 --> 02:32.090]  So, these are fake paper in the scientific publication.
[02:33.070 --> 02:35.490]  But it's very hard to tell sometimes.
[02:36.230 --> 02:40.310]  These paper double their rate of every 1.5 years.
[02:40.310 --> 02:48.770]  So, which is like 10 times faster than the total amount of scientific publications, making it very concerning.
[02:48.770 --> 02:52.250]  They're doubling at the rate of virus, basically.
[02:53.410 --> 02:54.670]  So, all right.
[02:55.710 --> 03:01.390]  So, today we're talking about many interesting light cases.
[03:01.610 --> 03:07.790]  I call them light cases because a lot of studies are still being investigated, which is why I need to be careful what I'm saying.
[03:08.750 --> 03:12.050]  So, first we'll talk about this guy, Elisa Maslia.
[03:12.250 --> 03:20.730]  He used to be the director for Neuroscience at National Institute of Aging, which is the largest institute for Alzheimer's research.
[03:20.730 --> 03:28.970]  And we'll talk about COVID vaccine, plastic collar for football players, YouTube, all this stuff.
[03:29.730 --> 03:31.630]  So, this one.
[03:31.790 --> 03:37.010]  So, this is this article from Science Magazine that came out about a year ago.
[03:37.230 --> 03:44.230]  So, this guy, he had a top job at National Institute of Aging.
[03:44.230 --> 03:49.790]  So, they have a budget of $2.6 billion in the last fiscal year.
[03:49.990 --> 03:54.890]  It is the largest funding agency for Alzheimer's research globally.
[03:55.350 --> 04:03.110]  He had over 800 papers and we flagged over 130 as potentially problematic.
[04:03.730 --> 04:13.810]  And his research, because he's a very, he's a top, he's one of the top researchers and he was also in charge of huge amount of funding monies from NIH.
[04:14.990 --> 04:17.530]  His study was highly influential.
[04:17.950 --> 04:28.590]  And since last year, since the paper came out last year and since we reported his paper to the NIH, over 10 have been retracted.
[04:28.590 --> 04:30.630]  These are some of the retracted papers.
[04:31.790 --> 04:39.570]  So, I just want to mention, this case was describing the story in the book by Charles Peeler.
[04:39.990 --> 04:41.710]  And you can check it out.
[04:42.010 --> 04:54.070]  And I was the lead on the Maslia case and other sleuths, Elizabeth Beek, Matthew Schreck, Kevin Patrick, they were the lead sleuths on other cases describing in the book.
[04:54.070 --> 04:59.810]  So, as you, same as your community, we're also collaborating in many cases.
[05:00.590 --> 05:06.990]  So, disclosure, the tool I use, I use very little tools, but this is one of the major tools I use.
[05:07.190 --> 05:09.350]  It is called Image Twin AI.
[05:10.410 --> 05:16.390]  It is basically, it's very hard for detecting, like, image duplication and stuff.
[05:16.390 --> 05:20.910]  So, now I will start to show studies.
[05:21.550 --> 05:26.630]  So, these are, this is, what you're seeing is brain histologies.
[05:27.210 --> 05:36.690]  So, what you're seeing is the same color boxes indicate the sections that are identical between these two experimental groups.
[05:36.690 --> 05:52.190]  So, basically, what you're seeing here is that, disregard what the researchers are describing, they are using the same tissue images over and over again to represent different treatment groups.
[05:52.330 --> 05:54.290]  And then this is definitely not okay.
[05:56.070 --> 06:02.430]  Same here, you can see line one and two are basically two type of, two different lines of mice.
[06:02.430 --> 06:08.050]  And then they cannot possibly have the same histological images, but they do.
[06:08.330 --> 06:10.230]  So, there's something not right there.
[06:10.970 --> 06:17.730]  And this is a particularly concerning case, as you can see this one, if I could point.
[06:17.950 --> 06:21.530]  This one was from the animal treated with two drugs.
[06:21.970 --> 06:24.790]  And then this one is from a control animal.
[06:25.130 --> 06:32.350]  So, you cannot claim that your treatment of two drugs works as well as when the animal is normal.
[06:34.510 --> 06:36.190]  And too far.
[06:36.670 --> 06:51.550]  Similar here, you can see that it's basically the label and, you know, when you have two studies, this is the same histology image using two studies and then the labels do not match what they represent.
[06:51.970 --> 06:56.330]  And then just like how do we even know any of them are real anymore?
[06:57.910 --> 06:59.070]  Same here.
[06:59.610 --> 06:59.770]  Same here.
[06:59.770 --> 07:03.190]  Non-TG, NTG is basically non-transgenic.
[07:03.350 --> 07:05.350]  This means like this is normal animals.
[07:05.650 --> 07:07.590]  And this is APP.
[07:07.910 --> 07:11.150]  This is an Alzheimer animal treated with something.
[07:12.950 --> 07:16.470]  So, this is a particularly concerning slide.
[07:16.690 --> 07:18.170]  You can see non-TG again.
[07:18.370 --> 07:21.610]  This is supposedly from a normal animal.
[07:21.790 --> 07:28.010]  And this is supposedly from an Alzheimer mouse treated with this drug called cerebral lysine.
[07:28.130 --> 07:32.890]  As you can see, for the most part, these two slides are identical.
[07:32.890 --> 07:35.990]  Where these two dots came from, we don't know.
[07:36.130 --> 07:37.030]  But they're not there.
[07:40.110 --> 07:47.110]  So, this is a study where they reuse the same axon image in one animal.
[07:47.390 --> 07:50.130]  One study was in mice and one study in rats.
[07:50.430 --> 07:51.650]  It doesn't make any sense.
[07:53.650 --> 07:58.690]  Similarly, this is a reuse of histological images in two studies.
[07:58.970 --> 08:05.910]  And then this sort of stuff, it happens a lot in lower tier journals from paper mills.
[08:06.110 --> 08:12.310]  But when it happens in your top guy at NIH, neuroscience is very concerning.
[08:12.310 --> 08:15.270]  This one has a lot of clone sections.
[08:16.970 --> 08:21.130]  So, you know, this is a cell image, a cellular image.
[08:21.350 --> 08:25.130]  You don't have, you shouldn't have identical section in a biological sample.
[08:26.370 --> 08:27.370]  Same here.
[08:28.070 --> 08:32.250]  You have a lot of like clone images, clone section in an image.
[08:33.770 --> 08:35.870]  This one already being retracted.
[08:36.250 --> 08:46.250]  It's very confusing because it doesn't really do anything for, because when you look at this, these slides here, when we read papers, we don't really look at that.
[08:46.410 --> 08:49.010]  It just, it doesn't really carry any information.
[08:50.390 --> 08:57.690]  Showing, duplicating this section doesn't make the, doesn't strengthen your result, which is, I don't know why people do that.
[08:59.470 --> 09:02.790]  Here, similarly, these are two slides.
[09:03.170 --> 09:06.530]  They were, they had different color and contrast.
[09:06.850 --> 09:10.930]  But again, these two dots are in this image, but not here.
[09:10.930 --> 09:16.830]  If you use a landmark, these are the same dots, but this little black dots is not here.
[09:17.430 --> 09:19.170]  And these two black dots are not here.
[09:20.230 --> 09:26.060]  So, these are, again, same color boxes indicate overlapping images.
[09:26.780 --> 09:30.840]  And then we're, we're looking at them as vehicle being controlled.
[09:30.980 --> 09:34.960]  Cerebral lysing is Alzheimer animal treated with a drug.
[09:35.380 --> 09:42.140]  And then these researchers are using the same histology images to, to represent both groups.
[09:44.880 --> 09:46.800]  Again, these are two studies.
[09:48.220 --> 09:49.520]  Labels are different.
[09:49.520 --> 09:50.520]  And, yeah.
[09:51.980 --> 09:57.800]  So, the problem is that we don't, we don't, like, we don't know what, where to, what to believe.
[09:58.920 --> 10:00.020]  So, here's the case.
[10:00.100 --> 10:01.760]  This is not in the NIH case.
[10:01.960 --> 10:03.300]  I found later.
[10:04.160 --> 10:06.740]  So, there's the square box.
[10:06.940 --> 10:10.980]  You cannot have square in your cells, I assume.
[10:14.100 --> 10:16.320]  Just some duplication there.
[10:16.520 --> 10:18.920]  Here's an enlarged image of a square.
[10:19.120 --> 10:20.760]  It shouldn't have in your cell.
[10:24.200 --> 10:26.330]  And then here is a protein expression.
[10:27.110 --> 10:29.370]  And you can see, this is very clumsy.
[10:29.830 --> 10:32.130]  Here's a copy and pasted edge.
[10:33.110 --> 10:35.310]  Because this is a gel image.
[10:35.550 --> 10:37.310]  It's basically a protein.
[10:37.630 --> 10:40.990]  The signal indicate the level of protein expression.
[10:41.630 --> 10:43.690]  And you can see this edge is sticking out.
[10:46.610 --> 10:49.670]  Here, I'm going to have a enlarged one.
[10:49.870 --> 10:53.390]  So, I use this free software called Forensically.
[10:54.090 --> 11:00.430]  So, you can, if you, if you look at details, you can see they are duplicated.
[11:00.770 --> 11:02.650]  The blue and red are duplicate.
[11:04.170 --> 11:07.510]  So, these are basically somebody did some patchwork.
[11:09.190 --> 11:10.970]  And move on to the next case.
[11:11.010 --> 11:12.350]  We have a lot to talk about.
[11:12.610 --> 11:13.970]  So, this is going to be fun.
[11:14.230 --> 11:17.190]  So, this is a study retracted recently.
[11:17.670 --> 11:27.410]  The study is showing that prenatal exposure to super high dose of COVID vaccine, by Pfizer, caused autism-like behavior in rats.
[11:29.090 --> 11:34.150]  And, Kenneth Owens retweeted some guys talking about this study.
[11:34.390 --> 11:38.450]  And she's like, maybe it's time to listen to all us anti-vaxxers.
[11:38.650 --> 11:41.430]  And stop listening to the big pharma draw a cartel.
[11:41.430 --> 11:42.230]  All right.
[11:42.570 --> 11:43.430]  So, which...
[11:44.170 --> 11:45.350]  So, Kevin Bass...
[11:45.350 --> 11:47.710]  So, this is a Kevin Bass tweet.
[11:48.030 --> 11:50.490]  And this is a study that you can see.
[11:50.810 --> 11:52.450]  That's a study we're talking about.
[11:52.570 --> 11:53.430]  That's being retracted.
[11:55.190 --> 12:08.150]  So, this has added to a lot of fuel to the anti-vax movement in saying that, look, COVID vaccine caused autism in kids or in rats.
[12:09.590 --> 12:11.410]  So, this is also on Twitter.
[12:11.650 --> 12:17.870]  Someone way before me criticized this study for giving the rats extremely high dose of vaccine.
[12:19.310 --> 12:25.670]  If you're a bigger guy and rats are about 250 grams, and that's like 400 times.
[12:25.890 --> 12:28.210]  So, they gave the animals a full dose of it.
[12:29.150 --> 12:34.730]  And the other sleuths indicate that the error bar seemed to be, like, of different length.
[12:34.750 --> 12:37.890]  So, he put a red line of the same length.
[12:38.090 --> 12:42.950]  And then you can see that the top part of the error bar and the bottom part of the error bar don't match.
[12:43.210 --> 12:46.990]  So, meaning that we don't know, like, your data or something wrong.
[12:46.990 --> 12:54.250]  But, even given this evidence, the journal decided that there's nothing to look at.
[12:54.450 --> 13:01.030]  So, this is one of our people posted on PubPier and just saying, like, okay, case closed.
[13:02.290 --> 13:06.200]  And for some reason, the publishers contacted me.
[13:06.200 --> 13:10.980]  So, this is a study used, like, this is a subject in there.
[13:11.120 --> 13:17.120]  If you prefer to spend time with this guy in the cup over here, then you're social and vice versa.
[13:17.120 --> 13:17.920]  You're autistic.
[13:18.280 --> 13:18.960]  That's the rationale.
[13:19.720 --> 13:26.440]  And so, the reason why they contacted me, because it was in my previous life, I wrote this thing.
[13:27.140 --> 13:34.620]  I wrote a methodology paper when I was struggling postdoc at NIH myself, trying to get my green card.
[13:34.620 --> 13:43.060]  So, I wrote about this thing and then they found it out and they asked me for my opinion as an author of this and also as a sleuth.
[13:43.820 --> 13:49.740]  So, I pointed out, that's my alias on PubPier.
[13:49.880 --> 13:56.380]  I pointed out that there's two experiments in this study and then the control data are identical.
[13:57.020 --> 14:08.080]  Again, the biological, like, in biological experiment, it is exceedingly rare to see identical means and SEM, which is Aribar.
[14:09.040 --> 14:11.840]  And then I pointed out this is being...
[14:12.380 --> 14:15.380]  But this is not very strong evidence in my opinion.
[14:15.640 --> 14:19.400]  So, the second one I put out was, I thought it was funny.
[14:19.400 --> 14:23.960]  So, this is a photo from the methodology paper that I wrote.
[14:24.240 --> 14:27.440]  So, this is a chamber of 40 by 20 centimeters.
[14:27.800 --> 14:32.740]  And then in the article that they use is 40 by 30 centimeters.
[14:33.340 --> 14:37.460]  Note that mice are about, like, 7 to 10 times smaller than rats.
[14:37.640 --> 14:48.780]  So, imagine if you have a box that is 50% larger than this one, but you have two animals that are 7 to 10 times bigger than these two guys.
[14:48.780 --> 14:49.940]  Where do you put everybody?
[14:50.460 --> 14:52.340]  Like, is it going to get scooch over?
[14:52.520 --> 14:52.880]  I don't know.
[14:53.020 --> 14:55.080]  That's not a social interaction.
[14:55.780 --> 14:56.300]  That's...
[14:57.380 --> 14:58.660]  I don't know how that works.
[14:59.280 --> 15:01.820]  So, anyhow, it got retracted.
[15:03.240 --> 15:08.020]  So, it was basically the case flipped from case closure retracted.
[15:08.620 --> 15:10.840]  It was just retracted last month.
[15:10.840 --> 15:13.940]  And then because I was...
[15:13.940 --> 15:16.200]  I am very mean person.
[15:16.420 --> 15:22.010]  So, I followed up and I made another comment after the paper was retracted.
[15:22.980 --> 15:23.880]  It basically...
[15:23.880 --> 15:25.020]  So, they have...
[15:25.020 --> 15:27.440]  So, their rationale was here.
[15:27.440 --> 15:30.880]  So, they have a sociability index.
[15:31.100 --> 15:39.420]  Basically, they are calculating the ratio of how much time the animal spent with another animal versus how much time he spent with another...
[15:39.420 --> 15:40.320]  With no animal.
[15:40.560 --> 15:50.420]  So, their saying here is that if the value is greater than one, showcase a preference for social interaction.
[15:50.420 --> 15:55.160]  So, if you spend more time in the chamber with an animal, then you're social.
[15:55.400 --> 15:57.140]  You're not autistic, basically.
[15:57.580 --> 16:05.880]  So, what they're showing here, the animal with pre-exposure to COVID vaccine has a ratio of 1.5, which is a lot greater than one.
[16:06.680 --> 16:12.720]  So, even if we assume there's nothing wrong with the data, there's nothing wrong with methodology.
[16:12.720 --> 16:17.420]  The rats didn't enjoy huddling in a very tight shoebox.
[16:17.660 --> 16:26.880]  Their own data never, ever supported their claim that COVID vaccine, prenatal COVID vaccine caused autism-like behavior in rats.
[16:27.040 --> 16:28.000]  It's just...
[16:28.000 --> 16:29.820]  Even their own data don't support it.
[16:29.960 --> 16:35.000]  So, because of me, I decided to look at what other study that these...
[16:35.000 --> 16:40.400]  The same research group from Turkey published, you know...
[16:42.220 --> 16:46.800]  And this study looked at long-term exposure to maternal fruit coats.
[16:47.220 --> 16:53.490]  And I was looking at this data here, and then I decided to blow it up to see what...
[16:53.980 --> 16:56.980]  You know, and you can see I'm writing forensically.
[16:57.320 --> 16:59.780]  The top part, it seemed to be a cap.
[17:02.020 --> 17:03.640]  You know, it just...
[17:04.620 --> 17:06.420]  I don't know how that happened.
[17:07.820 --> 17:09.660]  And here's another study.
[17:09.820 --> 17:17.280]  They are saying that prenatal exposure to liquid with high salt concentration, including...
[17:18.260 --> 17:22.380]  Including this Turkish Uyghur, or the Middle Eastern Uyghur with salt.
[17:22.620 --> 17:23.760]  Reduce sociability.
[17:24.040 --> 17:28.520]  Remember that sociability is the thing that you don't like to spend time with another animal.
[17:29.100 --> 17:35.710]  So, this study is showing that prenatal exposure to salt water, and...
[17:36.340 --> 17:39.260]  And then there's Uyghur, and it can reduce sociability.
[17:39.640 --> 17:45.460]  And so, this is a picture I took, the Uyghur I bought the day after I flagged this paper.
[17:45.520 --> 17:48.180]  Because I just need to try what this thing tastes like.
[17:48.320 --> 17:48.840]  It was delicious.
[17:48.840 --> 17:49.940]  The salt...
[17:50.640 --> 17:53.020]  The salt made it so good.
[17:53.280 --> 17:55.280]  So, I'm already weird, so I don't really care.
[17:57.980 --> 18:03.480]  And so, I look at that paper showing that the yogurt is reducing...
[18:03.480 --> 18:05.820]  Yogurt with salt reduced sociability.
[18:06.460 --> 18:07.980]  And then look at the error bar.
[18:08.180 --> 18:09.420]  So, the error bar should...
[18:09.420 --> 18:13.920]  The size of the error bar shown here should correspond to the number shown here, right?
[18:13.980 --> 18:14.560]  You would think.
[18:15.540 --> 18:16.060]  Here.
[18:16.560 --> 18:21.820]  This error bar of 3.2 apparently is a lot smaller than the error bar of 0.3.
[18:22.040 --> 18:23.300]  I don't know what happened there.
[18:23.860 --> 18:28.380]  So, again, it's all published information available to everyone.
[18:28.620 --> 18:29.600]  I didn't hack into anything.
[18:29.760 --> 18:30.560]  It just...
[18:31.360 --> 18:32.540]  Doesn't make any sense.
[18:33.200 --> 18:33.780]  And next...
[18:34.100 --> 18:35.120]  I'm going to move fast.
[18:35.360 --> 18:38.300]  So, our color that promised to protect the brain.
[18:39.620 --> 18:40.140]  So...
[18:40.140 --> 18:40.620]  Here.
[18:41.000 --> 18:44.540]  Three years ago, there's a New York Times article saying that...
[18:44.540 --> 18:45.480]  So, this thing.
[18:45.600 --> 18:47.360]  This is basically a plastic color.
[18:48.780 --> 18:52.060]  They wear it, supposedly, to protect your brain.
[18:52.260 --> 18:57.880]  Because it's supposed to constrict your neck and cause jugular vein compression.
[18:58.360 --> 19:02.300]  Supposedly, the jugular vein compression is supposed to help your brain.
[19:02.840 --> 19:03.900]  Don't know how that works.
[19:05.260 --> 19:05.780]  And...
[19:05.780 --> 19:06.640]  So...
[19:06.640 --> 19:07.120]  But, yeah.
[19:07.120 --> 19:08.120]  And then they have...
[19:08.640 --> 19:09.160]  It's like...
[19:09.160 --> 19:10.620]  It gives a sense of protection.
[19:10.880 --> 19:12.880]  I think the sense of protection is a key word.
[19:13.000 --> 19:14.160]  But it is...
[19:14.160 --> 19:19.420]  It is a company in Connecticut, up there in Norwalk, I think.
[19:19.840 --> 19:20.360]  It's...
[19:20.360 --> 19:25.060]  It is still trying to push this into the military, believe it or not.
[19:25.280 --> 19:27.080]  So, it is actually a serious business.
[19:28.600 --> 19:35.860]  A year ago, an article came out after we provided more evidence against the data.
[19:36.900 --> 19:40.080]  And I want to show you some of the key evidence I provided.
[19:40.440 --> 19:42.280]  So, here are two studies.
[19:42.280 --> 19:45.120]  And then these two data tables are identical.
[19:45.580 --> 19:46.580]  Are absolutely identical.
[19:47.020 --> 19:47.520]  If you don't...
[19:47.520 --> 19:48.680]  Ignore the header.
[19:48.860 --> 19:49.860]  I'll just look at the number.
[19:50.000 --> 19:50.580]  They're identical.
[19:51.300 --> 19:51.560]  Right?
[19:51.900 --> 19:52.220]  Okay.
[19:52.560 --> 19:53.740]  So, we could say...
[19:53.990 --> 19:58.220]  What this one is in high school female soccer player.
[19:58.580 --> 20:03.460]  This one is in male SWAT team people.
[20:04.020 --> 20:04.440]  So...
[20:04.440 --> 20:07.740]  And then they change the end group size here.
[20:07.740 --> 20:12.040]  They change pre-season to pre-training.
[20:12.540 --> 20:14.540]  They change the italic.
[20:14.640 --> 20:16.080]  They change a lot of things.
[20:16.360 --> 20:19.760]  So, these are not accidental copy and paste.
[20:19.940 --> 20:20.580]  People just...
[20:20.580 --> 20:23.240]  Didn't just upload the wrong table.
[20:24.160 --> 20:27.960]  Changes have been made indicating somebody thought about what they're doing.
[20:29.400 --> 20:31.680]  And then the reason why I was able to...
[20:31.680 --> 20:32.820]  This is a manual matching.
[20:32.920 --> 20:33.980]  I didn't use any tools.
[20:34.280 --> 20:35.300]  I was just manual matching.
[20:35.400 --> 20:37.100]  The reason I was able to do that...
[20:37.100 --> 20:38.200]  I don't always do that.
[20:38.620 --> 20:42.120]  It's because I look at one table for a very long time.
[20:42.360 --> 20:45.980]  I stared it out for a very long time because the number are weird.
[20:46.240 --> 20:46.900]  So, you can see...
[20:46.900 --> 20:50.240]  So, these are different experimental group of people.
[20:50.920 --> 20:53.780]  And whatever the measure is, some kind of measurement.
[20:54.120 --> 20:57.840]  They have identical data pre- and post-training.
[20:58.060 --> 21:01.480]  And basically pre- and post-the athletic season, supposedly.
[21:02.100 --> 21:04.260]  And then two groups have identical data.
[21:04.420 --> 21:06.420]  Two other groups have other identical data.
[21:06.420 --> 21:07.540]  They're just very weird.
[21:07.700 --> 21:10.760]  But I cannot prove something is impossible.
[21:11.000 --> 21:12.800]  I can only say that seems weird.
[21:15.160 --> 21:21.000]  For us, when you say something is weird, that means that nobody would take actions about anything.
[21:21.160 --> 21:23.360]  They would just go absolutely nowhere.
[21:23.940 --> 21:27.280]  This thing is still under investigation, by the way, after so many times.
[21:27.280 --> 21:28.860]  And then they put in a correction.
[21:29.160 --> 21:31.180]  So, they decided to say, oh, yeah.
[21:31.400 --> 21:33.760]  Those tables are not supposed to be the same.
[21:33.920 --> 21:35.720]  So, they corrected one of the...
[21:35.720 --> 21:39.020]  One of the table, making it even more weird.
[21:41.660 --> 21:42.140]  Yeah.
[21:42.600 --> 21:43.380]  So, first...
[21:43.380 --> 21:46.200]  Like you said, the first group, we don't have any variabilities.
[21:46.540 --> 21:49.540]  We have eight people, but we don't have any variability.
[21:49.780 --> 21:51.280]  Everybody scored a .905.
[21:51.280 --> 21:55.140]  You don't know what point that score is, but everybody got the same.
[21:55.460 --> 21:59.700]  And then, here's another group, pre- and post-season.
[22:00.020 --> 22:04.380]  You have identical median, means, and standard deviation.
[22:04.380 --> 22:09.180]  And then, even the range are absolutely identical, pre- and post-season.
[22:09.460 --> 22:11.120]  For that one group.
[22:12.160 --> 22:19.320]  Again, these are not something that the journal will take as a hard evidence, which is like, oh, you cannot prove it.
[22:19.340 --> 22:19.940]  It's possible.
[22:20.140 --> 22:24.140]  I cannot, but common sense sometimes works.
[22:24.540 --> 22:26.900]  Just finally, these people are also...
[22:26.900 --> 22:30.720]  This study are also supported by the company that make this.
[22:30.880 --> 22:32.960]  This case is still ongoing.
[22:34.460 --> 22:37.180]  And then, next one is...
[22:37.500 --> 22:38.640]  There's a YouTuber.
[22:39.560 --> 22:43.760]  So, this guy have half a million YouTube followers.
[22:44.340 --> 22:49.440]  They put like weird, bizarre videos of them eating at the hotel and stuff.
[22:50.300 --> 22:51.920]  So, they have like...
[22:51.920 --> 22:57.640]  So, they are a couple in a university in Sweden.
[22:57.640 --> 22:58.960]  And so, this is...
[22:58.960 --> 23:00.440]  This is Shankar Hama.
[23:00.700 --> 23:01.220]  He's...
[23:01.220 --> 23:01.660]  Shama.
[23:01.840 --> 23:04.660]  And then, he's a researcher in...
[23:05.280 --> 23:06.980]  Buspala University in Sweden.
[23:07.340 --> 23:12.180]  So, I got their entire book retracted with the help of Image to an AI.
[23:13.660 --> 23:14.460]  It's just...
[23:14.460 --> 23:15.680]  It is very basic stuff.
[23:15.880 --> 23:16.900]  It's just like very like...
[23:16.900 --> 23:19.440]  The reuse of the same image in...
[23:19.440 --> 23:21.480]  Here's like four book chapters.
[23:21.480 --> 23:25.100]  He reused the same thing over and over again.
[23:25.260 --> 23:26.800]  You can see the labels are all different.
[23:28.160 --> 23:33.200]  There's just very, very basic like image reuse.
[23:33.200 --> 23:35.620]  You can see that it's all the same here.
[23:36.300 --> 23:40.720]  So, you know, the labels don't mean anything anymore at this point.
[23:41.740 --> 23:46.500]  So, there's really nothing too interesting to talk about.
[23:46.940 --> 23:47.560]  But...
[23:47.560 --> 24:03.860]  So, long-storm short, the Swedish National Board of Assessment very quickly issued a statement after a year of investigation saying that these couple and some other people are guilty of research misconduct.
[24:03.860 --> 24:07.740]  And this speed of action is like...
[24:07.740 --> 24:10.460]  Doesn't happen in the U.S. at all.
[24:10.660 --> 24:11.960]  Like we are...
[24:11.960 --> 24:14.400]  We're moving things at a glacial speed.
[24:14.720 --> 24:16.920]  So, a lot of these cases...
[24:16.920 --> 24:20.100]  Like I said, I was lucky to...
[24:20.100 --> 24:23.180]  Be able to use the Image to an AI.
[24:23.540 --> 24:24.880]  So, I want to talk about the...
[24:24.880 --> 24:27.700]  Eyes over AI cases.
[24:28.040 --> 24:32.620]  So, after a lot of those cases, I had some kind of identity crisis.
[24:32.620 --> 24:33.340]  Because...
[24:34.340 --> 24:37.640]  Image to an AI is extremely expensive for people...
[24:37.640 --> 24:40.460]  Not for people for institute or for government.
[24:41.560 --> 24:42.080]  For...
[24:42.080 --> 24:42.720]  If...
[24:42.720 --> 24:43.660]  You know...
[24:43.660 --> 24:45.980]  If they don't know you as a sleuth...
[24:45.980 --> 24:46.220]  That's...
[24:46.220 --> 24:47.320]  That's how much...
[24:47.320 --> 24:47.860]  That's...
[24:47.860 --> 24:48.920]  That's a market price.
[24:49.100 --> 24:50.120]  So, I was wondering...
[24:50.120 --> 24:53.480]  Am I just someone with a free copy of Image to an AI?
[24:53.480 --> 24:55.300]  Or am I actually a sleuth?
[24:55.520 --> 24:55.680]  You know?
[24:56.580 --> 24:59.540]  To what extent am I depending on technologies?
[25:01.120 --> 25:03.220]  So, I started doing a lot of things.
[25:03.400 --> 25:05.480]  I started looking into things with just eyes.
[25:05.700 --> 25:07.080]  And so, I want to show some of the...
[25:07.080 --> 25:08.880]  I want to show you easy cases.
[25:09.160 --> 25:10.280]  Medium level cases.
[25:10.600 --> 25:11.720]  And advanced cases.
[25:12.040 --> 25:13.980]  So, the easiest ones...
[25:13.980 --> 25:16.540]  So, these are the chemistry squiggles.
[25:17.680 --> 25:18.780]  I call them all squiggles.
[25:19.980 --> 25:21.980]  So, you can see these are...
[25:21.980 --> 25:24.180]  These have the identical noises.
[25:24.980 --> 25:32.220]  So, people who have basic chemistry probably know that the noises are random events.
[25:32.460 --> 25:33.240]  They don't duplicate.
[25:33.500 --> 25:34.260]  They're snowflakes.
[25:34.360 --> 25:36.440]  Basically, noises are snowflakes.
[25:36.540 --> 25:40.580]  They don't duplicate in different traces, different experiments.
[25:40.580 --> 25:49.380]  Even if you run the same sample multiple times, they might have the same major peaks, but they wouldn't have the same noises.
[25:50.380 --> 25:54.600]  Again, these are five different samples.
[25:55.200 --> 25:56.660]  They all have the same noises.
[25:57.060 --> 25:58.700]  The only thing different is the color.
[26:01.100 --> 26:02.740]  This is a little bit trickier.
[26:02.920 --> 26:06.880]  I'll give you a second to see that these are actually the same.
[26:07.700 --> 26:14.320]  So, this is a little bit deceiving because you can see they change the intensity and change the vertical stretch.
[26:18.290 --> 26:19.970]  This is kind of just silly.
[26:23.530 --> 26:26.690]  This one, you can see the black trace.
[26:27.030 --> 26:31.790]  The noise in the black trace during each of the light-on phase.
[26:32.470 --> 26:33.910]  The noises are identical.
[26:34.270 --> 26:35.670]  So, these are pretty easy to see.
[26:36.590 --> 26:38.170]  I mean, they're published data.
[26:38.410 --> 26:42.050]  So, I mean, obviously, during publication, nobody saw that.
[26:42.630 --> 26:44.090]  This is just kind of goofy.
[26:44.090 --> 26:47.090]  You can see that these are just... these are not...
[26:47.770 --> 26:50.570]  I mean, you guys know that this...
[26:51.090 --> 26:52.630]  I don't even know how to describe it.
[26:54.290 --> 26:56.390]  And then, also, like...
[27:00.850 --> 27:03.430]  It's in a journal called Feel.
[27:03.750 --> 27:07.190]  So, I was thinking if you study feel, I was thinking you know more than...
[27:08.530 --> 27:09.490]  Yeah, so...
[27:15.380 --> 27:16.340]  Yeah, so...
[27:16.340 --> 27:16.820]  See, I can...
[27:17.680 --> 27:19.880]  Next week, I'm going to give a talk in Chinese.
[27:20.100 --> 27:22.120]  My scientific Chinese is pretty bad.
[27:22.260 --> 27:23.060]  I decided...
[27:23.060 --> 27:24.080]  It's doable.
[27:24.320 --> 27:26.980]  I can just point and people will laugh and that's all I have to do.
[27:29.400 --> 27:30.560]  It's very easy.
[27:31.700 --> 27:34.320]  And here is, again, like, backtracking.
[27:35.080 --> 27:39.940]  So, now we're moving on to slightly advanced cases.
[27:40.180 --> 27:41.960]  And this is a little trickier.
[27:42.120 --> 27:43.800]  You can see these are also the same.
[27:44.160 --> 27:47.040]  But they mix them together.
[27:49.040 --> 27:50.200]  This is annoying.
[27:50.360 --> 27:51.180]  This is very annoying.
[27:51.340 --> 27:54.040]  Because I just sit down and then drag them.
[27:54.580 --> 27:55.820]  AIs don't do that.
[27:55.820 --> 27:58.040]  Like, no AI can do this part.
[27:58.280 --> 28:01.860]  So, again, the same color boxes indicate the same section...
[28:01.860 --> 28:04.440]  The section was the same thing.
[28:04.880 --> 28:06.760]  This is very annoying to mark out.
[28:08.400 --> 28:12.340]  Again, you can see that it's like a patchwork.
[28:12.980 --> 28:14.240]  But if you look at...
[28:14.240 --> 28:16.940]  If you read the paper, the people wouldn't pay attention to that.
[28:17.020 --> 28:17.480]  Why would you?
[28:17.560 --> 28:18.980]  Why would you when you read this thing?
[28:21.500 --> 28:24.000]  It's not too hard, I guess.
[28:25.660 --> 28:27.320]  This is a manual matching.
[28:27.420 --> 28:28.440]  This is a little bit tricky.
[28:28.660 --> 28:31.040]  But because this thing is so weird looking.
[28:31.820 --> 28:34.400]  So, you can see there are some changes.
[28:34.800 --> 28:40.200]  But if you look at the details, these are just basically the same...
[28:40.200 --> 28:44.800]  Same basic structure that they either added something or took away something.
[28:47.540 --> 28:48.060]  The initial...
[28:48.060 --> 28:48.680]  So, the ?
[28:48.680 --> 28:50.080]  Is this a little trickier?
[29:03.070 --> 29:06.670]  And now, we are going to have the little advanced cases.
[29:07.310 --> 29:09.010]  So, these are...
[29:11.350 --> 29:13.550]  This is the same as this one.
[29:14.150 --> 29:15.830]  Now that's a little bit tricky.
[29:16.390 --> 29:19.330]  You had to kind of match a little prickly here.
[29:22.070 --> 29:28.450]  You kind of had to find some landmark to guide your eyes.
[29:29.070 --> 29:30.950]  But that's a little tricky.
[29:31.870 --> 29:35.690]  And this one I use the forensic laser.
[29:36.030 --> 29:40.470]  I took this chunk, I blow it up, and you can see these are two parts.
[29:40.610 --> 29:41.570]  There's a fracture.
[29:44.270 --> 29:47.130]  So this is a forensic laser I'm using.
[29:48.990 --> 29:52.150]  And then this is also kind of an advanced case.
[29:52.470 --> 29:58.470]  This is two... but it's actually not that hard once you know the style of the people.
[29:58.810 --> 30:03.850]  Like you kind of know where they want to play this game so you are able to find their papers.
[30:04.950 --> 30:08.490]  Because it's clearly used by the same people and you can see that.
[30:09.090 --> 30:11.310]  It's just very annoying to have to mark it out.
[30:13.210 --> 30:26.410]  And then this is also like... it's kind of hard to see because if you just look at this graph in the paper, why would you look at these little parts and blow it up and decide there's a fracture here?
[30:27.310 --> 30:31.650]  But, you know, it's... I don't know what it is, but it's not physics.
[30:34.550 --> 30:35.710]  Same thing here.
[30:35.910 --> 30:40.030]  There's a hairline gap.
[30:40.810 --> 30:42.890]  But again, like it's from here.
[30:43.090 --> 30:44.330]  Like why would you look at that?
[30:44.330 --> 30:52.150]  Most people when they read scientific papers decided there is a trust building when you read scientific papers.
[30:52.350 --> 30:54.330]  You don't really necessarily look at that.
[30:54.530 --> 30:55.550]  So same thing here.
[30:57.270 --> 30:58.410]  More of this.
[30:58.650 --> 31:00.830]  This is really annoying.
[31:01.670 --> 31:04.250]  These are only all the same.
[31:04.490 --> 31:05.390]  There's all the...
[31:07.830 --> 31:12.110]  And like... we have a squiggle shortage.
[31:15.530 --> 31:17.630]  Just keep resusing it.
[31:22.000 --> 31:25.320]  It's just... you can just use random number generator.
[31:31.090 --> 31:35.590]  Again, this is... you can see... if you blow it up, you can see the fractures.
[31:37.430 --> 31:48.030]  And this one, yeah, it's really not that hard, but you can see these... there's a... there's a gap.
[31:48.330 --> 31:50.170]  Because this is energy count.
[31:50.490 --> 31:55.570]  So the readout is when they emit energy, this is count.
[31:55.570 --> 31:57.890]  So... so line shouldn't go horizontally.
[31:57.890 --> 31:59.150]  You only go vertically.
[31:59.730 --> 32:02.510]  You cannot have horizontal line go that way.
[32:03.930 --> 32:14.970]  This is a little bit... this is not difficult for people who took chemistry and physics or paid attention to them during high school or college, but that wasn't me.
[32:16.830 --> 32:20.630]  But I... someone gave me a chart that I learned to use.
[32:20.630 --> 32:24.910]  And you can look at the element and you can look at the element location of different...
[32:24.910 --> 32:26.170]  that they have peaks.
[32:26.350 --> 32:28.510]  So I match them and I point it out.
[32:29.530 --> 32:33.230]  And... it's kind of a... it's a shame and the editor-in-chief.
[32:33.530 --> 32:33.870]  It's a look.
[32:33.970 --> 32:36.990]  I didn't even take high... I didn't even take college chemistry.
[32:36.990 --> 32:38.290]  But here you go.
[32:38.450 --> 32:39.150]  This is evidence.
[32:40.770 --> 32:48.070]  And... this is also a little bit advanced because this is basically the same data that flipped it...
[32:48.970 --> 32:52.450]  into... different directions and changed the...
[32:52.690 --> 32:56.050]  the scary part of these two studies that...
[32:56.050 --> 33:00.010]  the two... the two papers don't have overlapping authors.
[33:00.250 --> 33:01.710]  They're not by the same people.
[33:01.710 --> 33:05.650]  So this is when we run into the problem of paper mill.
[33:05.870 --> 33:10.030]  It's like... where this came from.
[33:10.530 --> 33:13.090]  So that's a scary part of...
[33:13.090 --> 33:16.790]  And... and then... so I posted this on... on LinkedIn.
[33:17.310 --> 33:19.150]  Some guys are fighting me and I...
[33:20.930 --> 33:23.490]  and I... and I...
[33:23.490 --> 33:24.890]  and I... I... I... I... I... I...
[33:24.890 --> 33:28.910]  presented this data to show that I just took this line from a...
[33:28.910 --> 33:33.350]  obscure graph and line them up together to...
[33:33.350 --> 33:34.290]  to show that...
[33:34.290 --> 33:37.150]  do you really think that this is still possible?
[33:37.430 --> 33:40.750]  That these were done by two independent group of...
[33:40.750 --> 33:42.690]  these are two independent experiment?
[33:45.590 --> 33:46.430]  So... okay.
[33:46.630 --> 33:48.570]  Now we're moving to why are...
[33:48.570 --> 33:50.390]  there are so many bad papers.
[33:51.710 --> 33:52.650]  Peer review.
[33:52.850 --> 33:56.050]  Because I always think a peer review is some sort of like...
[33:56.050 --> 33:57.130]  this...
[33:58.850 --> 34:02.030]  there's a credibility that is given when you...
[34:02.030 --> 34:04.090]  when you say something is peer reviewed.
[34:05.210 --> 34:07.810]  So again, this is a paper by my friend...
[34:07.810 --> 34:09.390]  Reese Richardson to show.
[34:09.570 --> 34:10.450]  And so...
[34:10.450 --> 34:12.150]  the problem is profit margin.
[34:12.470 --> 34:14.050]  So a lot of people here are...
[34:14.730 --> 34:16.490]  kind of anti-capitalistic.
[34:16.870 --> 34:19.930]  So I don't want to blame capitalism for everything.
[34:19.930 --> 34:21.110]  but...
[34:21.110 --> 34:21.390]  the...
[34:21.390 --> 34:23.470]  the profit margin is insane.
[34:23.870 --> 34:24.210]  This is...
[34:24.210 --> 34:27.950]  this is tobacco industry level of profit margin right here.
[34:28.750 --> 34:30.250]  It's absolutely insane.
[34:33.130 --> 34:34.930]  And you can also find...
[34:34.930 --> 34:35.310]  easily...
[34:35.310 --> 34:36.290]  this is from Facebook.
[34:36.590 --> 34:39.210]  You can find authorship for sale...
[34:39.620 --> 34:40.670]  in a lot of places.
[34:40.990 --> 34:42.290]  They're trading authorship.
[34:42.290 --> 34:42.650]  So...
[34:43.290 --> 34:43.610]  so...
[34:43.610 --> 34:44.530]  imagine if you're...
[34:44.530 --> 34:45.650]  if you're...
[34:47.410 --> 34:47.970]  rich...
[34:47.970 --> 34:49.210]  guy from Saudi Arabia.
[34:49.430 --> 34:52.010]  You can just basically buy yourself a lot of papers.
[34:52.690 --> 34:53.070]  And then...
[34:53.070 --> 34:54.650]  you can claim that you're...
[34:54.650 --> 34:55.370]  you're...
[34:55.370 --> 34:56.510]  well-published...
[34:56.510 --> 34:57.170]  scientist.
[34:58.050 --> 34:59.690]  All you did is spending money.
[34:59.930 --> 35:01.290]  But you could find...
[35:01.290 --> 35:02.430]  this is all over the place.
[35:02.570 --> 35:04.490]  People even use names and stuff.
[35:05.610 --> 35:06.170]  So...
[35:06.170 --> 35:07.570]  where do we flag paper?
[35:07.850 --> 35:09.990]  Our platform is mostly...
[35:09.990 --> 35:11.030]  this thing called...
[35:11.030 --> 35:11.530]  pop peer.
[35:11.730 --> 35:13.470]  This is where we can...
[35:13.470 --> 35:14.230]  anonymously...
[35:14.230 --> 35:14.950]  flag papers.
[35:15.730 --> 35:16.330]  So...
[35:16.330 --> 35:16.970]  for example...
[35:16.970 --> 35:18.050]  this is one of the paper...
[35:18.050 --> 35:18.530]  I flag.
[35:19.190 --> 35:21.030]  These are two papers...
[35:21.030 --> 35:21.230]  with...
[35:21.230 --> 35:23.370]  non-overlapping...
[35:24.030 --> 35:24.630]  authors...
[35:24.630 --> 35:25.770]  but overlapping images.
[35:26.710 --> 35:27.270]  So...
[35:27.270 --> 35:27.650]  this...
[35:27.650 --> 35:28.730]  is a very basic...
[35:28.730 --> 35:29.590]  website...
[35:29.590 --> 35:30.670]  where you flag papers.
[35:31.450 --> 35:31.790]  And...
[35:31.790 --> 35:32.170]  but...
[35:32.170 --> 35:32.350]  like...
[35:32.350 --> 35:32.850]  you guys...
[35:32.850 --> 35:33.910]  when you do something...
[35:33.910 --> 35:34.250]  as...
[35:35.110 --> 35:35.710]  activists...
[35:35.710 --> 35:36.910]  you got haters, right?
[35:36.910 --> 35:37.750]  that's just...
[35:37.750 --> 35:38.630]  that's a...
[35:38.630 --> 35:39.210]  that's a beauty in formula.
[35:39.310 --> 35:40.150]  We got our haters.
[35:40.730 --> 35:41.290]  And...
[35:41.290 --> 35:43.710]  here is where you guys might be interested in.
[35:43.890 --> 35:44.390]  So...
[35:44.390 --> 35:46.510]  this is this guy called Science Guardians.
[35:47.110 --> 35:49.050]  They have a very sleek website.
[35:49.290 --> 35:50.650]  You can see everything looks beautiful.
[35:51.470 --> 35:52.030]  And...
[35:52.030 --> 35:53.210]  they call themselves...
[35:53.210 --> 35:57.190]  the global hub for upholding the highest standard of scientific integrity.
[35:59.090 --> 35:59.650]  And...
[35:59.650 --> 35:59.930]  and...
[35:59.930 --> 36:00.870]  they basically...
[36:00.870 --> 36:04.090]  saying that retraction watch is also one of our...
[36:04.090 --> 36:04.690]  thing...
[36:04.690 --> 36:06.450]  and pop peer is one of our platform.
[36:06.450 --> 36:06.950]  So...
[36:06.950 --> 36:09.350]  they basically are saying that these are all...
[36:09.350 --> 36:09.690]  they...
[36:09.690 --> 36:10.490]  we are...
[36:10.490 --> 36:11.110]  the...
[36:11.110 --> 36:12.630]  the pop peer mob.
[36:15.610 --> 36:16.350]  So...
[36:16.350 --> 36:17.050]  so...
[36:17.050 --> 36:18.050]  so...
[36:18.050 --> 36:19.510]  Reese Richardson, he's the author...
[36:19.510 --> 36:20.190]  he's...
[36:20.190 --> 36:23.270]  the author of the PNS paper that I just showed you.
[36:23.470 --> 36:27.350]  And then this is a tweet by Science Guardian calling...
[36:27.350 --> 36:28.410]  uh...
[36:28.410 --> 36:29.530]  uh...
[36:29.530 --> 36:30.270]  Reese...
[36:30.270 --> 36:31.730]  the pop...
[36:31.730 --> 36:32.110]  like...
[36:32.110 --> 36:33.250]  mob member.
[36:33.250 --> 36:34.470]  her...
[36:34.470 --> 36:34.830]  so...
[36:34.830 --> 36:36.250]  this is basically trashing...
[36:37.470 --> 36:38.190]  us...
[36:38.190 --> 36:38.510]  um...
[36:38.510 --> 36:39.090]  this is...
[36:39.090 --> 36:39.970]  Elizabeth Beek...
[36:39.970 --> 36:40.990]  Elizabeth Beek...
[36:40.990 --> 36:41.870]  uh...
[36:41.870 --> 36:43.190]  I follow her footsteps.
[36:43.490 --> 36:44.030]  She's OG.
[36:44.230 --> 36:45.250]  She's the most...
[36:45.870 --> 36:46.270]  uh...
[36:46.270 --> 36:47.070]  prolific...
[36:47.070 --> 36:48.590]  well-respected...
[36:48.590 --> 36:49.250]  uh...
[36:49.250 --> 36:50.190]  fearless...
[36:50.190 --> 36:50.790]  I can...
[36:50.790 --> 36:51.110]  I can...
[36:51.110 --> 36:52.370]  I can go on forever...
[36:52.370 --> 36:52.970]  the...
[36:52.970 --> 36:53.690]  uh...
[36:53.690 --> 36:54.450]  of us...
[36:54.450 --> 36:54.710]  like...
[36:54.710 --> 36:55.610]  she started everything.
[36:55.810 --> 36:57.410]  She pioneered this whole...
[36:57.410 --> 36:58.090]  sluicing...
[36:58.090 --> 36:58.670]  uh...
[36:58.670 --> 36:59.690]  grassroot movement.
[37:00.190 --> 37:01.570]  This is also like...
[37:01.570 --> 37:02.870]  grassroot underground movement.
[37:03.290 --> 37:04.530]  And then they call her...
[37:04.530 --> 37:04.870]  perpetrator...
[37:05.590 --> 37:06.090]  five...
[37:06.090 --> 37:06.310]  EB.
[37:07.690 --> 37:08.150]  uh...
[37:08.150 --> 37:08.470]  so...
[37:08.470 --> 37:09.830]  so here is like...
[37:09.830 --> 37:10.450]  uh...
[37:10.450 --> 37:10.930]  I need...
[37:10.930 --> 37:11.110]  I...
[37:11.110 --> 37:11.430]  I...
[37:11.430 --> 37:14.110]  I'm not coercing anybody to do anything.
[37:14.390 --> 37:14.890]  But...
[37:14.890 --> 37:15.370]  uh...
[37:15.370 --> 37:16.030]  we...
[37:16.030 --> 37:17.130]  we sluice...
[37:17.130 --> 37:17.890]  we're not hackers.
[37:18.070 --> 37:19.390]  We don't know who they are.
[37:19.390 --> 37:21.670]  We don't know what's behind Science Guardian.
[37:22.210 --> 37:22.570]  So...
[37:23.710 --> 37:24.430]  um...
[37:25.190 --> 37:26.910]  you're welcome to contribute.
[37:27.690 --> 37:28.110]  So...
[37:28.110 --> 37:29.290]  yeah.
[37:29.890 --> 37:30.370]  Um...
[37:30.370 --> 37:31.150]  so...
[37:31.150 --> 37:31.690]  wow...
[37:31.690 --> 37:33.310]  I blew it really fast.
[37:33.770 --> 37:34.150]  Um...
[37:34.150 --> 37:35.810]  so you can find me...
[37:35.810 --> 37:36.750]  you can email me.
[37:36.930 --> 37:38.530]  I'm pretty active on LinkedIn.
[37:39.310 --> 37:39.670]  Um...
[37:39.930 --> 37:43.790]  and I have a good amount of time for Q&A.
[37:55.200 --> 37:55.860]  Um...
[37:55.860 --> 37:58.420]  for questions, please line up at the microphones.
[37:59.060 --> 37:59.540]  Um...
[37:59.540 --> 38:01.680]  and I'll be starting out with a question from the live stream.
[38:01.880 --> 38:03.060]  Which is a two-part.
[38:03.280 --> 38:05.320]  Which is first, who funds this sluicing work?
[38:05.640 --> 38:07.740]  And how should it be funded in your opinion?
[38:09.180 --> 38:09.660]  Uh...
[38:09.660 --> 38:11.440]  this is currently a grass...
[38:12.280 --> 38:12.760]  uh...
[38:14.000 --> 38:14.480]  activity.
[38:15.040 --> 38:15.820]  Like we're not...
[38:15.820 --> 38:16.460]  we're all volunteer.
[38:16.700 --> 38:17.080]  Like you...
[38:17.080 --> 38:18.460]  like a lot of you guys are volunteers.
[38:19.000 --> 38:20.300]  And we don't really get paid.
[38:20.880 --> 38:21.120]  Um...
[38:21.120 --> 38:22.260]  because getting paid...
[38:22.260 --> 38:24.260]  makes things a little bit more complicated.
[38:24.660 --> 38:25.080]  So...
[38:25.080 --> 38:25.960]  currently we're not.
[38:25.960 --> 38:26.360]  like...
[38:26.360 --> 38:27.000]  I mean...
[38:27.000 --> 38:27.920]  I can't go on...
[38:27.920 --> 38:28.500]  on that forever.
[38:28.780 --> 38:29.220]  But that...
[38:29.220 --> 38:31.260]  but that's currently we're not getting paid.
[38:31.620 --> 38:32.040]  Yeah.
[38:34.220 --> 38:34.780]  Um...
[38:34.780 --> 38:34.980]  Hi.
[38:36.460 --> 38:37.020]  Um...
[38:37.020 --> 38:37.120]  Oh...
[38:37.120 --> 38:38.380]  I was wondering what the like...
[38:38.380 --> 38:39.120]  retraction...
[38:39.120 --> 38:41.420]  can you talk more about the retraction process...
[38:41.420 --> 38:41.880]  ...from like...
[38:41.880 --> 38:42.880]  how you...
[38:42.880 --> 38:43.660]  I guess...
[38:43.660 --> 38:45.260]  accuse or whatever the right word is...
[38:45.260 --> 38:45.780]  like...
[38:45.780 --> 38:47.080]  who is notified upstream...
[38:47.080 --> 38:47.440]  and then the...
[38:47.440 --> 38:49.900]  like the citations later as well...
[38:49.900 --> 38:50.040]  like...
[38:50.040 --> 38:50.780]  So...
[38:50.780 --> 38:51.120]  yeah...
[38:51.120 --> 38:52.080]  that's a very good question.
[38:52.280 --> 38:53.460]  So what we do is that...
[38:53.460 --> 38:53.980]  we first...
[38:53.980 --> 38:55.000]  we flag the paper...
[38:55.000 --> 38:57.240]  so that people can see that...
[38:57.240 --> 38:58.920]  and then we'll send a report...
[38:58.920 --> 39:00.260]  to the editor-in-chief...
[39:00.260 --> 39:00.740]  and then...
[39:00.740 --> 39:02.800]  every publisher has a...
[39:02.800 --> 39:04.020]  integrity team...
[39:04.020 --> 39:05.260]  that receive complaints...
[39:05.260 --> 39:06.940]  or receive this kind of report...
[39:06.940 --> 39:08.960]  and they're the people that handled...
[39:13.160 --> 39:13.680]  um...
[39:13.680 --> 39:14.720]  and then ultimately...
[39:14.720 --> 39:16.080]  if they decided to retract...
[39:16.080 --> 39:17.620]  the journal will retract...
[39:17.620 --> 39:18.940]  and sometimes it's a long fight...
[39:18.940 --> 39:19.900]  it can go on for two...
[39:20.420 --> 39:20.800]  ten years...
[39:21.500 --> 39:21.880]  but...
[39:21.880 --> 39:22.080]  yeah...
[39:22.080 --> 39:23.080]  it's getting faster now.
[39:23.200 --> 39:23.920]  What about like...
[39:23.920 --> 39:24.840]  like upstream...
[39:24.840 --> 39:26.040]  because there's papers that might have...
[39:26.040 --> 39:26.880]  cited that paper...
[39:26.880 --> 39:27.380]  are they notified?
[39:27.640 --> 39:28.360]  That is...
[39:28.360 --> 39:29.640]  that is just...
[39:30.960 --> 39:31.440]  yeah...
[39:31.440 --> 39:32.000]  no...
[39:32.000 --> 39:32.320]  yeah...
[39:32.320 --> 39:32.600]  it's...
[39:32.600 --> 39:33.220]  no...
[39:33.220 --> 39:34.920]  there's no mechanism...
[39:34.920 --> 39:35.980]  that is effectively...
[39:35.980 --> 39:37.480]  I know what you're talking about...
[39:37.480 --> 39:37.760]  but...
[39:37.760 --> 39:38.140]  yeah...
[39:38.140 --> 39:38.560]  it's a...
[39:38.560 --> 39:39.840]  it's a chain of reactions.
[39:39.840 --> 39:40.440]  Okay.
[39:40.940 --> 39:41.260]  Alright.
[39:42.080 --> 39:42.700]  Um...
[39:42.700 --> 39:43.800]  I'm just wondering...
[39:43.800 --> 39:44.890]  like don't the big publishers...
[39:45.380 --> 39:46.260]  of these papers...
[39:46.260 --> 39:47.760]  like vet the papers...
[39:47.760 --> 39:48.680]  before they're published?
[39:48.880 --> 39:49.460]  They...
[39:49.460 --> 39:50.460]  they...
[39:50.460 --> 39:51.280]  should...
[39:51.280 --> 39:52.380]  that's a problem...
[39:52.380 --> 39:53.960]  because it is peer review...
[39:53.960 --> 39:55.300]  peers are...
[39:55.300 --> 39:55.860]  like...
[39:55.860 --> 39:56.940]  scientists...
[39:56.940 --> 39:57.040]  right?
[39:57.220 --> 39:58.840]  Peers don't get paid...
[39:59.360 --> 39:59.800]  so...
[39:59.800 --> 40:01.110]  sleuths are not getting paid...
[40:01.980 --> 40:02.600]  peer...
[40:02.600 --> 40:03.160]  reviewers are not...
[40:03.160 --> 40:04.110]  also not getting paid...
[40:04.440 --> 40:05.500]  and people aren't busy...
[40:05.500 --> 40:06.320]  why would I...
[40:06.320 --> 40:06.720]  you know...
[40:06.720 --> 40:08.400]  the publishers are the one...
[40:08.400 --> 40:09.130]  that get making money...
[40:09.130 --> 40:11.230]  and then they don't pay...
[40:11.230 --> 40:12.290]  the peers...
[40:12.290 --> 40:12.590]  so...
[40:12.590 --> 40:13.050]  why would...
[40:13.050 --> 40:14.550]  peer reviewer...
[40:14.550 --> 40:15.770]  spend a lot of time...
[40:15.770 --> 40:16.250]  working on things...
[40:17.030 --> 40:17.390]  so...
[40:17.390 --> 40:18.850]  independent of what...
[40:18.850 --> 40:20.310]  peer reviewers do...
[40:20.310 --> 40:21.230]  the publishers...
[40:21.230 --> 40:21.910]  unfortunately...
[40:21.910 --> 40:23.050]  don't do a lot...
[40:23.050 --> 40:23.910]  in terms of...
[40:23.910 --> 40:24.170]  I mean...
[40:24.170 --> 40:25.270]  they check typos...
[40:25.270 --> 40:25.690]  and stuff...
[40:25.690 --> 40:26.890]  but that's pretty much about it...
[40:26.890 --> 40:27.950]  they don't really check...
[40:27.950 --> 40:28.710]  images...
[40:28.710 --> 40:29.910]  and citations...
[40:29.910 --> 40:31.430]  they're catching up...
[40:31.430 --> 40:32.410]  but the rate of...
[40:32.410 --> 40:33.210]  they're catching up...
[40:33.210 --> 40:34.570]  is so much slower...
[40:34.570 --> 40:35.430]  than the fraud...
[40:36.670 --> 40:37.230]  so...
[40:45.610 --> 40:47.010]  and like...
[40:47.010 --> 40:47.310]  research...
[40:47.310 --> 40:48.070]  and like...
[40:48.070 --> 40:48.490]  I guess...
[40:48.490 --> 40:49.070]  what goes on...
[40:49.070 --> 40:50.130]  with publication process...
[40:50.130 --> 40:50.750]  and stuff like that?
[40:51.510 --> 40:52.470]  you can...
[40:52.470 --> 40:53.730]  once you find me...
[40:53.730 --> 40:54.310]  also shows...
[40:54.310 --> 40:55.010]  and you can find...
[40:55.010 --> 40:56.010]  the chain of us...
[40:56.010 --> 40:56.930]  sort of like...
[40:56.930 --> 40:57.450]  you guys...
[40:57.450 --> 40:57.790]  it's like...
[40:58.430 --> 41:00.010]  it's all right here...
[41:00.010 --> 41:01.690]  once you find that...
[41:01.690 --> 41:02.470]  so...
[41:02.470 --> 41:03.710]  you can contact me...
[41:03.710 --> 41:04.590]  and I'll...
[41:04.590 --> 41:05.410]  guide your way...
[41:05.410 --> 41:05.890]  through there...
[41:05.890 --> 41:06.230]  thank you...
[41:06.910 --> 41:07.550]  yeah...
[41:07.550 --> 41:09.250]  you said that...
[41:09.250 --> 41:10.310]  a lot of these...
[41:11.070 --> 41:11.710]  images...
[41:11.710 --> 41:12.890]  are not really...
[41:12.890 --> 41:13.270]  viewed...
[41:13.270 --> 41:13.970]  by...
[41:13.970 --> 41:15.450]  people in the industry...
[41:15.450 --> 41:16.290]  not really...
[41:16.290 --> 41:17.010]  paid attention to...
[41:17.010 --> 41:18.290]  what's the point then...
[41:18.290 --> 41:18.650]  what's the...
[41:18.650 --> 41:19.610]  who do you think...
[41:19.610 --> 41:20.390]  they're targeting...
[41:20.390 --> 41:21.370]  these images at?
[41:23.270 --> 41:23.910]  so...
[41:23.910 --> 41:24.850]  it is very...
[41:24.850 --> 41:25.390]  unfortunate...
[41:25.390 --> 41:26.250]  because these...
[41:26.250 --> 41:27.190]  AI tools...
[41:27.190 --> 41:27.530]  are highly...
[41:27.530 --> 41:28.050]  it is...
[41:28.050 --> 41:28.810]  widely available...
[41:28.810 --> 41:29.510]  to everybody...
[41:29.510 --> 41:30.370]  the publishers...
[41:30.370 --> 41:31.430]  definitely...
[41:31.430 --> 41:31.790]  should...
[41:31.790 --> 41:32.210]  I mean...
[41:32.210 --> 41:32.930]  they...
[41:32.930 --> 41:33.590]  can afford...
[41:33.590 --> 41:34.390]  a 10 copy...
[41:34.390 --> 41:34.710]  of them...
[41:34.710 --> 41:35.330]  I don't know why...
[41:35.330 --> 41:36.010]  they're not...
[41:36.010 --> 41:36.610]  using them...
[41:36.610 --> 41:37.590]  as much as they...
[41:37.590 --> 41:38.730]  they should...
[41:38.730 --> 41:39.770]  so...
[41:39.770 --> 41:40.170]  it's...
[41:40.170 --> 41:41.150]  it's a very...
[41:41.150 --> 41:41.810]  bizarre...
[41:41.810 --> 41:42.650]  situation...
[41:42.650 --> 41:43.110]  like...
[41:43.110 --> 41:46.290]  but you...
[41:46.290 --> 41:47.270]  the publisher...
[41:47.270 --> 41:47.790]  why...
[41:47.790 --> 41:48.890]  why wouldn't they...
[41:48.890 --> 41:49.350]  do the...
[41:49.350 --> 41:50.070]  пере-screening...
[41:50.070 --> 41:50.550]  before...
[41:50.550 --> 41:52.270]  before publishing...
[41:52.270 --> 41:53.870]  I'm sorry...
[41:53.870 --> 41:54.990]  I misspoke a little bit...
[41:54.990 --> 41:56.070]  the people who...
[42:00.370 --> 42:08.410]  My understanding was that the people in the industry, as they're reading these publications, they don't really necessarily pay attention to all of these squigglies.
[42:09.130 --> 42:11.630]  And yet they're putting in these fake squigglies.
[42:11.710 --> 42:14.150]  I ask this question all the time.
[42:14.530 --> 42:19.370]  I guess sometimes it's just sort of like, oh, we did this experiment, here's the thing.
[42:19.570 --> 42:38.170]  Even if you just like... I guess there's a lot of like, you show that you are able to do the experiment, that you have the data for the experiment, even if the experiment was just like, let's just say, I have a radio or something, like, nobody will open your radio and check every part.
[42:38.310 --> 42:40.090]  You would just say, I have a radio, right?
[42:40.550 --> 42:43.010]  I guess that's sort of like that.
[42:43.630 --> 42:46.270]  It doesn't really carry a lot of meaningful information.
[42:47.330 --> 42:49.530]  I don't really know the answer to that.
[42:49.790 --> 42:56.990]  Do you think it could be for the lay public, so they can then kind of like advertise and make it look, you know, kind of sexier?
[42:56.990 --> 43:04.270]  Yeah, my analogy is sort of like from putting on makeup to looking like Michael Jackson, right?
[43:04.450 --> 43:06.530]  Where is a fraud, right?
[43:06.630 --> 43:08.710]  Like putting on makeup, nobody blink, right?
[43:09.150 --> 43:10.950]  It's completely...
[43:11.790 --> 43:14.090]  Some of them look like putting on makeup.
[43:14.550 --> 43:20.050]  Maybe that's the intention, but you cannot put on makeup on scientific images.
[43:20.350 --> 43:23.790]  It's just your Michael Jackson immediately, anyway.
[43:23.950 --> 43:24.210]  Thank you.
[43:24.210 --> 43:25.710]  Weird analogies.
[43:27.190 --> 43:27.630]  Hi.
[43:27.850 --> 43:28.930]  Thanks for your talk.
[43:29.530 --> 43:35.270]  I'm wondering, has anyone studied the types of papers that are fraudulent?
[43:35.490 --> 43:39.970]  Like I just, in the first one, there seemed to be a clear political bias.
[43:40.510 --> 43:43.430]  And I'm wondering, is it because certain areas of research are more lucrative?
[43:43.730 --> 43:46.950]  Or is there, you know, threat actors at work here?
[43:47.110 --> 43:47.690]  That sort of thing?
[43:48.990 --> 43:52.890]  Yeah, that's kind of a super complicated question.
[43:53.150 --> 44:03.570]  Like very hot topic research paper definitely got cited a lot because it's like, you know, like papers on Alzheimer's and COVID vaccines are like widely cited.
[44:04.650 --> 44:06.970]  Some other papers are not really white cited.
[44:07.270 --> 44:11.610]  They're just added to the person's resume to show that they're highly prolific.
[44:12.310 --> 44:15.450]  Some of the paper are homeless because nobody even read them.
[44:15.990 --> 44:18.750]  You know, it's just like they're just junks in the literature.
[44:18.750 --> 44:20.110]  They're the forever plastic.
[44:20.690 --> 44:21.570]  But you're right.
[44:21.670 --> 44:25.710]  Some of the paper can have very significant harm to public health.
[44:30.590 --> 44:44.990]  So with the problems of capitalism causing this, I just wanted to ask if you're familiar with the work of Alexandra Elbikyan and Sci-Hub and all of the things that are, I guess, taking on the economic base of the publisher cartel.
[44:45.850 --> 44:47.250]  Feel free to send it to me.
[44:47.350 --> 44:47.970]  I'd rather...
[44:47.970 --> 44:49.190]  I would love to read it.
[44:49.390 --> 44:49.510]  Yeah.
[44:49.510 --> 44:51.990]  Well, Sci-Hub is basically...
[44:51.990 --> 44:52.250]  Oh, yeah.
[44:52.250 --> 44:52.810]  Sci-Hub.
[44:52.890 --> 44:56.210]  Sci-Hub is the one that a lot of my friends use.
[44:56.570 --> 45:00.130]  I am fortunate because I am at Columbia.
[45:01.670 --> 45:02.230]  Hopefully...
[45:02.230 --> 45:02.770]  But...
[45:02.770 --> 45:17.530]  And I was also wondering if there's any, like, grassroots effort among academics to, like, get Columbia to stop sending all of the money to these publishers that don't even do due diligence or at least force the hand of the publishers.
[45:17.530 --> 45:18.610]  Or is it...
[45:19.550 --> 45:20.050]  Or...
[45:20.050 --> 45:20.230]  Or...
[45:20.230 --> 45:20.890]  Or...
[45:20.890 --> 45:21.010]  Oh...
[45:21.010 --> 45:23.590]  I cannot say this online.
[45:23.730 --> 45:23.970]  Yes.
[45:23.970 --> 45:24.690]  I can talk about...
[45:24.690 --> 45:26.170]  Yeah, I cannot...
[45:26.170 --> 45:28.550]  Because, you know, I want to keep my job.
[45:29.510 --> 45:31.690]  but I can chat afterwards.
[45:33.250 --> 45:35.790]  Thank you for your research and your talk.
[45:36.670 --> 45:55.890]  I wanted to know if in the publication process if the researchers have to submit like their original materials like they're embedded within like their... within their paper like the pictures of the PDF and if not would that be helpful to people doing fraud research like you to have the original like pictures PDFs embedded?
[45:57.490 --> 46:10.610]  Yes, so now major journals are moving to the direction where they want to... they data deposit, they want raw data, they create like large amount of cloud to store raw data, raw images.
[46:10.970 --> 46:27.030]  But the problem is that the volume of paper and then the volume of us, it is like we have, you know, we are such a tiny, like you guys, we're such a tiny community compared to the vast of the capitalistic power out there.
[46:27.230 --> 46:31.950]  So sometimes it's very overwhelming to even think about the things out there.
[46:32.250 --> 46:33.210]  Yeah, but I agree.
[46:35.330 --> 46:36.630]  Two-part question.
[46:36.990 --> 46:39.170]  First of all, this is not new.
[46:39.390 --> 46:47.410]  So is there a fruitful area for going back and retrospectively just like you could take plagiarism software and look at old historical books and so on.
[46:47.550 --> 46:54.510]  Can you do this retrospectively and look at older articles because there's probably no statute of limitations on retraction of older stuff.
[46:54.630 --> 46:58.090]  And there's probably a lot of stuff back there that would show up really quickly.
[46:58.790 --> 46:59.730]  Second question.
[46:59.970 --> 47:05.590]  Have any of the people where it's been obvious fabrication ever come forward and explain their thinking?
[47:05.730 --> 47:09.030]  What the heck were they thinking when they did this just as a learning opportunity?
[47:09.290 --> 47:11.490]  Or have they all been totally silent?
[47:12.270 --> 47:13.810]  Second question first.
[47:14.050 --> 47:18.970]  Second question, the answer to that is always my postdoc or my graduate student did it.
[47:19.050 --> 47:20.210]  I have no clue what happened.
[47:20.210 --> 47:24.170]  And that's like 100% of the answer to that question.
[47:24.890 --> 47:29.190]  The first is like, yes, there is a lot of things.
[47:29.510 --> 47:32.850]  We, the Maslia case, we went all the way back to the 90s.
[47:32.990 --> 47:40.070]  I think the image fraud didn't really start kicking until Photoshop was used widely.
[47:40.070 --> 47:42.710]  So before that, the image fraud is probably not a thing.
[47:43.730 --> 47:44.230]  Yeah.
[47:44.430 --> 47:50.770]  And I think people back then did not make so much money being a professor or being a publisher as they are now.
[47:50.990 --> 47:54.590]  So that also changed in recent years.
[47:54.850 --> 47:59.670]  So in the older paper, there may not be as much incentive to do sort of stuff.
[47:59.670 --> 48:00.370]  Thank you.
[48:03.130 --> 48:07.990]  Are there any consequences for any of these journals for publishing garbage?
[48:08.550 --> 48:11.670]  And like if, what kind of consequences would you like to see?
[48:12.590 --> 48:16.710]  The consequence for them is like me shaming them on LinkedIn and Twitter.
[48:17.010 --> 48:18.290]  That's pretty much about it.
[48:18.690 --> 48:26.890]  And then sometimes they invite me to give talks at their conference and then to show that we're collaborating with the sleuths.
[48:26.890 --> 48:29.830]  And, but no, not really.
[48:29.970 --> 48:30.150]  Yeah.
[48:30.190 --> 48:32.990]  There's no consequences in terms of impact factor or whatever.
[48:33.270 --> 48:38.770]  It's just, it's just kind of, we're trying, we're hiring more people, we're going to do better and yada, yada, yada.
[48:38.990 --> 48:41.270]  But yeah, that's just, no.
[48:41.510 --> 48:41.690]  Yeah.
[48:41.810 --> 48:42.090]  Thank you.
[48:47.110 --> 48:58.830]  So I have no scientific background, but I know many nurses and medical professionals who are anti-vax and believe in a lot of things that I just think are fundamentally not based in science.
[48:58.830 --> 49:01.630]  And they send me scientific papers.
[49:02.970 --> 49:07.870]  And, you know, they're the authority, but I'm not because I don't have any kind of like medical background.
[49:08.350 --> 49:12.690]  So I guess my question is, in that case, like what is something a lay person can do?
[49:12.790 --> 49:17.990]  Like easy steps to take a scientific paper and just check like the fundamentals if they check out.
[49:17.990 --> 49:18.270]  Hmm.
[49:19.290 --> 49:24.270]  So like a lot of the things that, all the chemistry stuff, I don't understand any of it.
[49:24.510 --> 49:25.690]  And I can crack into it.
[49:25.810 --> 49:30.510]  A lot of things, it's just like, you know, it's just how you guys crack into things.
[49:30.550 --> 49:36.350]  It's all about the little, little, little details about like, you know, like how you crack.
[49:36.350 --> 49:40.810]  I don't know how you crack the code, but like, you see how I look at the error bars and stuff.
[49:41.070 --> 49:43.730]  Like, hmm, something is weird there.
[49:43.950 --> 49:49.250]  It's just like there's always, there's always some cracks on the edges that doesn't look right.
[49:50.810 --> 49:55.850]  And then, you know, I think people always hide the claim of like the scientific papers.
[49:56.090 --> 49:58.830]  There's so many shaky science, scientific papers.
[49:58.830 --> 50:02.030]  So, yeah, you can hit me up on socials.
[50:02.650 --> 50:02.870]  Okay.
[50:03.190 --> 50:03.290]  Yeah.
[50:04.810 --> 50:05.110]  Hello.
[50:05.710 --> 50:12.050]  I noticed that a lot of your methodology focuses around looking at graphs, charts, figures, that stuff.
[50:12.210 --> 50:16.230]  I don't know if that's because that's the most common place to keep these kinds of things, but...
[50:16.230 --> 50:19.210]  Because data don't work, and I learned the hard way.
[50:19.370 --> 50:30.170]  So my first two years as a sleuth, I worked on exactly one case, and I hit so many walls, and I was like completely traumatized because I was talking about data and nobody wanted to talk about data.
[50:30.490 --> 50:35.430]  And then, yeah, so like when you talk about data being impossible, then how can that be?
[50:35.590 --> 50:37.970]  And then people really don't want to listen.
[50:38.150 --> 50:42.330]  And I agree with you, the image stuff is just a tiny fraction.
[50:42.430 --> 50:42.690]  Right.
[50:42.870 --> 50:46.790]  So what I meant to ask is, do you ever look at other figures?
[50:46.930 --> 50:52.250]  I noticed you had at least one table in this presentation, for example, where numbers were duplicated.
[50:52.250 --> 50:55.650]  That would seem to make the statistics, which were different, invalid.
[50:55.910 --> 50:56.030]  Yeah.
[50:56.030 --> 51:02.270]  Do you ever find issues like this when simply doing the math, ultimately, using the figures presented?
[51:02.390 --> 51:12.470]  Yeah, so I guess we're kind of the different sleuths at different skill sets, and so I'm more of a visual person.
[51:12.470 --> 51:19.250]  So, like there are other people who are very good at reference, some people are very good at, like, you know, like we're...
[51:19.250 --> 51:24.470]  Again, like you guys, different guys have different skill set like, you know, so...
[51:25.190 --> 51:25.630]  Thank you.
[51:25.990 --> 51:26.150]  Yep.
[51:27.110 --> 51:28.690]  All right, thank you for the talk.
[51:29.150 --> 51:29.650]  A round of applause.
[51:29.650 --> 51:30.210]  Thank you very much.