author: niplav, created: 2022-02-06, modified: 2023-01-17, language: english, status: notes, importance: 6, confidence: unlikely
The optimization power of attention-grabbing systems available over the internet is growing faster than any normal human's ability to defend against such attention grabbing, and seems likely to continue to grow. If one values ones own time, it therefore seems worthwhile to limit ones access to the internet, more so than people usually do.
New religions and mystery cults explode across the planet; much of the Net is unusable, flattened by successive semiotic jihads.
— Charles Stross, “Accelerando” p. 169, 2005
Many people have noticed that they are using the internet and online services (e.g. through apps) more than they would ideally like, in a non-optimal manner, with effects on work productivity, reducing selective sustained attention spans, and with a large opportunity cost.
The basic argument can be summarized as follows: There is an economic incentive to get people addicted to things. Currently, information technology appears to be the fastest advancing sector of the economy, and is therefore the most potent way of getting people addicted.
If one wants to avoid falling into addictive patterns, then, it seems advisable to be very cautious around information technology, and in the more paranoid case, one wants to stay out of the addictive "event horizon": the state at which it has become basically impossible to unplug—which means that one ought to prepare to leave the internet.
The world economy is currently growing at ~3% a
year,
which implies that during the average
lifetime
of a human, it grows $1.03^{70} \approx 9$-fold (and
that is before considering possible changes in the growth
mode
from exponential to hyperbolic
growth).
While human minds are extraordinarily adaptive (surviving as hunter-gatherers, farmers, industrial workers, scientists, in space, etc.), they appear to have some hard-coded characteristics that are not easily changed during a human's lifetime (or even by evolution—counterexamples include the evolution of lactase persistence, alcohol dehydrogenase and sickle cell anemia). Humanities' ability to create technology outstrips each individual's ability to understand that technology—the Commodore 64 is said to be the last computer that can be fully understood by one person in its totality, and AI systems are becoming so complex that explainable AI and interpretability research are becoming their own fields.
While economic growth has been on the whole a delightfully advantageous process, this makes individual humans more prone to manipulation, deception, and addiction to these kinds of processes.
This would be no problem if human brains were the OpenBSD or seL4 of cognitive systems, but alas—we were created by evolution as spaghetti towers, and until we change something about that, our minds will remain unstructured heaps of pasta.
Not only do we bring up an impressive barrage of cognitive biases, but it seems likely that our brains are designed to deceive and hide information from themselves. And knowing about cognitive biases and flawed thinking usually helps little with actually mitigating them, sometimes even amplifying the problem.
Coherent, comprehensive theories of cognitive bias mitigation are lacking.
— English Wikipedia, “Cognitive Bias Mitigation”, 2021
This applies even more strongly to addictive tendencies and emotional outbursts.
At the moment, technology is just moving more quickly than evolution or cultural evolution (which is why the tricks of pickup artists continue to work: the defenses against them operate on the timescales of cultural or genetic evolution, and the pickup artists also usually put in more optimization work than the women they approach).
Similarly, companies usually put in much more effort to manipulate their customers than the customers put into defending against that manipulation.
Producers of a good have an incentive to modify their (potential) customers' preferences so that those customers want to consume more of their product. Examples for this are not hard to find, with perhaps the most striking one being the opioid epidemic, especially in the United States, but other examples include gambling, nicotine, gustatorily appealing fast food, and the complete practice of advertising.
As long as companies can legally get away with it, they have an incentive to modify the preferences of their (potential) customers as to desire more of the product of that company.
He's the guy who patented using genetic algorithms to patent everything they can permutate from an initial description of a problem domain – not just a better mousetrap, but the set of all possible better mousetraps. Roughly a third of his inventions are legal, a third are illegal, and the remainder are legal but will become illegal as soon as the legislatosaurus wakes up, smells the coffee, and panics.
— Charles Stross, “Accelerando” p. 11, 2005
Existing legislative bodies move too slowly and don't understand the problem well enough to do much about it, and often the resulting polarization & outrage is in their interest.
Why, then, focus on addiction to information specifically? If the whole economy is growing faster than any human can defend, and companies have an incentive to get me addicted, shouldn't I worry about everything, and retreat into a bunker while the world around me falls into a spiral of self-reinforcing behavioral loops?
Well, yes, maybe! But, barring complete independence from the rest of society, there are reasons to focus on addictive information systems in particular.
The current setup of the internet (at least in western countries) allows most types of information exchange to occur freely, without obstruction, and nearly instantaneously from any part of the planet to any other. The same can not be said for physical objects, making both regulation easier and feedback loops slower. Thus, if the provider of an addictive service is regulated in a country, it can often change servers and appear on the web again in a short time (as shown by the difficulty of shutting down The Pirate Bay, sci-hub, or Library Genesis (this method is not entirely foolproof, as shown by the censorship of TikTok, especially in India)).
While there is much ado about privacy, for example the right to be forgotten or the General Data Protection Regulation, humane technology isn't very much on the radar of legislators yet—a summary on Wikipedia lists only some minor actions by east-Asian countries (again of their western counterparts) and some attempts of large technology companies at self-regulation.
A new Nielsen Company audience report reveals that adults in the United States devoted about 10 hours and 39 minutes each day to consuming media during the first quarter of this year. […] The report reveals a dramatic one-hour increase over last year in how often the average American adult gorges on media in a day. During the same time period last year, Nielsen reported that people spent about nine hours and 39 minutes engaging with gadgets.
—Jacqueline Howard, “Americans devote more than 10 hours a day to screen time, and growing”, 2016
Children as young as two to four years old already average two and a half hours of screen time a day, and it soon climbs to an average of three hours and five minutes for ages five to eight. While most screen time is TV and YouTube viewing, children ages five to eight spend forty minutes a day gaming.³⁸ We don't yet know what young children's attention spans are when watching, say, YouTube, but we do know from laboratory studies that young children are mor susceptible to distraction than older children, and when distracted, it takes them longer to regain their focus on the original object.
—Gloria Mark, “Attention Span” p. 83, 2023
Moved here.
So, dear reader, you might ask what I now want from you. I'm glad you asked! (Or at least my model of you very conveniently did so).
If you are reading this, it's likely you have already taken some measures to prevent you from wasting too much time on the internet or internet-connected services. If you haven't, you reading this text this far is evidence that you should.
My advice is in its strongest form quite drastic: Leave the internet and everyone connected to it behind, acquire a library (be it physical or digital) and live in a secluded hut by a remote lake, lest you spend the rest of your life glued in front of a screen fuming at how other peoples' opinions can be so wrong, or just mindlessly scrolling while the hours (and days, and weeks…) trickle by…
People are going to make their restrictions around blockers/grayscreening etc. stricter and stricter.
As an example, here's the things I have downloaded for offline reading:
To download a website, I usually used the command wget:
wget --page-requisites -e robots=off -p -k -c -r --mirror -p --html-extension -R '*replytocom*,*like_comment*,*p=*,*test-preview=*,*test-expand=*,*url=*,*share=*,*commentId=*,*sortedBy=*,*after=*,*before=*,*limit=*,*filter=*,*showPostCount=*,*useTagName=*,*///*,*social-media/*,*updated-max=*,*max-results=*,*reverse-paginate=*,*showComment=*,*filters_and=*,*following%5C*,*https:%5C*,*fsnapshot*' -P . www.example.com
This excludes downloading the same page for every comment with wordpress
sites, and different versions of the allPosts pages for the EA forum and
LessWrong. Unfortunately, both the EA forum and LessWrong are not very
amenable to being downloaded this way, and I will have to experiment
with other approaches. It also sadly can't distinguish between resources
(like images) hosted on other sites and other websites, and so those are
missing (if -H is turned on, it begins downloading the entire internet).
To download a webpage, I use this command:
wget -H -e robots=off -p -k www.example.com/resource.html
Optionally: Separate work with internet, work without internet, fun with internet, fun without internet spaces
Started 2021-05-21.
I definitely have (had?) a problem with wasting
time on the internet, barely consciously scrolling
reddit/twitter/lesswrong/4chan/…. Mercifully, I don't have exact
numbers, but I would estimate that since I started using the internet
~10 years ago, I have spent ≥~1½h a day on it unproductively, giving
me at least $10*365*1.5h=5475h$ wasted. Had I poured half of that
into honing an existing skill, or acquiring a new one… But what is
important is to avoid wasting another 5000h, instead of crying about
the ones that had already fallen into a screen.
AI systems become capable enough to sap out most of human attention/ability to work, this prevents more capable AI systems from being built, but also destroys any options for progress.
This might end up in an equilibrium.