In machine learning and computer vision, the default is to put your work on ArXiv immediately upon completion. There is a lot of openly available research depending on the field. I find most fields in computer science are good for this.
In the case of the Mario example, I very much doubt that the author was not able to find other work because of closed journals.
Research can seem non-transparent to non-researchers because when problems are new, they are often poorly understood. Academic papers discuss novel problems and contextualize them based on other cutting edge work. Reading and understanding these papers requires a lot of context and takes time. Research is a skill that takes years to learn.
After some time has passed and we collectively gain a better understanding of a problem, academic papers may seem abstruse and overly complicated, but when these works were first published, this was the best way to understand them. For somebody looking for a recipe solution to a problem, an academic paper is likely not an the ideal place to look, which is why we write books, blog posts, etc... as we come to better understand a problem and its solutions.
Discoverability is hard. I mean, forget academic papers, take a look at just the web, and how many years it took Google to create a search engine that made discoverability easy, and that exploited (originally) the graph structure of the web. Papers do cite other papers, but how do you check a paper's relevance? Moreover, papers often describe something very specific, a specific approach, and searching for that approach is a lot harder than looking for a page that is "Haskell tutorial".
An open journal isn't the answer here (it is the answer for other things though imo) and I'm not sure why you're getting so angry.
You also can't make use of data if it doesn't have appropriate metadata indicating how it can be used. Metadata is more than just "good"- it's essential for actually being able to run an analysis. Take fMRI for example- someone could just throw a brain image up on the internet. In fact, one particularly uninformed individual (who possibly broke Virginia state law) did so here: https://github.com/dcunited001/mri-scans
However, the raw data is not useful without knowing what scan sequence was used, slice-timing order, etc. Some of this is encoded in the data itself (the example above uploaded DICOM, which has somewhat rich metadata, although by no means is it comprehensive enough), but other data that is necessary is not (Phillips in particular is somewhat infamous for doing non-standard things with their scanners that aren't recorded in the DICOM headers, and in general it's a Herculean task getting MRI vendors to agree on standards). You could try performing an analysis with contextless data, but you're likely just going to make really spotty conclusions.
> I'm not in academia
Why should that matter? The circumstances of data collection are also important for business intelligence data or financial modeling data. I'm certain that the lack of this also contributes to poorly formed conclusions that cause real hurt to businesses out there.
In any semi-decent peer reviewed venue, you would cite a wide variety of papers that solve a similar or related problem or introduce a concept related to the method in the paper.
The related work section of a paper is one of the most important parts since it puts the research into context. By citing other work, the authors explain what has already been done, and what contribution their work makes.
A related work section should also illustrate the downsides, limitations, and differences of other cited research. Limitations of other works are often poorly understood since very few people have had the time to evaluate them beyond the initial experiments done before first publication.
Research is not simply about presenting new techniques but also understanding the trade-offs that arise when choosing different solutions to a problem.
This of course comes up during the peer review process as well. The referee informs you of a paper that is tangentially relevant, but you couldn't find that during your literature search because it is paywalled. How relevant was it actually to the work you performed or to people reading your paper if it is not readily possible for you (and possibly your readers) to access the paper?
Citations in academia are often times more the currency of the trade than actual scholarship. There is a reason review articles exist, not every paper needs to be a review article in its own right.
Reproducibility is important but a completely different topic and I don't know why you're bringing it up here.
Note that I don't say that SethBling is wrong. Seen as tinkering, this is perfectly fine, and he did a great job with the video. No hard feelings. I just use it as an example of how it is not research, because of the lack of scholarship.
Mayhaps, you need to do a bit more "youtubing" to get your stuff out there and more accessible. A couple of my current favourites on youtube is 3Blue1Brown and Mathologer. They are both entertaining as well as being informative and have set me on investigatory paths for other research in various fields.
And fuck that. If that's the attitude that academia is going to take, clearly we need to kill it off, tinker for a while, and start fresh.
Maybe we can refresh some standards for academic writing while we're at it.
The number of pretentious title-holders (even at the lowest minimum levels) I've met seems unending. I could get into that one... but I won't.
Instead you've just reminded me of yet more subjects to stack on my reading list.
>Last year, some guy made an experiment with evolving neural networks for Super Mario Bros and made a YouTube video out of it. The video certainly reached more people on the internet than my work did; it makes no mention of any previous work. Seen as tinkering, that work and video is good work; seen as research, it is atrocious because of the complete lack of scholarship. The guy didn't even know he was reinventing the wheel, and didn't care to look it up.
Then publish your damn results in a damn open journal, dammit.