aschreyer · HackerTrans

aschreyer

no profile record

comments

vor 10 Jahrendiscuss

The most likely risk are not the majority of psychedelics themselves, but the quality of the substance you are getting as well as potency and "therapeutic" window. As they are controlled substances, what you are getting is often synthesised from reagents that are known carcinogenics and poorly purified. In many cases, it is not the real substance but a replacement or a mixture of various other chemicals that is supposed to mimic superficial effects. In addition, some psychoactive substances are active in the ug range and therefore have a high likelihood of leading to severe side effects because of overdosing. Also, the dosage window and thus the difference between hypnotic and anxiogenic/severe side effects can be in the range of mgs as well. Therefore one should be cautious if you cannot get access to them in a controlled/legal environment and certainly should not try every research chemical or analogue that appears on the internet.

vor 11 Jahrendiscuss

A big problem are the perverse incentives researchers are facing. The only thing that matters is publishing in prestigious journals, otherwise your career is pretty much over (no real chance to get tenure). Universities often don't even check the publications, and researchers usually omit the actual title of the publication and only keep the name of the journal. Moreover, if you want to get tenure you need to work on "hot" topics (as in CRISPR at the moment or cancer in general), meaning you don't necessarily spend too much time on one subject. I'm sure researchers are often aware that the research results are not ready for publication yet, but it's either publish or perish, and you usually get away with publishing sub-par results because no one really follows up on them either. Cancer and Alzheimer are two fields were billions have been wasted by research grants and pharma companies because the initially promising drug targets for example could not really be validated. I'm afraid this will stay that way as long as the career dilemma researchers are facing does not change.

vor 11 Jahrendiscuss

The "second concern held by some" is clearly a rhetorical figure of speech to make it sound as if they are not really sharing it, although the rest of what the authors say only deals with this exclusively. The first concern complete disappears. They are also using the term parasite or parasitically throughout the paper, which is not really helpful in this context.

The real concern of the authors, as it appears to me, is that you "own" the data that you produce and should have the exclusive right to use it - they call it "obvious extension of the reported work".

How would data sharing work best? We think it should happen symbiotically, not parasitically. Start with a novel idea, one that is not an obvious extension of the reported work

On the other hand, if you have a novel idea, you are supposed to work "symbiotically" with the authors with relevant coauthorship:

Third, work together to test the new hypothesis. Fourth, report the new findings with relevant coauthorship to acknowledge both the group that proposed the new idea and the investigative group that accrued the data that allowed it to be tested

The problem is that science does not work this way. You cannot own the facts, as someone famously said. Imagine people in computer science or mathematics held the same attitude, especially in artificial intelligence/machine learning.

vor 11 Jahrendiscuss

You could translate it as "how dare you trying to sabotage my career by attacking my research", something like this. It is fairly common that researchers build their careers on single results/breakthroughs and those have to be fiercely defended against everyone. For example, by rejecting/delaying research grants/papers that are competing with your own opinion/results, thanks to anonymous peer review. Sounds crazy, but I have seen this more than once.

vor 11 Jahrendiscuss

I think this paper summarises everything that is wrong with academia and why a lot of people have enough and are looking for ways out.

A second concern held by some is that a new class of research person will emerge — people who had nothing to do with the design and execution of the study but use another group’s data for their own ends, possibly stealing from the research productivity planned by the data gatherers, or even use the data to try to disprove what the original investigators had posited.

According to the authors, trying to reproduce results or possibly disproving them is worse then theft. That already says it all. Especially in medicine, where reproducibility is severely lacking and research fraud is certainly far from uncommon.

There is concern among some front-line researchers that the system will be taken over by what some researchers have characterized as “research parasites".

Wow, strong language here. Imagine everyone who used someone else's results to advance science would be called a research parasite. The authors have an extremely cynical view on science and are simply equating it to business and career.

To be honest, I am not surprised that this is coming out of NEJM or another medicine/molecular biology journal as this is a fairly common attitude once you rise in the academic ranks in those areas.

vor 11 Jahrendiscuss

You don't necessarily have to only lift in order to gain health benefits. Generally I would simply recommend to pick a sport and try to become the best you can be and the rewards will follow soon, not only physically but also psychologically. Could be anything, cycling, running, tennis, martial arts - lifting works great of course. I myself switched from lifting to gymnastics strength training and couldn't be happier.

vor 11 Jahrendiscuss

Maybe it is because the anglophiles assume that because they feel Shakespeare is such an outstanding figure in their own culture he must have the status in every other and cannot possible fathom something else. The truth is though, here in Germany (and surely in countries like France), Shakespeare is overwhelmingly known only through English media/culture such as TV shows, movies etc. and as far as I know not on any educational plan at all. Why would it be? There is such as vast amount of German writing (in the past at least) that one can easily do without him. I guess the problem is also that there was historically little overlap between English (i.e. British) and continental poetry (although translating Shakespeare into German was done by poets such as Schiller) and coincidentally, philosophy as well, at least that would be my impression.

vor 11 Jahrendiscuss

I think it would be too easy to say that it was only the German government responsible for what happened. Below is a key quote:

What has happened? What has happened is quite simple. Those elements of the population which are both brutal and stupid (and these two qualities usually go together) have combined against the rest.

And another quote from Hannah Arendt:

Only the mob and the elite can be attracted by the momentum of totalitarianism itself. The masses have to be won by propaganda.

The problem might be that the "mob" is getting stronger for various reasons and there is an active elite that is eager to exploit them for their own gain.

vor 11 Jahrendiscuss

Einstein's most genius insight was probably not to make any initial assumptions and simply to follow his thought experiments regardless of the outcome (even if they contradicted contemporary dogma). The idea that time is not constant and space can be bend was certainly genius and as far as I know rejected by others who thought the maths must be wrong. He certainly relied on others for the mathematical foundation but he always mentioned that. Science even today is strongly ego-driven and he surely had the personality to take on the other scientists of his time.

vor 11 Jahrendiscuss

> In Germany, it is thanks to European monetary policy that artificially depresses the Deutsche Mark, allowing for cheaper exports and lower costs for German companies, as well as a captive market whose currency is overvalued in exactly the same way destroying local industry (i.e. German loans pay for German cars to be sold to the Europeans but the other way round is much harder). Hello VW, goodbye Rover. On top of this Germany is absorbing East Germany which is a handy source of incredibly cheap labour (in the same way that illegal labour is in the US food and farm industry) AND massive subsidies for whoever wants to go take advantage of it (I know a company that opened a very, very expensive plant 200km from Berlin because the subsidy was such that they came out profitable even if the plant lost money for 20 years).

Fascinating story, if only Germany had started producing cars in 2000 with the introduction of the Euro. How do you think VW, Audi etc. fared in the 70's and 80's (and the rest of the German economy in general)? In addition, the Euro is pretty much irrelevant because the cars are mostly produced in the region where there are sold - do you think the exchange rate is relevant for the VW factories in Mexico, Tennessee and Pennsylvania for the north American market? How does a cheap Euro make the import of resources from outside the Eurozone cheaper, steel, aluminium etc.?

vor 12 Jahrendiscuss

In Germany the cutting edge research is usually done by specific research organisations that do no or only very little teaching such as the Max Planck Institutes, Helmholtz Association, Fraunhofer, etc. Of course, as a student you do not really benefit from that at the undergraduate level - those world-class researchers are usually too busy to teach and leave it to their assistants.

vor 13 Jahrendiscuss

PL/Python is NOT supported: http://www.databasesoup.com/2013/11/first-look-at-postgresql...

vor 13 Jahrendiscuss

PL/Python is very unlikely to appear soon unfortunately because it is an untrusted language in PostgreSQL.

EDIT: you are right, PL/Python is supported as well. I only read the "Language Extensions :PL/Perl, PL/pgSQL, PL/Tcl" part at the top.

vor 13 Jahrendiscuss

It is not stupid it actually makes sense for ETL workflows where you load raw data into unlogged tables and transform them in the database into the final (logged) representation.

vor 13 Jahrendiscuss

Could be an embassy.

vor 13 Jahrendiscuss

[Current PostgreSQL user]

Oracle-like partitioning syntax support (including composite) would be great - although possible through inheritance and triggers/rules, it is the only bigger feature missing in PostgreSQL in my opinion. The current form requires a lot of overhead but is definitely worth the effort when working with 1B+ rows.

vor 13 Jahrendiscuss

Thanks, that was a good example indeed! Improving the memory traffic was actually really important because the real-life application has N=20M+, up to 200M. Around 85M calculations per seconds is pretty impressive I have to say and the example really helps to understand how to write efficient Numba code.

vor 13 Jahrendiscuss

The Ultrafast Shape Recognition (USR) algorithm is a very simple yet interesting application used in drug discovery that I tried speeding up with Numba (the similarity calculation part). The NumPy implementation looks roughly like this:

    def usr(X, y, S=0.9, N=10):
        scores = 1.0 / (1.0 + 1/12.0 * np.abs(X - y).sum(axis=1))
        scores = scores[scores>=S]
        scores.sort()
        
        return scores[-N:][::-1]

Where X.shape could be (2000000, 12) (or more rows) and y.shape (12,). The idea is to retrieve the top N most similar hits above a similarity score of S.

vor 13 Jahrendiscuss

It is a bit unfortunate that you have to "unroll" your code to get the most out of Numba. Hopefully at some point it will be able to translate NumPy functions such as np.abs(X - y).sum(axis=1)) efficiently into LLVM. Those extreme performance improvements are misleading though; in my experience the speed up compared to concise, optimal NumPy code is more in the range 50-100%.

vor 13 Jahrendiscuss

Here "organic form" means organic as in organic chemistry, not organic food.