HNHacker News
TopNewBestAskShowJobs

DanielleMolloy

2,467 karma · joined March 23, 2012

submissionscomments
DanielleMolloy··on Fake review factories that run on Facebook and post five-star Amazon reviews
They have to clearly separate between Marketplace and Amazon again. Everything used to be fine when people knew from whom they bought at any time. They are removing the line in multiple ways right now (one of the worst being comingling).
DanielleMolloy··on Fake review factories that run on Facebook and post five-star Amazon reviews
I check Fakespot, and click on some reviewer profiles. The idiots doing full-time positive product reviewing are so easy to spot by their hilarious review history that you start to assume that Amazon doesn’t do anything against them on purpose. They also mainly flock on the crap within a single month or so (but this is one of many Fakespot metrics).

I tried buying a new bicycle light last week and literally every product was crap with fake reviewing. After wasting an hour on this I ended up not buying from Amazon. Rolling out the red carpet for the Chinese crap and counterfeit industry on their platform is likely to become Amazon’s death if they don’t admit their mistake and turn the wheel now.

DanielleMolloy··on Riksbank Prize in Economic Sciences in Memory of Alfred Nobel awarded
Also see: https://en.wikipedia.org/wiki/Nobel_Memorial_Prize_in_Econom...
DanielleMolloy··on Physics Nobel won by laser wizardry – laureates include first woman in 55 years
+ This nobel prize is basically about a real life microscopic tractor beam, which already appeared in the original Star Trek in the 60s. Could be another invention inspired by science fiction.
DanielleMolloy··on Physics Nobel won by laser wizardry – laureates include first woman in 55 years
Thanks for the pointer, the Lemaitre case looks pretty bad. There seem to be several similar cases that were brought to light in recent years, maybe the committee has learned from such mistakes.

Especially for PhD scientists it is inspiring and motivating that you can make nobel prize-worthy discoveries at this level. This should not be hidden for keeping up power hierarchies.

DanielleMolloy··on Physics Nobel won by laser wizardry – laureates include first woman in 55 years
+ Arthur Ashkin, at 96, is the oldest nobel prize laureate ever ( https://www.nobelprize.org/prizes/lists/nobel-laureates-by-a... ), and declined an interview this morning because he is "busy doing research". He did much of is research at Bell Labs and has his own lab at home now.

+ The most cited Strickland / Mourou paper is this one: Compression of amplified chirped optical pulses. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.673... It has a humble length of 3 pages and has collected 4600 citations in 23 years.

+ This is a good time to remember Lise Meitner: https://en.wikipedia.org/wiki/Lise_Meitner#Nobel_Prize_for_n... and check out James Watson's 'opinion' of Rosalind Franklin: https://www.youtube.com/watch?v=fBWPzdyyPic But also note that Marie Curie was one of very few nobel laureates who got awarded twice, and she got her awards very early on in the history of the nobel prize ( https://en.wikipedia.org/wiki/Category:Nobel_laureates_with_... ).

DanielleMolloy··on Hundreds of academics at top UK universities accused of bullying
There were three recent scandals at MPIs a few weeks to months back, the most recent one being: http://www.sciencemag.org/news/2018/08/she-s-world-s-top-emp...

I'm glad that this new Guardian article highlights that it is a structural problem. When reading up on the MPI cases it seemed as if it was a problem of few labs only. Due to the power imbalance, as a to-be-PhD student you are strongly advised to experience several labs before making a decision where to go. Don't only look at metrics such as reputation. It is incredibly hard to find a good PhD advisor, and intuition trained by good and bad experiences does help.

Similarly bad are people in those labs who don't recognize the behaviour as such because they either have never experienced a non-dysfunctional lab or see benefit in licking upwards (they may hope they can be in the intimidating position one time). Every party enabling this seems to confuse abusiveness with strength and assertiveness. Is it Stockholm syndrome?

The supervisor's perspective is that grant pressure can be very stressful. But as soon as multiple people agree to the point that they file formal complaints the problem can't be the PhD student anymore.

I have still proportionally seen more functional than dysfunctional labs in science though, great labs do exist. I would expect that the problem is much worse in industry than in academia.

DanielleMolloy··on How counterfeits benefit Amazon
Based in central Europe, have received counterfeit products that were indistinguishable from the real one. On the surface counterfeits have become difficult to identify, it is not about blurry and misspelled labels anymore. Especially if it comes to electronics they are not less dangerous. Non-electronic items like books get counterfeited too.

But I had more trouble with fake reviews selling me mediocre (often Chinese) products. I check the same US Amazon product on Fakespot for almost everything now (but the fake reviewers will game that system too). Amazon is clearly outsourcing their job to the customers.

Been a (happy) Amazon customer since I'm legally allowed to buy there and liked marketplace in the beginning for second-hand products and buying from smaller scale shops (the scale of the family winery in Saarland), but rolling out the red carpet for the Chinese counterfeit "economy" is just dangerous given that Amazon's number one rule that made them successful is keeping customer trust. Co-mingling - I don't see how anybody at Amazon could have believed that 1-day shipping is a more important metric than selling genuine products.

I wish they temporarilly blocked Chinese marketplace sellers as a whole until these issues are solved, by whatever entity. As a customer I don't care that this is unfair towards a handful of moral sellers, I care about not getting shocks from electronic equipment that I bought at full price. When I want to buy terrible Xiaomi products for some reason (e.g. because I fell for the Apple counterfeit's fake reviews) I can still go to eBay or Alibaba like back in the days.

The other solution would be visibly and strongly separating marketplace again, like in the beginning, or putting it on a separate dedicated website that crosslinks with main Amazon.

DanielleMolloy··on A Cornell Scientist’s Downfall
Starting to get mixed feelings about these cases. Impostors and fraudsters certainly need to be exposed. But publicly ripping apart few individuals who, together with their peer reviewers, in some cases may simply not know better (e.g. because their whole field has low standards – nutrition science seems to be one of these fields –; or because they are inexperienced and their co-authors and supervisors don't care about details and correct science and just want to have their names on the paper) can not be the solution for an overall malfunctioning scientific communication and incentive system, or for bad methods that are used by entire fields.

The current situation also invites all sorts of abuse. Need to say that certain parts of the open science movement appear to be a bit angry, not to say sadistic. Wansinks case was one that was ripped apart on Twitter. Wansinks research has received lots of media attention before - would his statistical methods have been investigated and debunked without that? How many Wansinks are out there, still having their job, continuing to apply bad methods (while not committing active fraud), never to be called out for their mistakes? I've been in Japan during the Obokata stem cell case. Her public apology was broadcasted on national TV. In konbinis you could see various tabloid newspapers with her name and her crying face on the cover for weeks. There seemed to be different mechanisms at play than a genuine interest in correct science. One of Obokata's supervisors, Yoshiki Sasai, committed suicide as a consequence of the case.

DanielleMolloy··on Praying mantis seen hunting for fish in wild for first time
> [1] I had an African flower mantis as a pet ;)

types—this—into—google—images What the..! This kind (flower mantices) really looks incredible!

DanielleMolloy··on Thousands of scientists publish a paper every five days
Some agencies issuing grants have started only allowing mentioning n (e.g. n=10) important publications in grant applications. I think this is a good temporary workaround for some of the bad incentives in the current system.

You will give a negative impression if:

+ you can not find 10 high-quality articles in your body of work

+ you are not the first author on any of the 10 articles

+ influential work is more positively rewarded than with the h-index

DanielleMolloy··on Deep image reconstruction from human brain activity (2017)
Blog post that has some info on the new GAN-based mind reading ideas from last year (including this one):

https://www.mindcodec.com/using-gans-brain-reading/

DanielleMolloy··on The Age That Women Have Babies: How a Gap Divides America
This can be attributed to people who work themselves being independent from a potentially abusive person's income; and to widely spread atheism, much more than in most European states (currently still 50% atheists (approximately, depending on the survey)).
DanielleMolloy··on The Age That Women Have Babies: How a Gap Divides America
Ideological indoctrination aside, it may suprise some people here that the german GDR had 92% working women, many of them in technical and scientific fields and many having three children. One of the things they did right here is offering proper childcare widely around typical work time windows. The society and often husbands too felt responsible for children, it was not seen as "a mother's thing" like many in this thread seem to do. This has implications up until today, with several prominent and well-educated female politicians and public figures being GDR-born. It was not all perfect (housework and childcare was implicitly still seen as a mother's thing, mothers could request special housework days, sometimes questionable things like full-week kindergardens and so on), but much better than what young people face today when moving to west-german states (where the jobs for educated people can be found) where old-fashioned mindsets and kindergardens closing at 4pm and at random full days throughout the month persist. A single generation of children born among both men and women working in decent jobs did this.

Angela Merkel is one of these GDR children, she studied physics and has a PhD in quantum chemistry. She is childless, but her mother - despite being very old now - is still teaching English because she has never stopped being passionate about her chosen profession (I know an interview where Merkel says that her mothers work attitude always inspired her). Well, she never had to throw her education away because society or her husband expected it either. Another well-known politician I can think of here is Sarah Wagenknecht. Peter Scholze who just won the Fields medal was born in Dresden (growing up in Berlin) - his father is a physicist and his mother is a computer scientist working as network technician.

DanielleMolloy··on Creatures: Artificial life software agents (1997 paper on AI game)
So, does current AI research have access to a similar artificial environment playground simulating:

* body biochemistry

* genetics / artificial evolution

* perceptual systems

* motivations / drives

* the interplay of different brain lobes

* environmental processes like weather

Can't think of one.. Has AL+AI research disappeared? The success with OpenAI DOTA is all great, but is DOTA (with its clearly defined pragmatic goal states) really more interesting for AI research than an autonomous artificial agent?

Afaik this game saw one of the few applications of multi-layer perceptrons / neural networks in commercial games.

DanielleMolloy··on New MPI scandal: World's top empathy researcher accused of mistreating employees
Many more details in this piece on Buzzfeed: https://www.buzzfeed.com/de/pascalemueller/mobbing-max-planc... (german though)
DanielleMolloy··on 'Intelligent' crows to pick up litter at French theme park
I don't have an explanation, but I recommend looking into the work of Onur Güntürkün who is doing research on bird brains.

Güntürkün: Cognition without cortex. https://www.cell.com/trends/cognitive-sciences/abstract/S136...

Press release: http://news.rub.de/english/press-releases/2016-03-03-complex...

His talks are pretty good, there may be higher quality recordings than this one though: https://youtu.be/nHTTVFDjQ1Q

DanielleMolloy··on Systems Neuroscience Is About to Get Bonkers
As an example, this is a great article detailing out how the visual system and convolutional neural networks use similar layers (representations) of processing: https://neurdiness.wordpress.com/2018/05/17/deep-convolution...
DanielleMolloy··on That $35 that scientific journals charge goes 100% to publisher, 0% to authors
This. I hope more people start thinking along these lines so that we can find a solution that respects and exploits the useful boundaries and incentives in the system, resulting in better science.

> For that to happen, we would need a group of already reputed and respected researchers to review the submitted articles. But because the journal fees would be avoided, that means the money could be collected to pay those reviewers.

Scientists are motivated by curiosity for their field, prestige, and having a stable job position were they can follow these motivations without fear. But they are not motivated by money (grant applications are a necessary evil). Due to this I don't think that you can start paying professors with stable positions, but you could start permanently hiring those many highly qualified mid-career academics who are still on the way towards a permanent position (and may never reach it). The review quality will be improved if people review full time, which can lead to journal prestige, faster review times, reviews with a broader look on science.

DanielleMolloy··on That $35 that scientific journals charge goes 100% to publisher, 0% to authors
Trust building and a seal of quality through peer review.

Basically, a paper that is published in a journal like Nature or Science or the field-specific top journals is assumed by the community to have been reviewed at much higher standards than those published in lesser known journals, just because it is much, much more difficult to publish in these journals. Thus a paper there will be a signal that you are a capable scientist.

Also, peer review is necessary as protection against crap. I am not saying that this is a fool-proof system, but science has to have one, preprints should exist but science needs to be reviewed and approved by professionals in some way, and the professional reviewers need to be incentivised to do a decent job.

Peer review may be the only asset keeping journals alive at the moment. I've been wondering whether traditional journals can be overcome by making high-quality peer review the focus of the business model.

DanielleMolloy··on Thank You, Robert Voit, Creator of Paint Shop Pro
I have not, thanks for the pointer. GIMP is still a software I would like to like; will give it a try.
DanielleMolloy··on Thank You, Robert Voit, Creator of Paint Shop Pro
PSP7 was a favorite piece of software that I had been using for years when still mainly on Windows systems. It was useful and intuitive when having to make changes to plenty of raster images, or when doing pixel graphics; creating alpha transparent sections, clean ups, and so on. I also used Paint.NET a lot for batch edits.

It still seems a bit difficult to find a replacement for these two on macOS and Linux. Intuitive interfaces focusing on primary image editing functions and stable output to various formats are probably what is missing for non-professional users like me.

GIMP seems like a great project but has a terrible overblown UI. I tried Photoshop / Illustrator for a year as you get the CS4-subscription at a low price from my university. A few features, e.g. how embedded images were handled are nice, but the software is massive and the interface is so weirdly complex that I had to search for guides for the simplest tasks (who is profiting from having to "study" CS4...?). With PSP7, which is decently powerful I went through a couple of tutorials and was set up for years. Also, I had spent half an hour to deactivate all the spyware that comes with CS4, this is a no-go.

You can read that I am just really ignorant about modern graphics editing workflows. I'm not a graphics professional, but need to make "production level" graphics every few months. I need relatively technical vector illustrations (e.g. of experiment flows, NN models) and clean figures of scientific results (raster) in a visualization-heavy field.

For vector graphics I now really like Inkscape with its SVG basis. I still go back to Dia a lot though whenever the illustration fits their limited (but sufficient) scope. Both can be used on Linux and macOS.

Still searching for a powerful but plain & simple (and stable!) raster graphics editor though. Currently Krita looks nice, but it is clearly in development stage. Would pay for either Acorn or Pixelmator if I wasn't concerned that they will also start emulating the UI of Photoshop in the future.

DanielleMolloy··on Zuckerberg defends Facebook users' right to be wrong – even Holocaust deniers
The german school system is not about indoctrination or assigning guilt.

The post war german generations were not teached that they are guilty, but that they do carry responsibility for being aware of how easily facism and dehumanisation happens and escalates, and of keeping the memory of such evil present.

And yes, this includes the responsibility of being aware that similar things may happen again, clearly detecting extremism by the root, calling it out and hopefully averting it. Probably every country in the world should put such emphasis in their history lessons, but nobody else ever really will for obvious reasons.

So this kind of became a responsibility of the people born in modern germany. This is why it is one possible Abitur topic in politics, history and german lessons, why it is one element of the curriculum in (3 to 4)+2 different middle school years + Abitur (saxony speaking) and why there are high school visits to concentration camps.

DanielleMolloy··on Zuckerberg defends Facebook users' right to be wrong – even Holocaust deniers
Holocaust denial is, on the root, pure antisemitism, which has been an evil element of every Western society since the beginning of the diaspora. The ideology behind it is pretty much equal to that part of nazi ideology (and I'm not one of those who is quick to run around calling everybody nazi these days).

As someone whose school system has provided them with some insight on how this all happened I see the US absolutist view on free speech when it is about outright lies like Holocaust denial as a bit dangerous. Preventing deluted individuals from spreading a life-threatening lie is not censorship. I see that the US view is coming from a very humanist view here, but tolerance needs to have limits set by basic human rights values, otherwise it will be destroyed by itself sooner or later.

The approach to "educate the deniers" is indicating to them and the listeners that you are taking them and their delusions seriously, and dissiminate them in the process. In the current debates about e.g. evolution you clearly see what happens in wider society if you talk to the 1% "skeptics".

DanielleMolloy··on Keeping a plaintext “did” file
It will create a second file in the directory, renamed with information where it comes from. No problems here as it is merely a text file.

There used to be an Android app for todotxt, but it does not seem to exist / work anymore after the Dropbox API changed.

DanielleMolloy··on Keeping a plaintext “did” file
http://todotxt.org has a format for a done.txt list, too. Useful before project meetings to get an overview over what you did in the last days.

TodoTxtMac creates this for me: https://mjdescy.github.io/TodoTxtMac/

I've been using it together with Dropbox (for access from my phone) for 3 years now.

DanielleMolloy··on Thailand cave rescue: all 12 boys and coach successfully rescued
Actually this started when a random guy on Twitter asked him whether he could help: https://twitter.com/MabzMagz/status/1014251869760249856

I got the feeling that he just jumped in on that, without further thinking (which at least to me seems sympathetic). Got the feeling that it rather hurt their PR as people were quick to call it a PR stunt, which is foreseeable if you mainly think about PR.

DanielleMolloy··on The relativistic discriminator: a key element missing from standard GAN
Interesting, especially the performance on bigger images, and it looks like a low-effort modification of many standard GAN losses. Seriously, I want to give this a try right now. What do GAN researchers think about this paper?

It should also be appreciated that it comes with code and a short blog post.

DanielleMolloy··on Opinion: Elsevier are corrupting open science in Europe
The business model of these large publishing companies is not selling access, it is peer review. Science and Nature keep their reputation despite Elsevier because it is notoriously hard to pass their peer review process, and the academic world trusts it.

We may live in a better academic world if rigorous peer reviewing was a full-time paid job. Presumably it would not be difficult to do better than 90% of the current peer reviewing by letting people with full-time experience (and the overview of the state of science that comes with it) do it. It may also solve a couple of other problems with anonymous peer reviewing.

(eventually with ever-more complex methodology and involvement of complex programming routines and tools we still have to get rid of the scientific paper as the main form of communication)

DanielleMolloy··on Van Gogh’s Japanese Idyll
The article missed showing the imho coolest of van Goghs Japan-inspired works, "Père Tanguy":

https://en.m.wikipedia.org/wiki/Portrait_of_Père_Tanguy

https://www.wikiart.org/en/vincent-van-gogh/père-tanguy

"Almond blossoms" is really beautiful, and indeed very reminiscent of cherry blossoms. I like its bittersweet background story a lot.

← PreviousPage 8 of 12Next →