Comment by DoctorOetker | Hacker News Reader

Comment by DoctorOetker | Hacker News Reader

Parent Full thread

> To observe entanglement between top quarks, the ATLAS and CMS collaborations selected pairs of top quarks from data from proton–proton collisions that took place at an energy of 13 teraelectronvolts during the second run of the LHC, between 2015 and 2018. In particular, they looked for pairs in which the two quarks are simultaneously produced with low particle momentum relative to each other. This is where the spins of the two quarks are expected to be strongly entangled.

I believe it would be much easier to generate public support and curiosity if the data could be made more accessible.

Suppose the research articles explicitly told what URL was used to request the relevant slice of the data, which selectors to use etc. so that Average Joe could download the slice of events investigated (and thus learn how to issue other slices of events) and bulk download them for analysis.

Allow the public to interact with the data and the juices -err- funds will start flowing.

Make it more didactic and less ex cathedra.

> so that Average Joe could download the slice of events investigated (and thus learn how to issue other slices of events) and bulk download them for analysis

Is there any precedent for such openness generating public support?

Not saying it wouldn't be scientifically productive. Or even engaging for enthusiasts. But I'm having trouble seeing a politically-relevant number of Average Joes diving into collision datal. (That is already publicly accessible [1].)

[1] https://opendata.cern.ch/record/80020

An out-of-band URL to the data without instructions on how the slice of interest was requested is not what I described.

If average people can follow the steps in detail, they will probably not contribute observations of significance, but the ivory tower effect would diminish. Otherwise people will just think of all the particle accelerators as very expensive nonsense they don't support.

And yes, I think there is a lot of precedent: look how the LLM community is blooming and all the money being poured in. Average Joe can interact with a chatbot online, Average Joe runs into tutorials and step by step instructions to run a local LLM, etc.

> look how the LLM community is blooming and all the money being poured in. Average Joe can interact with a chatbot online, Average Joe runs into tutorials

Average Joe is consuming chat bots. He is not raising money or taking tutorials.

The mere act of consuming brings money to the field.

Average Joe delegates part of his opinion to the friends or family or idols that are the closest proxies to a certain field. If those think poorly of a certain initiative because it feels exclusive or not down-to-earth enough it will fan-out to all the Average Joes they know.

I'm not retarded but I am still not finding the 2017-2018 data through the URL you cited, guess how I will describe it to any acquaintances? Guess in what sense I will precondition the (dis)continuing validity of my comments upon the presence or absence of clear instructions so average joe can reproduce a finding?

If it feels like Average Joes are requesting to hold their hand like a toddler, one is being exclusionary.

Tell Average Joe that CERN invented URLs.

The first big application of URLs was precisely this use case: distributed access to particle event data in databases and tape silos across different international institutions. The mainframes (mostly IBM 3090) would then crunch the data to analyze and filter the interesting collisions.

But that would convince them to sponsor the particle physicists of the day, not of today?

We already make the data available publicly...

http://opendata.cern.ch/

please show me the low barrier to entry

areas of improvement:

* where do you download the 2017-2018 data for ATLAS?

* make the methods and data section part of the PDF or provide additional PDF's (some people will opt to "download PDF" for later perusal, and then be disappointed they have to use the browser to render the methods section to PDF as well)

* provide explicit URL's in the data section

* data and code availability upon request: why would you want to organise the bothering of physicists and engineers with N (interested parties) times code and data requests if you can provide them as downloads 1 time?

Reply on news.ycombinator.com