[Popular STEM] Curating the Internet: STEM digest for January 2, 2021

A YouTube video and transcript discussing the question of whether "time" is real; Studies find that machines are changing the language that people use; An argument that increases in scientific research volume might have diminishing returns in terms of scientific progress; Descriptions of how people shape technology and are shaped by it in return; and Encryption advances by IBM and others make it possible for computers to process encrypted files without decrypting them first

Administrivia

Experimental: Please leave a relevant and substantive reply to this post in order to be considered for a gratuity from @penny4thoughts after the post pays out.

Links and micro-summaries from my 1000+ daily headlines. I filter them so you don't have to.

First posted on the Steem blockchain: SteemIt.

1. Is Time Real? What does this even mean? - Sabine Hossenfelder discusses the concept of time, first pointing out how difficult it is to define time without making use of circular definitions. She then moves into Einstein's result that time is a dimension of Space-Time, which makes it possible to measure time, without circular-definitions, by observing the intervals between certain recurring events - like a year being the interval during which the Earth revolves around the Sun. According to Einstein, then, time is real. But Hossenfelder goes on to mention that most physicists think that Relativity is an approximation for a deeper theory, like Quantum Gravity. In some of those deeper theories, time is not needed in order to explain real-world observations, so the question of "Is time real?" boils down to a question of which of these competing theories is correct. So, in conclusion, time certainly feels real, but it might be an emergent propeprty that arises out of some deeper phenomenon that has yet to be fully understood.
   
   

   Here is the video:


2. How machines are changing the way companies talk - Here's a possible example of Goodhart's Law in action. When a measure becomes a target, it ceases to be a good measure. Knowing that natural language and AI tools will be parsing their mandatory and voluntary public disclosures, this article reports that executives have started modifying their language to include signals that will lead to advantageous reports from the machines. The conclusion was reported in an October paper and was reached by examining 360,000 SEC filings from 2003 and 2016. During this time, filing downloads increased from 360,000 to 165,000,000, this amounts to an increase from 39% of downloads to 78% in the SEC's EDGAR database. In 2011, study found that a majority of words that a Harvard dictionary identified as negative weren't actually negative in a financial context. Sure enough, after that study was released, the frequency of those "negative" words decreased in the corporate financial disclosures. Another study this year found that startup's pitches with "positive words" are more likely to receive funding, and yet another study found that hedge funds that automate the download of corporate filings do better than those that do not. -h/t Communications of the ACM


3. Peer-reviewed papers are getting increasingly boring - Estimates are that the number of scientific researchers is doubling every 16 years with an even greater increase in the number of publications. However, this may not be translating to corresponding advances in scientific knowledge. Is it possible that research has past a point of diminishing returns? Daniel Lemire reports on a number of papers that might suggest that reform is in order. In particular, Cowen and Southwood argue that the higher volume translates to a greater cost and makes it harder to find the true scientific advances. Bhattacharya and Packalen suggest that the trend towards more papers and incremental advances drives researchers away from high-risk exploratory projects that could create the most robust advances. Thurner et al. echo this theme, finding that the percentage of out-of-the-box projects is declining and Collison and Nielsen suggest that science is slowing enormously on a per-dollar or per-hour basis. Against this backdrop, Lemire suggests that researchers should challenge the status-quo by looking past numbers of papers and numbers of citations as metrics for a researcher success. This might be done by broadcasting research meetings online or emulating the model of movie critics where analysts delve past the superficial numbers and into the practical merit of a research paper. While potentially useful, these techniques are expensive. In closing, the author suggests two guiding principles:

   1. Seek objective feedback regarding the quality of your own work using “customers”: people who would tell you frankly if your work was not good. Do not mislead citations or “peer review” for such an assessment.
   
   2. When assessing another research, try your best to behave as a customer who has some distance from the research. Do not count inputs and outputs as a quality metric. Nobody would describe Stephen King as a great writer because he published many books. If you are telling me that Mr Smith is a great researcher, then you should be able to tell me about the research and why it is important.



4. How electric lighting changed our sleep, and other stories in materials science - Coincidentally tying together "time" from the first article, and "technology shaping people's behavior" from the second, this article introduces us to a book by Ainissa Ramirez, Alchemy of Us. Before the electric light, people slept in two segmented blocks overnight. So-called "first sleep" lasted for three or four hours, after which people would awaken for an hour or two before going back to bed for "second sleep", which lasted until morning. Electric light, however, lengthened the day and caused people to stay awake later, combining first and second sleep into a longer solitary time window. This history is supported in literature all the way from Homer's Odysee until novels and news sources in the 19th century. The book contains a number of demonstrations of ways in which, "we shape materials, and are shaped by them in turn". Another example from the book is how, in the days before radio, Elizabeth Ruth Belville (a.k.a. The Greenwhich Time Lady) made the rounds daily to see her 200 clients in London who paid for the privilege of synchronizing their own timepieces with Belville's pocket chronometer. Click through to the article containing this, and several other stories like that along with an interview of Ramirez.


5. IBM Makes Encryption Paradox Practical - Fully homomorphic encryption (FHE) enables processing of encrypted files without the need to decrypt them. IBM is the latest firm to approach this challenge, and earlier this month the company announced an online demo where companies can explore the technique with their own encrypted data. Traditional cloud security approaches have presented a security risk because even if data is encrypted at rest, it is normally decrypted for processing, which exposes it to prying eyes. With FHE, however, companies can rest assured that even during processing, data is still secure from prying eyes. The article contains this explanatory excerpt:

   Yet with a workable and reasonably efficient FHE system, even the most heavily encrypted data can still be securely processed. A customer could, for instance, upload their encrypted genetic data to a website, have their genealogy matched and sent back to them—all without the company ever knowing anything about their DNA or family tree.

   IBM has reported previous successes in the field staring in early 2020, and the latest announcement makes the technique available to customers who might be less, "crypto-savvy". It's not clear whether IBM's offering is technically superior to other alternatives, but they seem to have taken the lead in offering it as a service to customers, which may provide a long-term competitive edge. Describing the significane of the advance, the article closes with this paragraph:

   FHE’s applications may be increasing with time, though. In a data-rich, privacy-poor world, it’s not hard to recognize the appeal of a novel technology that lets people have their secret cake and eat it too.

   -h/t RealClear Science
   

   ---

   In order to help bring Steem's content to a new audience, if you think this post was informative, please consider sharing it through your other social media accounts.

   ---

   And to help make Steem the go-to place for timely information on diverse topics, I invite you to discuss any of these links in the comments and/or your own response post.

   Beneficiaries

   • Experimental - penny4thoughts - 36%
   • Public Steem API - justyy - 5%
   • Public Steem API - steemit - 5%
   • SteemWorld.Org Tools for Steemizens - steemchiller - 5%

   ---

   About this series

   ---

   Sharing a link does not imply endorsement or agreement, and I receive no incentives for sharing from any of the content creators.
   
   

   Follow on steem: remlaps-lite, remlaps

   ---