Unfortunately most advice tends toward conscious and deliberate rejection of such sources, which continue that assault on attention.
As anyone who's implemented an adblocker or similar online annoyance filters is aware, automating the process of distraction rejection is a vastly more effective and less-attention-costly approach. Rather than individually reject cookies, or cookie notices, or trackers, or ads, or various interstitials / "recommendations" / nags / pop-ups / fly-overs, and the like, I've applied and created sets of tools which remove those without my further conscious awareness. Users of PiHole may occasionally check the dashboard blocking statistics and be amazed at how much not only useless but actively counterproductive crud has been avoided.
Information overload requires cheap, fast, regret-free rejection tools.[1] I've come to suspect that worldviews and models specifically function in this manner, identifying key information which we should focus on, and costlessly discarding the rest.[2] The article does nod to this briefly, particularly in the note referencing Gigerenzer & Gaissmaier, 2011, but on balance misplaces its emphasis, most especially in its suggestions.
Another surprisingly effective tool is randomness, whether in sampling or in decisionmaking directly. Random sampling is a statistical tool for choosing a very small set of observations of interest, where the absolute number is utterly relevant rather than the sampled fraction. That is, a sample size of 30, or 300, or 3,000, is equally effective whether the total population size is ten thousand, ten million, or ten billion.[3]
Within a media environment, in which there are publishers, channels, and distributors, the challenges shift somewhat.
One exceedingly useful tool is noise rejection, or as I put it on Mastodon, "block fuckwits".[4] By blocking those who are misleading, annoying, toxic, or simply cannot or will not engage productively, noise levels fall tremendously. One key feature of the Fediverse is a panopoly of filtering, blocking, and limiting tools, from keyword filters to lists to muting (temporarily or permanently) to blocking (by or against individuals, or entire instances, gradated as silencing or complete exclusion). As a result antisocial behaviour carries sharp consequences.
Timeboxing, and other mechanisms of setting a media budget can also help. Most of use hugely overestimate our message-processing capabilities. Even highly prolific individuals or groups typically have an upper bound of a few hundred incoming messages per day, and perhaps 50 outbound (most of which are exceedingly short --- acknowledgements or yes/no decisions). See Walt Mossberg and Stephen Wolfram, both averaging about 150--300 inbound daily emails. Content moderators seem to have an upper capacity of about 700-800 items/day, which affords about 35--40 seconds of attention to each item, if no other tasks are included at all in an eight-hour workday.[5] Scheduling email and media breaks at specific times of day, perhaps gradated between high-priority (boss, clients) and lower (the rest of the world) contacts is also hugely useful. My email schedule is largely now a few times per year....
A feature lacking from many mainstream tools and platforms is precisely this ability, and especially at the aggregate level. Within, say, YouTube, application stores from Apple and Android, or marketplaces such as Amazon, there is not ability to permit or deny presentations of offerings from a specific channel. Nor is there any way for groups to aggregate their own blocking behaviours. Put another way, providers, merchants, and channels carry no high-cost risk of low-value offerings. If the cost of putting out a crap app, video, book, or merchandise was immediate and permanent blocking by individuals or whole groups, behaviours might improve.[6] Contrast alternatives such as the Debian GNU/Linux package repository where specific sets of policies and behavioural requirements eliminate entire classes of end-user-hostile behaviours. Debian's organisational orientation and premise is focused on its users.[7] Documents aren't enough, the project also has institutional and social conventions for achieving those goals.
Removing apps and disabling notifications is a start. Blocking sites entirely is even more preferable. A fairly recent development has been the emergence of alternate front-ends to major commercial social media platforms which permit directed access whilst avoiding much of the social manipulation, and incidentally, privacy-abusing surveillance, of them. Threadreader and Nitter for Twitter; Invidioius, Piped, yt-dlp, and freestanding media apps (mpv, xine, mplayer, etc.) for YouTube (and a host of other video and audio platforms), Teddit for Reddit, and more.[8]
PiHole or similar adblockers, mentioned above, can be used not only to block advertising, but entire sites and domains exhibiting toxic practices. These blocklists are often collaboratively generated and distributed, which extends the reach and effectiveness on them; they're de facto means for citizens of the Internet to reject the terms and practices imposed by Internet monopolies and publishers. Power law distributions mean that even relatively short lists are surprisingly effective.
Hacker News itself incorporates a similar mechanism through a list of 67k+ banned sites.[9]
Legislation, regulation, class-action lawsuits, and other forms of collective action are the underrepresented remedies lacking from most discussion of this topic. Much as neither Soviet Communism nor Western corporate capitalism finally addressed labour exploitation or environmental concerns, but rather collective unionisation, organising, advocacy, protest, lawsuits, legislation, and regulation did, I think we're facing a similar situation here. Market failures aren't resolved by marketing-it-harder solutions. Boycotts and wallet advocacy have some utility, but are ultimately limited especially where monopolies are absolutely pervasive. That's a slow, long, and expensive battle, but most battles worth fighting are.
________________________________
Notes:
1. A point I've made before: <https://news.ycombinator.com/item?id=35365180> <https://news.ycombinator.com/item?id=35365180> <https://news.ycombinator.com/item?id=22208255> <https://news.ycombinator.com/item?id=31254795>
2. <https://news.ycombinator.com/item?id=22208255> (Listed in note 1 as well.)
3. This concept is central to modern statistics, but hugely misunderstood by the general public. The sample size is sensitive to the size of the effect you're trying to measure, such that if only a very small portion of the entire population is behaving in some manner of interest, you'll need to scale your sampling such that a representitive number of that subgroup are included. Still, small samples can be highly effective. I'd seen this directly when assessing how many Google+ profiles had ever been active, or were recently active. Those numbers turned out to be about 9% and 0.6% of the totals. I'd had a reasonable estimate of the first value after checking only 100 randomly-selected profiles, and within 1,000 or so, of the second. I ultimately checked about 50,0000 profiles (using Google's own sitemaps). A subsequent study increased that sampling tenfold (and ran some more robust checks and processes to ensure effective random sampling), but largely confirmed my own results. See: <https://web.archive.org/web/20150130125653/https://ello.co/d...> and <https://blogs.perficient.com/2015/04/14/real-numbers-for-the...>.
4. <https://toot.cat/@dredmorbius/104371585950783019> I'll note that blocking can be made absolutely without prejudice, as a recognition that another party may be harmed by, or react poorly to, your own posting. Or with varying degrees of prejudice, where appropriate. It's also entirely at the discretion of the blocking party, and grousing about being blocked is virtually always exceedingly poor form, and a validation of the action.
5. Wolfram: <http://quantifiedself.com/wolfram/> Mossberg from an interview I'd at least 7 years ago based on my prior HN references. NYTimes moderation: <http://www.nytimes.com/interactive/2016/09/20/insider/approv...>
6. Why this is hasn't been sufficiently studied to my view, though Cory Doctorow's concepts of enshittification and chokepoint capitalism demonstrate how forcing low-relevance search results on both customers and vendors can, at least in the short term, maximise the profits of the monopolist.
7. See Debian's Philosophy <https://www.debian.org/intro/philosophy>, Social Contract and Free Software Guidelines <https://www.debian.org/social_contract>, Constitution <https://www.debian.org/devel/constitution>, and Policy Manual <https://www.debian.org/doc/debian-policy/>.
8. LibRedirect is a browser extension that automatically ... redirects ... to alternatives for these and other services: <https://libredirect.github.io/>