the age of internet optimism is long over. it's almost hard to believe now that the quaint techno-utopians of the nineties who believed the internet would bring about a more decentralized and democratic societyadam curtis voice "but this was a fantasy" even existed, in the wake of the more-or-less complete corporate takeover of the internet. a good portion of the internet that didn't just devolve into spam and advertising is now devoted to people complaining about the internet. the media is constantly running articles about how social media is making teens depressed or turning everyone into an extremist or spreading "dangerous misinformation". self-help gurus are now writing books about how to quit social media and reclaim your life. but despite the apparent consensus, most people seem consigned to wallowing in their online misery forever, content with blowing off steam by making the occasional ironic comment about how the internet sucks and they're suffering but they are still "stuck here forever". the last thing anyone expects is for it all to just collapse one day. to borrow the phrasing from mark fisher, "it's easier to imagine the end of the world than the end of the internet".
enter maverick writer and perennial pessimist sam kriss, who kicked off his new blog with a dramatic proclamation: "The internet is already over". beneath some of his more extravagant prose (you must forgive a prophet his theatrics), i found his fundamental arguments persuasive to some extent, although i'm not sure i would agree that the internet will vanish completely, only change form.
probably the most... errr... grounded argument kriss makes is from the business perspective. the corporate internet, at least the parts that actually make money, is deeply reliant on a single golden goose: advertising. it's now a 500 billion dollar per year market, and "...provides 80% of Google's income and 99% of Facebook's". things are good for google and facebook because they're so dominant and the market is enormous, but what if something were to happen to online advertising... like companies finding out that it doesn't work.
online advertising's killer product has always been so-called "targeted ads", the thing internet companies greedily collect all your personal browsing data for, so they can beam you ultra-specific ads that you can't help but click, compelled by some hypnotic power. if it worked, it would be quite valuable indeed. but this was a fantasy.
on this topic, our old friend adam curtis from the sidenote above has this to say:
The truth is that people within the advertising and marketing industry are extremely suspicious about whether online advertising has any effect at all. The internet has been captured by four giant corporations who don't produce anything, contribute nothing to the wealth of the country, and hoard their billions of dollars in order to pounce on anything that appears to be a competitor and buy it out immediately. They will get you and me to do the work for them – which is putting the data in – then they send out what they con other people into believing are targeted ads. But actually, the problem with their advertising is that it is – like all geek stuff – literal. It has no imagination to it whatsoever. It sees that you bought a ticket to Budapest, so you're going to get more tickets to Budapest. It's a scam.
but you can't just trust what any old british guy with an authoritative voice says in an internet interview, which is why i decided to take a page from curtis' booknot literally, here's what curtis says later in the interview:
I know everyone says there's a massive amount of information on the internet but actually, if you analyse it, there is very little. It tends to be the same thing repeated over and over again. This is something to do with Google rankings; it's also something to do with the speed with which journalism must be done. So what you have to do is go and read books. Read really boring, old books. and find a book on this topic. turns out there is one, titled Subprime Attention Crisis by tim hwang. the title of the book alludes to the subprime mortgage crisis that kicked off the global recession in 2008, because the hwang's main argument is that the same kind of process may sink online advertising as well.
the key factor that caused the subprime mortgage crisis was that risky debt (in the form of huge mortgages that the borrowers were likely to default on) was packaged up and obfuscated under so many layers of financialization (using confusing new financial instruments with complicated names like CDO, "collateralized debt obligation") that big investors like banks bought tons of them without knowing how risky they actually were. they blindly trusted the credit rating agencies who said they were relatively safe investments, because untangling the chain of financial abstraction to find out what the main underlying assets even were was too difficult.
meanwhile, hwang sees the same sort of financialization and obfuscation going on in the online advertising market, except a lot of it is driven by the usage of algorithms and automatic bids in lightning-fast ad auctions, happening so fast and automated that no one is quite sure what exactly is being bought. on top of this is a similar problem with commodification, which requires standardization and classification of goods that may be difficult to standardize or classify, like mortgages or attention. how do you define a risky vs. a non-risky mortgage? even if you manage that, what if you run into malicious debt salesman gaming the system to get risky mortgages classified as non-risky ones? the same thing happens in online advertising, how do you measure attention? what is one unit of attention? is some attention more valuable than other attention? and of course, the fact that measurement of attention is particularly nebulous makes it easy to sell attention for more than it's worthsomewhat related: "One analyst noted that Facebook claimed to be able to reach 25 million more eighteen- to thirty-four-year-olds in the United States than should exist according to the U.S. census.".
despite the opacity of the market, studies have been run to try and determine the efficacy of online advertising, since it's so important to many stakeholders. what the studies have found is far from reassuring, for advertisers at least. a selection from the book:
One data set drawn from Google's ad network suggests that the average click-through rate for a comparable display ad in 2018 was 0.46 percent.by comparison, the first banner ads in 1994 had a click-through rate of 44%. can you even imagine?
Once, the world was not as it has since become. Once it worked in a way different from the way it works now; its very flesh and bones, the physical laws that governed it, were ever so slightly different from the ones we know.
And now, overused by the greedy, the magic has left and online advertising no longer works.On mobile devices, close to 50 percent of all click-throughs are not users signaling interest in an advertisement, but instead accidental "fat finger" clicks—users unintentionally clicking on content while using a touch-screen device.
In 2009, one study estimated that 8 percent of internet users were responsible for 85 percent of all advertisement click-throughs online.
In 2013, a controlled experiment on more than a million customers to evaluate the causal effect of online ads concluded that a customer "between ages 20 and 40 experienced little or no effect from the advertising." This was in spite of this demographic's proportionally heavier usage of the internet. In contrast, the study found that customers older than sixty-five, despite constituting only 5 percent of the experimental group, were responsible for 40 percent of the total effect observed as a result of the advertising.
In 2014, Google released a report suggesting that 56.1 percent of all ads displayed on the internet are never seen by a human.
One study by Deloitte from 2017 suggests that fully three-quarters of North Americans engage in "at least one form of regular ad blocking."
A study conducted by Adobe in 2018 concluded that about 28 percent of website traffic showed "non-human signals," indicating that it originated in automated scripts or in click farms.
In 2017, Forrester Research concluded that the previous year, as much as 56 percent of all ad dollars spent on display advertising were lost to fraudulent or otherwise unviewable inventory.
is this enough to make marketers "suspicious"? apparently so, because as it says in one of the articles kriss links to, some of your big corporate ad spenders (the kind that make you wonder why do they even bother advertising, everyone already knows who they are) tested the waters: "P&G; turned off $200 million, Chase slashed ad reach by 99%, Uber turned off $120 million in ads for app installs". the result? "not one business noticed a change in outcomes."
facebook seems to have seen the writing on the wall. google is somewhat shielded by better diversification and dominance in the more robust "search ad" market, but as mentioned before, 99% of facebook's revenue is from digital ads, mostly targeted. hence, the recent desperate rebrand into "meta platforms inc." and the flailing forays into the "metaverse", which despite all their money and developers is still somehow barely distinguishable from a unity game made with free assets. meanwhile, amid a drop in users for the first time ever and warnings about advertisers cutting spending, investors are pulling out and facebook (excuse me, META) stock is down well over 50% this year so far.
the same in fact goes for google and many other companies, who were riding highs from covid lockdowns that ultimately ended up proving that even when people were shut in all day with nothing to do, they weren't engaging that much more with social media or digital ads. the issue is, if the internet/tech hype dies down, it could bring down a lot more than just your digital advertisers. the dirty little secret of the industry is that the tech companies that aren't making mountains of cash selling digital ads right now are... not making money at all. twitter has been kind of a big deal for a while now, and they've only turned a profit in the years 2018 and 2019. snapchat has never turned a profit. uber might be one of the worst, straight from wikipedia: "Uber has posted hundreds of millions or billions of dollars in losses each year since 2014 except for 2018, when it exited from the markets in Russia, China, and Southeast Asia in exchange for stakes in rival businesses."uber lost 9.14 billion in 2022 alone. i joke that at this point it's more or less provided as a public service to the world by billionaire venture capitalists. it's probably the most successful unsuccessful company of all time, but in the bizarro world of modern business and finance it has somehow still made many people fabulously wealthy anyway. same goes for lyft, grubhub, and doordash. but these are only the companies whose finances we know about because they're publicly traded. who knows when it comes to the many that are subsidiaries (linkedin, twitch) or still privately held (reddit, discord)?
how do these companies possibly stay in business when it seems half of what they do is just light piles of money on fire? the answer is that investors and financiers keep giving them money to burn, in the form of venture capital and cheap debt. a lot of this is a consequence of the 2008 recession, which led to over a decade of low interest rates (essentially zero from 2008 to 2015) and thus favorable loans. low interest rates also meant, however, that investors with a lot of money had trouble finding worthwhile places to park it, since the yields on many other investment classes like bonds and treasuries dropped as well, as fundamentally they are types of interest-yielding debt. you can't just leave money sitting around because it'll lose value due to inflation, so investors poured money into the one thing that still got good returns: buying stock (equity) in companies. this is where "venture capital" comes from, investors with lots of money going "unicorn hunting" in silicon valley, trying to find early-stage privately-held companies raising money by selling stakes in their companies before they go public. the hope is that down the line, the companies will become successful and the investors will make their money back and more, especially if they got in earlyinterestingly, a company doesn't ever have to become profitable for an investor to exit out with a tidy profit, as long as they find a bigger chump to sell their stock in it to for more than they paid. even though uber has never made profit, most of its early investors and founders/employees have done very well for themselves offloading uber stock on the stock exchanges after uber went public, since they got their shares for much less than what they are now selling for on the stock market.. the only issue is that successful companies are few and far between, however when you win you win BIG (returns of several thousand percent on your investments), so it becomes a numbers game: venture capitalists throw loads of money at dozens of companies, and the one big one out of those more than makes up for all the rest being failures.
as such, as long as unprofitable tech companies are able to keep up the charade that they will one day become profitable, the investment dollars will keep flowing. the most important metric is generally growth in the number of users, measured in a lot of ways like "daily active users". it doesn't matter how effective you are at actually extracting money from those users, as long as the number of users continues growing, investors will be happy to continue throwing money at you. so far, this has been easy for a lot of the big social media companies to keep up thanks to network effects, the fact that there are a lot of people in the world, and of course almost all of them are free to use (sometimes they also just fudge the numbers by including bots as users, as we are now finding out with twitter). meanwhile, tech companies like uber and doordash are able to continue gaining users because they offer their services at a loss. this is why uber loses hundreds of millions of dollars every year, they do not charge the true cost of delivering their service, but as long as the cheap prices continue enticing new users to join and old users to stay, the venture capital money will keep coming in to cover the losses.
when the userbase stops growing, though, that's when the whole scheme collapses. the fact of the matter is, the current tech/internet/social media company business model may simply be long-term unviable because there's no way to effectively monetize it outside of digital advertising, and as we saw earlier, the digital advertising gravy train is probably not going to last forever. what happens when the userbases stop growing? the money from venture capitalists and investors will stop flowing. when facebook's "daily active users" fell for the first time ever in 2021, from 1.930 billion to 1.929 billion, its share price got pummelled, dropping 30% in a single day. if this is what happens to facebook, which measures its user counts in billions and is currently massively profitable, what will happen when the user count falls for one of the countless tech companies that's been hemorrhaging money? to make the situation even worse, interest rates are now going up as well to curb inflation, so there won't be any cheap loans to plug the gaps anymore and investors may start looking at other investments like bonds again now that the yields are getting better.
the only option for those companies will be to try and become profitable before the money runs out, but the problem is that it may not even be possible. this essay started out by recognizing the massive negative vibe shift towards social media and the modern internet (the "gig economy" is also taking a lot of heat lately), however many people still feel trapped in it. perhaps all they need to quit is a little nudge, like some intolerable ads or fees or other changes that desperate companies make in the scramble to monetize their previously-heavily-subsidized services.
if you're young like me, i wouldn't blame you for thinking the current internet may last forever. after all, it's basically been heading the same direction towards corporatization and social media walled gardens for as long as you can remember. but in the grand scheme of things, all those companies and the internet itself are still quite young, younger than many people alive today. it still could easily be, as sam kriss proposes in his post, nothing but an extended fad. the rise was quick, but the collapse may be even quicker as the exodus begins and the dominoes cascade. the internet may feel like the world to many now, but even the world can change almost overnight: in 1989, everyone still thought the soviet union would last forever. the one thing you hope, though, is that it doesn't all get replaced with something worse.
note: i originally promised some thoughts about the future of the internet at the end of this essay so here they are now. the above essay was written prior to important recent developments like the release of chatgpt and elon musk’s acquisition of twitter, which have made me feel a little more confident about speculating now. maybe i should split all this into its own post because it got kind of long.
the great hurdle for social media moving forward into an era with more emphasis on profitability is that ads may not be enough to do the trick, but people are vehemently opposed to paying anything for social media. this is despite the fact that many use it for hours a day, are essentially addicts/dependent on it, and many even get significant value or benefits out of it. people literally build whole businesses and brands off of marketing themselves for free by making good social media posts, or get hired for jobs. even people like prolific poster “wint (@dril)” on twitter, who has basically built a modest career off the back of his short-form tweet comedy, won’t even shell out fifteen dollars a month for it. it’s hard to tell if this is simply something that’s been normalized due to over a decade of free social media, or if it’s some deeply entrenched psychological barrier. it could be that all it takes is one company to successfully charge for social media, and then all the dominoes fall into place making that the new normal.
the recent meteoric rise of tiktok is a fascinating case. it represents a significant departure from the social media paradigm of the past, in fact i even think that calling it social media may be a misnomer. there is comparatively little actual socialization or construction of a personal social graph through “following” and “followers”. instead the focus and the emphasis is purely on the content, curated not by anyone you know or follow but by the individualized and thus highly-effective algorithm. pure entertainment creation and consumption. everyone gets sorted into their own bubble. this has allowed tiktok to avoid to some extent the pitfalls that make people dislike other social media, like petty flamewars and the like. in fact it seems to have far and away the highest user satisfaction of any social media platform, practically unprecedented and yet another reason it stands out. the question of eventual profitability still remains, one obstacle being that apparently short video ads generate significantly less advertising revenue than other types. but its popularity among users means that potentially tiktok could be the first to pull off a subscription-based social media model – it has already been proved that people will pay $15 a month for access to entertainment services like streaming platforms, so why not extend that to short video as well?
llm/ai like gpt presents both the problem and the solution with regards to widespread spam on the internet. one of the main use cases of the internet, consulting its vast stores of human knowledge through searches, is rapidly becoming unfeasible due to mountains of SEO spam. openai inadvertently solved this issue by condensing all the information on the internet into one dataset that can be easily and effectively queried using natural language by pretty much anyone, with little risk of spam or scam in the output. obviously this does not bode well for google, a flailing calcified behemoth of a company, eminently dependent for decades on a single extremely productive golden goose that is now showing the early symptoms of terminal disease, and the business sense they don’t have because they never needed to develop it allowed them to be easily outmaneuvered by savvy microsoft in laying a claim on openai.
twitter is in a particularly unique position among social media, which as you may guess is because it’s now owned by elon musk. this means that it’s particularly nimble and capable of undergoing massive changes on a whim if mr. musk wills it, since he’s more or less the king of twitter. unlike most corporate ceos, he doesn’t have to answer to countless committees or bureaucrats, to a board of directors, to large investors and shareholders, all because he is in essence the sole owner of twitter. he only needs to answer to his creditors, the banks that partially helped finance the deal, and they will be satisfied as long as they get paid. in fact owing several billion dollars to the banks actually gives elon the upper hand over them as well, according to the old adage “if you owe the bank $100 that’s your problem; if you owe the bank $100 million, that’s the bank’s problem”.
this all means that twitter is now a grand experiment in the future of social media, no longer beholden to many of the forces like excessive “stakeholders” and bureaucracy that have strangled innovation at other large tech companies. so far, though, the results haven’t been very promising although to be fair, not much has changed. elon musk clearly has some big ideas and talks a big game about twitter becoming the “arbiter of truth” or “the everything app”, but so far the actual changes seem either petty (all the blue check stuff) or misguided (allowing megatweets with thousands of characters). the only big thing that happened was 80% of twitter employees getting fired or quitting without the website going down in flames like some people prophesied, which seems to reveal that one potential path to tech profitability could be by cutting expensive tech employee costs, since evidently headcounts must be seriously bloated if you can let go of that many people with barely any effect on the site.
however, it's hard to deny musk's track record as a miracle-maker who succeeds despite long odds. nobody believed in tesla or spacex for years, and look at where they're at now. it's a completely different market from building cars and spaceships and being one of the richest and most famous people in the world means he no longer has the strong underdog energy that he had while building tesla or spacex fifteen years ago, personally throwing his last several million dollars in while teetering on the edge of backruptcy. he may just pull something off yet again. i wouldn't neccessarily bet for him, but bet against him at your own peril.
apple – through some mystifying marketing magic, they have managed to become enormously popular while still retaining prestige, an oxymoronic superposition previously thought to be untenable. the extensive apple “ecosystem” of hardware and software is essentially a walled garden and represents perhaps the most valuable tract of real estate in the world, thanks to the devout loyalty to the brand and the premium prices residents are willing and able to pay. apple retains exclusive, authoritarian control over this prime real estate which gives it pretty much the most unassailable moat in not just technology, but business as a whole. investors recognize this and that’s why it is the world’s most valuable company by market capitalization.
in the past, apple has profited greatly by gatekeeping access to the walled garden, for example by taking a huge cut from every app on the app store. however, i expect that going forward, we may see some more rent-seeking behavior on the part of apple, trying to extract additional value from the loyal customers they’ve captured. this could go a number of ways – selling additional services, charging for things that used to be free, or even the classic method of selling ads. the gatekeeping will get fiercer as well, something that was already seen last year when apple made some privacy changes that kneecapped facebook’s data collection ability and sent its share price tumbling. apple makes the dubious claim that this was to protect users since privacy has been all the rage, but the way i see it, they’re just trying to keep all that data for themselves to monetize. maybe in the future they’ll even let facebook back in on it, just for a hefty price.
apple has also made some gestures towards getting into AR/VR recently, which to date has been the graveyard of tech empires: microsoft, google, facebook, and valve have all tried and failed to make it a Thing in the past couple years. but none of those companies are as sexy as apple, or have the stellar hardware innovation track record that apple does, so if anyone can do it, it’s apple. i just worry that it may be a struggle without jony ive at the helm anymore. it’s also possible that they’ve lost the design touch, since it’s now been nearly seven years since their last major innovative new hardware product (airpods, released 2016).