Some bits about science

The Scientific Paper Is Obsolete. Here’s What’s Next

It’ll be some time before computational notebooks replace PDFs in scientific journals, because that would mean changing the incentive structure of science itself. Until journals require scientists to submit notebooks, and until sharing your work and your data becomes the way to earn prestige, or funding, people will likely just keep doing what they’re doing.

It is incredibly depressing that we live in a world where scientific knowledge is still shared mostly by means of PDF documents, but the title of this article is misleading at best. The future of science communication will not be built on yet another proprietary document format. On the other hand, the Web platform has all the technical capabilities needed to create any sort of “computational” papers, but it still lacks appropriate authoring tools to empower scientists to do it by themselves.
The reports of the scientific paper’s death have been (unfortunately) greatly exaggerated.

via theatlantic.com
April 17, 2018

Complexity Bias: Why We Prefer Complicated to Simple

Complexity Bias

Complexity bias is a logical fallacy that leads us to give undue credence to complex concepts.
Faced with two competing hypotheses, we are likely to choose the most complex one. That’s usually the option with the most assumptions and regressions. As a result, when we need to solve a problem, we may ignore simple solutions — thinking “that will never work” — and instead favor complex ones.
To understand complexity bias, we need first to establish the meaning of three key terms associated with it: complexity, simplicity, and chaos.

Nice piece on the risks of being seduced by unnecessary complexity, especially in the broader context of language. It reminded me of an old essay by Italo Calvino, “L’antilingua”—literally: “the anti-language”—in which he comically shows the effects of replacing simple words with increasingly grotesque jargon. To paraphrase Calvino, the anti-language is the language of people who prefer saying “utilize” instead of “use”, people who are scared of showing familiarity with the subject of their talk. According to him, speaking the anti-language is a sign of being out of touch with life, and ultimately represents the death of language itself.

via fs.blog
April 5, 2018

The impossibility of intelligence explosion

The basic premise is that, in the near future, a first “seed AI” will be created, with general problem-solving abilities slightly surpassing that of humans. This seed AI would start designing better AIs, initiating a recursive self-improvement loop that would immediately leave human intelligence in the dust, overtaking it by orders of magnitude in a short time. Proponents of this theory also regard intelligence as a kind of superpower, conferring its holders with almost supernatural capabilities to shape their environment (…)

This science-fiction narrative contributes to the dangerously misleading public debate that is ongoing about the risks of AI and the need for AI regulation. In this post, I argue that intelligence explosion is impossible — that the notion of intelligence explosion comes from a profound misunderstanding of both the nature of intelligence and the behavior of recursively self-augmenting systems.

Exhaustive post by Françoise Chollet (author of Keras, currently working on Deep Learning at Google) confuting the dangerous misconceptions that fuel the AI debate.

via medium.com
December 10, 2017

The world is poorly designed. But copying nature helps

Vox and Roman Mars

Nice video from Vox and 99% Invisible’s Roman Mars about biomimicry.

It reminds me of Bruno Munari’s analytical study of plants and fruits in his must read Design as Art, in which he meticulously describes and praises the essential features of natural objects as a source of inspiration:

This object [an orange, an almost perfect object where shape, function and use display total consistency] is made up of a series of modular containers shaped very much like the segments of an orange arranged in a circle around the vertical axis. Each container or section has its straight side flush with the axis and its curved side turned outwards. In this way the sum of their curved sides forms a globe, a rough sphere.

via youtube.com
November 10, 2017

Music was better back then: When do we stop keeping up with popular music?

Mainstream artists are at the center of a circle, with each larger concentric ring representing artists of decreasing popularity. The average U.S. teen is very close to the center of the chart — that is, they’re almost exclusively streaming very popular music. Even in the age of media fragmentation, most young listeners start their musical journey among the Billboard 200 before branching out.

And that is exactly what happens next. As users age out of their teens and into their 20s, their path takes them out of the center of the popularity circle. Until their early 30s, mainstream music represents a smaller and smaller proportion of their streaming. And for the average listener, by their mid-30s, their tastes have matured, and they are who they’re going to be.

Two factors drive this transition away from popular music. First, listeners discover less-familiar music genres that they didn’t hear on FM radio as early teens, from artists with a lower popularity rank. Second, listeners are returning to the music that was popular when they were coming of age — but which has since phased out of popularity. Interestingly, this effect is much more pronounced for men than for women

Great analysis and insights on this fascinating phenomenon, which I suspect is more about the lack of desire to discover new things, than it is about popularity itself.

via skynetandebert.com
November 8, 2017

The Myth of a Superhuman AI


The most common misconception about artificial intelligence begins with the common misconception about natural intelligence. This misconception is that intelligence is a single dimension. (…) Intelligence is not a single dimension. It is a complex of many types and modes of cognition, each one a continuum.

The main problem with AI still remains finding a suitable definition of the concept of “intelligence”.

via wired.com
August 16, 2017
View all Bits