The STEM scam and its effects on “the future”


Any English major – from those who actually gritted out 4+ years of theory and close reading, to those “lucky” enough to have stemmed their humanities urge by becoming a supposedly in-demand science, technology, engineering, and mathematics (STEM) student instead – has inevitably heard this passive-aggressive question disguised as an honest inquiry. Really, it deserves every pixel of the “mocking Spongebob” meme:

Public contempt for the humanities – and the corrollary of equally public, and thereby visibly embarrassing, adulation for the STEM fields – has been an unmistakable trait of how Americans see education for my entire life. Oh noez, English and Art History don’t teach you “practical” “skills”! We need more chemists and computer programmers!! You know, “skilled” workers!!

The entire STEM/skills discourse – I think I’ll just call it STEMills for short, to try and capture the inherent logic of STEM boosters who envision colleges and universities primarily as sleeker diploma mills, from which students can emerge (having paid dearly along the way) with a sort of certificate of “skills” that translates 1-to-1 into workplace responsiblities – emerged in the shadow of the Great Recession and the ascendant power of Silicon Valley. Rather than confront the reality of Wall Street having destroyed people’s livelihoods and upturned global politics, numreous myths emerged about how:

-There just weren’t enough skilled workers out there; that’s why it was so hard to find a job. (False; if anything, STEM degree holders are in oversupply).
-STEMills would put struggling graduates and more experienced workers ahead of their humanities-focused peers. (False; earnings for humanities major usually catch up by age 40).
-Higher education shouldn’t be wasting its time on “useless” majors such as English. (False; the humanities are enormous money-makers for institutions! They subsidize everything else).

STEMills discourse is propaganda. It posits:

  1. Only “technical” (read: relating to STEMills) knowledge is of any value.
  2. Relatedly, jobs are mostly about applying specific “technical” skills, rather than knowing how to interact with other people (which is what they’re really about; yes, even the STEMmiest job).
  3. Said “technical” domain develops mechanically, or perhaps “naturally” is the better word, outside of politics or culture; we can but acquiesce to what it “wants,” whether that’s faster speeds and feeds in devices or a more convenient experience for customers.
  4. To meet those demands, we need a lot of (cheap) workers. Therefore, higher education is obligated to train students for industry, toward the implicit goal of lowering industry’s labor costs by creating a glut of STEMills graduates.

These are all ideological assumptions, and they’re incoherent. For example, we have the capitalist drive to lower labor costs (more STEMills graduates!!) juxtaposed with the preregoative to destroy humanities programs that are low-cost, high-margin blockbusters for universities.

Adam Kotsko has helped square this circle better than I can:

“Why kill inexpensive programs that often serve as profit-centers while doubling down on the huge capital investments associated with STEM? The answer is that the health of individual institutions is a subordinate concern, at best, for the business interests who set the agenda for U.S. higher education—as evidenced by the fact that so many institutions are perpetually on the verge of failing even as aggregate demand for higher education has never been higher. In addition to the benefits of STEM fields over liberal arts programs already mentioned, the hegemony of STEM creates a situation where colleges and universities are more dependent on major donations, which corporations and wealthy individuals are obviously in the best position to provide.”

To build on Kotkso’s point, the drive to STEMify higher education and society is also a drive to increase inequality. The interests of wealthy donors become preeminent, and that means a dominant ideology under which “good” degrees are mostly ones that serve, rather than question, the goals of capital. Students in the “wrong” fields face the prospect of massive debt with career prospects that have been limited not by any natural law of supply and demand, but by ideological constraints. David Walsh, a postdoc at UVA, has summed up this point well in regard to the history major in particular:

STEMills degrees aren’t inherently more practical than humanities one, nor are they even in every case more profitable, despite the intense ideological war waged against them. There are numerous humanities majors – including yours truly – who have highly remunerative jobs, and not just in fields that map perfectly to what we studied. I acquired all of my knowledge of computer science and electrical engineering outside of formal education, and so did the friend who helped me land a pivotal role a decade ago. Moreover, the writing chops I acquired in a graduate gender studies class ended up transforming how I constructed narratives and used words, vastly improving how I wrote about topics of all kinds – including “technical” ones like software. Knowledge acquisition isn’t always linear; everything you learn in school doesn’t map perfectly to something in the real world, and sometimes the most important capacities are honed somewhere else and under circumstances that didn’t seem important at the time. This very blog, for instnace, which I started mostly as a platform from which to talk about video games, was instrumental in getting me a job at a marketing firm.

But anyway, as Walsh emphasizes, history and the rest of the humanities are now coded as a liberal field eager to question entrenched power, making them the polar opposite of the more politically digestible – for the current ruling class, at least – STEMills fields. Going down the humanities route doesn’t doom one to unemployment and high debt – though it definitley felt that way in 2011 for me! – but it does mean fighting an ideological struggle against every variant of the “But what are you going to do with that?” crowd, from your annoying distant relatives to skeptical employers. What’s worse, the people who complain about the uselessness of non-STEMills knowledge continue to profit immensely from the humanities.

Facebook recently changed its main company name to Meta and has become fixated on the concept of a metaverse, which is kind of hard to explain clearly (and hence why I’m skeptical of its eventual utility). But the metaverse is an idea from fiction – from the pen of lowly geography major Neal Stephenson’s novel “Snow Crash,” in fact. Keanu Reeves had something to say about this recently:

So much of current Silicon Valley/tech ambition is just trying to literalize the work of humanities nerds:

-The metaverse and NFTs are straight out of Stephenson. The ubiquitous Microsoft Teams meetings that a work-related metaverse might eventually entail were predicted by David Foster Wallace’s “Infinite Jest.”

-Mars exploration – one of the most impractical endeavors imaginable – are just extensions of a sci-fi lineage going back to H.G. Wells, a journalist who got his start reading Thomas More and Plato.

-Algorithmic “on this day” memories and Spotify playlists that try to include only “good vibes” music were predicted far in advance by Rana Dasgupta’s “Canterbury Tales”-inspired novel, “Tokyo Cancelled.”

-Space travel and “futurstic” vehicles and devices often turn to Star Trek and Star Wars for inspiration; these properties were not the products of grizzled STEM majors, but of police officer and television writer Gene Roddenberry and film student George Lucas, respectively.

It’s also worth noting that many of thse works were dystopias, especially “Snow Crash” and “Infinite Jest.” So our STEMills overlords are trying to recreate, literally, the worst of all worlds. And they’re doing so by chasing the visions of works of fiction, not following some clear-cut line of scientific progress. Meanwhile, the social conditions that enable the pursuit of technological innovation in the first place are not delivered from on high or determined by some scientific fact of nature – they were built over centuries by people, from ideas drawn from every corner of human knowledge, including – gasp – the humanities. Good luck creating an iPhone, much less a Mars rover, in a society without direct government support, open immigration, and a baseline of social harmony – not to mention money-making, subsidy-providing humanities programs at virtually every college and university.

Eight years (!!) ago, I was already exhausted with the spectacle of STEMills acolytes boasting about the “future” and talking about the worthlessness of the humanities, even as they themselves couldn’t see that same “future” without relying on works of literature:

“By “the future,” commentators usually mean “a reality corresponding to some writer or creative artist’s widely disseminated vision,” which shows the odd poverty of their own imagination as well as the degree to which they often underestimate the power of creative artists/humanities types to drive technological evolution. But can human ingenuity really aspire to nothing more than the realization of a particular flight of fancy? Should we congratulate ourselves for bringing to life the technology from a reality that doesn’t exist?”

So yes, what ever will I or millions of other do with our English degrees – probably helping to prop up the STEMills charade for many more years. It’s such an American notion, of the strong, rugged, self-sufficient individual – or in this instance, the STEMills industrial complex – who in reality is just part of the same humanistic world as everyone else.

Protected: The private archipelago of nostalgia

This content is password protected. To view it please enter your password below:

The limits of software

In 2011, venture capitalist Marc Andreessen wrote a widely discussed essay provocatively called “software is eating the world.” One of the illustrations for the web edition of this piece looked like this:

An Angry Bird

It’s a pixelated version of the main (red) bird in the breakthrough mobile game, “Angry Birds,” one of the so-called “killer apps” of early iOS and Android devices. “Angry Birds” was as good a microcosm of the then-booming app economy as any other piece of mobile software – it was disarmingly easy to use, cheap, and completely unsustainable.

If you want to play the original “Angry Birds” today, the most straightforward option is to play one of the console ports (e.g, for Wii U or Wii), which offer a frozen-in-time look at the game near the height of its popularity in the early 2010s. The game itself, which retailed for a few dollars, can no longer be bought in the Apple App Store and has been replaced instead by a free-to-play sequel called “Angry Birds 2.”

The first “Angry Birds” did well within the distinct limitations of mobile phones, which lack the buttons, analog sticks, and various oddities of consoles and PCs, despite having very powerful CPUs and GPUs. But it was a dead end, too. There wasn’t a lot of room to go further with innovation in control schemes, plus mobile devices were already so capable in the 2010s that there was never an opportunity for a quantum leap in graphical fidelity like the ones that swept through console gaming from roughly 1996-2006, when 3-D graphics and then HD became mainstream.

But despite seeming not to change that dramatically on the surface – how much different, really, is the iPhone 11 from the iPhone 6 you might idly wonder – mobile devices and their games have changed rapidly under the surface, dropping support for older 32-bit apps, adding new APIs, etc. on a yearly basis. Their app ecosystems have also evolved just as fast, from paid software in the early days (Angry Birds was once a one-time purchase) to free-to-play and subscription models.

Despite being only a bit over a decade old, “Angry Birds” seems like a relic now. It couldn’t go outside the limits of rock-bottom mobile app pricing nor the restrictive controls possible on a slab form factor phone, and now ironically the best place to experience it as it was is on one of the consoles that it and the other early mobile games were supposed to make obsolete. The Wii U, which sold a meager 12 million units, has a touch screen controller with a stylus that it basically perfect for playing “Angry Birds” and the fact that that console is discontinued means that there aren’t any of the commercial or technological pressures that will necessitate ongoing updates to its version of the game for it to remain playable. Meanwhile, the iOS version to “Angry Birds” from 2010 might as well not exist anymore.

Andreessen himself recently published a new essay lamenting how no one “builds” things anymore, mainly physical things that would have been helpful to have in hospitals, which after all need much more than software. In other words; he seemed to be moving on from the “software is eating the world” optimism because software is, despite being infinitely malleable, limited in its own ways.

It’s limited in the audiences it can reach, as developers like email client makers HEY recently found out.

It’s limited in its preservation, as “Angry Birds” demonstrates.

It’s limited in how it can be commercialized, as the dominance of free-to-play games shows.

It’s limited by the types of hardware that can be built and shipped in a world beset by climate and public health crises.

Software has limits, just like the earth itself. Andreessen’s pivot since 2011 is a strange but welcome way of someone coming to terms with that notion.

The 2020 Democratic Primary, Part 2: Hillary Clinton Doesn’t Live Here Anymore

The state of Illinois makes it very easy to vote. Rather than stand in line on election day, you can vote absentee, via mail, or early with no excuse, with early voting allowed for two weeks prior. ID is not required, either.

All of this is to say I voted for Hillary Clinton well ahead of Election Day 2016, both because I really wanted to see her win – more so than I did John Kerry in 2004 or Barack Obama in 2008 – and because I was terrified of the prospect of a Donald Trump presidency. I early voted a straight Democratic ticket despite having, only 8 months prior, cast my primary ballot for Bernie Sanders.

(Clinton had won the Illinois primary that March by a narrow margin. Although she had carried Chicago, she was badly outrun elsewhere in the state, to the point that even her landslide margins among the Prairie State’s significant black vote were barely enough for victory. She then carried it easily in the general election, albeit by a lesser margin than Obama had in either of his runs.)

After I early voted, I stopped by at a Popeye’s Chicken on the NW side of Chicago to get lunch and went back home to eat and continue working. When I opened my laptop up, I saw a notification that James Comey had reopened his investigation into Clinton’s email server. It was Friday, Oct. 28, 2016, and Election Day was only 96 hours away. Despite the confidence and civic pride I felt at my early vote, I now dreaded the backlash and the possibility of a loss.

While many elaborate explanations have been offered for what the 2016 election was Really About (trade, immigration, industrial policy, etc.) it was ultimately just about Hillary Clinton.

“Likable Enough”

My contradictory feelings toward Clinton – all-in for her against Trump, very against her in the primary – sort of captured the uniquely polarizing nature of her candidacy – even among Democrats! While the right-wing was never going to cross-over to her in significant numbers, the intraparty tensions in the Democratic Party of 2016 meant that she struggled mightily both to win the nomination and to rally the party’s base in November.

Bernie Sanders, a gadfly who never formally became a Democrat, won almost half the pledged delegates in the 2016 primary despite starting off with no name recognition or money and going up against the most famous woman in the world. That’s crazy! But it wasn’t necessarily an endorsement of his agenda. Most of all, it was an endorsement of him as Not-Clinton.

Sanders won easy victories in the Western states that had bucked Clinton in 2008, while also adding the Rust Belt states of Minnesota, Wisconsin, and Michigan, the Left Coast other than California (where he nevertheless came within 7%), portions of Appalachia, and most of his home turf in New England. These areas all shared large populations of white working class voters, many of whom are likely no longer even Democrats. They went hard for Sanders in 2016, but mostly in protest, it seems.

For example, Sanders won easily in West Virginia, and leftist publications like Jacobin interpreted this as the white working class getting on board with socialism. The notion of an ancestrally Democratic state like WV swinging back over to the party after years of GOP dominance – but only in response to an avowedly socialist campaign – is as nostalgic and romantic to the modern American left as the attachment to outmoded 20th century tactics like canvassing , which I discussed in the first part of this series.

But mostly, it was just conservative and moderate Democrats giving a big thumbs down to Hillary Clinton, whether because of sexism or reservations about her record.

How do we know this, other than the fact that many of the states Sanders carried in 2016 ended up voting GOP against Clinton? Well, without getting to into the weeds, his 2020 campaign has fallen off mightily from its 2016 heights, losing or barely winning many of the states where he ran strongest in 2016:

  • He upset Clinton in Michigan by crushing it in rural areas of the state; this year, he might not even win a single county in the Wolverine State.
  • He won New Hampshire with a bare 26%, the lowest ever in the history of that primary, after getting over 60% in 2016.
  • He will probably have dropped out by the time Wisconsin, one of his most substantial victories in 2016, even votes.

Basically, Bernie 2016 was two things: 1) a strong core of young and very liberal supporters and 2) a massive sink for anti-Clinton votes. He outlasted the other also-rans in the D primary that year because of #1, but it was #2 that took him to the brink of the nomination. The anti-Clinton bloc was so strong that I feel confident saying even Lincoln Chafee, the soft-spoken Republican turned Independent turned Democrat turned Libertarian, whom I once ran into at a Starbucks in Providence, could have racked up a bunch of primary and caucus victories had the race come down to him and Clinton. Barack Obama was on to something when he sarcastically said Clinton was “likable enough” in one of the 2008 debates.

The absence of Hillary Clinton from the 2020 election meant that the Sanders 2020 campaign needed a new blueprint – but it never got one.

What Went Wrong?

An early article about Sanders 2020 by Edward-Isaac Dovere outlined the weird strategy of the campaign, which boiled down to somehow maintaining that strong core of support in a highly fractured field and coming away with a victory despite only getting 30% of the vote. Basically, a plurality victory.

At the same time, the campaign talked about how it needed unprecedented turnout from unlikely voters to win. Taken with the above point about pluralities, these strategies seemed to indicate that Bernie 2020 was scared of the Democratic primary electorate and needed to basically win through a trick – i.e., some combination of there being too many candidates in the race (all of them splitting votes), and a surprising turnout that would lead to Sanders vastly outperforming his polling.

In terms of the fractured field, it sort of worked through the first three contests. Sanders basically got a draw with Pete Buttigieg in Iowa, won a narrow victory over him in New Hampshire, and then won by bigger margin in Nevada despite only getting 33% of the vote overall. But then Joe Biden’s massive South Carolina win led to rapid consolidation. Most of the former Buttigieg and Klobuchar voters went to him, as did those who had flirted with voting for Mike Bloomberg. Once the race became Sanders vs Biden, Bernie was doomed, if only because Biden doesn’t have the unique liabilities that Clinton did.

Turnout was never enough to lead Bernie to the types of sweeping victories his campaign likely imagined. The alarms should have gone off right after Iowa, in this regard. Despite raising $100 million dollars and knocking half the doors in the entire Hawkeye State, the Sanders campaign still lost to Pete Buttigieg, who mostly got earned media coverage and ran just a few TV ads. In New Hampshire, Bernie volunteers knocked a majority of all the doors there but likely would have lost outright had Amy Klobuchar not destroyed Pete Buttigieg at the final debate, lowering his numbers. Combined, those two got over half the vote.

Nevada seemingly went to plan, but Biden actually outran his poll numbers and won black voters by double digits – that, too, should have sounded the alarm that the Sanders campaign was in trouble. Black, moderate and suburban voters were waiting to consolidate behind a not-Bernie candidate, and they did so in South Carolina. In a way, the 2020 primary ended up being a reverse of the 2016 one, with an anti-Sanders bloc being a huge but not decisive force shaping it.

Why the GOP 2016 Primary Was Different

Perhaps the Sanders campaign imagined winning the primary in a similar way to Trump 2016. Trump won a lot of early states with only 30-35% of the vote, while Marco Rubio, Ted Cruz, and John Kasich split what amounted to a majority of the vote. Had two of those latter three dropped out before Super Tuesday, a la Buttigieg and Klobuchar in 2020, Trump might have been stopped, but it still would have been close due to his broad appeal across the GOP primary electorate.

Sanders 2020 was a very different beast from Trump 2016, and not just on policy grounds (obviously). First, he never really got to the 35% level Trump did, instead winning with 26% and 33%. His support was never as broad either, instead being concentrated in white and Hispanic voters under 50. And most of all, he wasn’t the choice of the party’s dominant faction – whereas Trump was the favorite of elderly conservatives (the GOP base), Bernie was not the choice of suburban moderates or older black voters, who between them now form the D base in many populous states.

Essentially, the GOP establishment could not have easily stopped Trump in 2016 due to his popularity with voters and the inability of his rivals to take collective action. In 2020, the Democratic establishment had the voters on its side, and the non-Bernie candidates were more than happy to defer to Biden, in a way that GOP candidates had not done for Ted Cruz in 2016.

The 2020 Democratic Primary, Part 1: Human-Wave Tactics Are So 20th Century


This is the first in a series of posts I’ll do on the 2020 Democratic Party primaries.

On a Saturday morning in March 2016, I stepped out of my Chicago home, got into an Uber, and took a ride to the Ukrainian Village. My destination was a small field office for the Bernie Sanders presidential campaign, which only that past Tuesday had pulled a shocking upset in the Michigan primary.

Despite polls showing Hillary Clinton leading by double digits, Sanders had pulled off a narrow win, buoyed by good margins in the state’s predominantly white rural areas and suburbs. Meanwhile, Clinton hadn’t been able to run up the score enough in Detroit.

The sheer size of the polling error had made me and likely millions of other Bernie supporters question if the race had fundamentally shifted in his favor, despite his struggles not long before on Super Tuesday, when he had only been able to salvage his home state and a few low-turnout caucuses. So I decided to go all-in on the Illinois primary that coming Tuesday.

My task that Saturday was simple: Canvass a list of homes – mostly apartment buildings – in the Ukranian Village. I got a clipboard, addresses, and campaign literature, and set off on my way with a few others.

That morning, I probably canvassed 100 addresses. The response rate was abysmal – I think maybe 2 or 3 people actually answered, and none of them even needed convincing. The number of non-answers was so high that I ran out of literature long before I had exhausted the list.

Basically, it was a huge waste of time. Bernie lost Illinois.

And yet, canvassing is a staple of any serious campaign, right?

The left’s nostalgia for a lost world

In the past, this sort of strategy was indeed integral to numerous  high-profile campaigns, Obama 2012 among them. Obama’s “ground game,” as it was called, was painted as one of the key reasons for his narrow win over Mitt Romney. And it’s even more important in low-turnout local and state elections, where vital offices can be won with just a few thousand votes.

But in presidential races? I don’t think it’s important anymore. My March 2016 morning would have yielded better results for the Sanders campaign had I simply been paid to write a blog post critical of Hillary Clinton and gotten it picked up by media personalities on Twitter.

The absolute best case against the value of the so-called ground game, intensive as it is on canvassing, is the Donald Trump 2016 campaign. Despite Donald Trump being a billionaire, his campaign was notably light on the paid advertisements and overwhelming field operations that people traditionally expect from high-budget operations. He had no ground game, or at least much less so than Ted Cruz and especially Hillary Clinton.

But he did have earned media. Every day, his face was all over CNN, Fox News, Twitter, you name it – he didn’t have to pay for any of that exposures and it all amounted to him getting billions in free coverage. The narrative was conducted on his terms, playing up issues he cared about most – immigration, Clinton’s email server – and never really veering out of his control except for the brief Access Hollywood cycle.

Now flash forward to 2020. Did the Democrats learn from this? For the most part, no.

Bernie Sanders and Elizabeth Warren in particular ignored Trump’s success with earned media and instead went hard on field operations – i.e., lots of local offices and canvassers, like me and the team in 2016. Warren’s Iowa operation was called a “field of dreams,” while Sanders’ team knocked half the doors in Iowa (at least) and were ubiquitous across the Hawkeye State in the weeks leading up to the primary.

To me, this approach reflected the left’s own nostalgia for something that can’t be recaptured: The days of people-powered campaigns, with lots of (often unionized) labor on the ground knocking doors, making calls, leaving flyers, etc. Carrying around a clipboard and talking to people about solidarity and universal healthcare is invigorating! But it’s also a 20th century notion, and politics have changed immensely. Just ask Donald Trump – and Pete Buttigieg.

Pete’s earned media

Alone among the 2020 Democratic candidates in the early part of the race, Buttigieg showed a remarkable ability to stay in the headlines despite having much less money and supporters than either Warren or Sanders. He went on every TV show, podcast, or other media opportunity he could, and the press loved him in response. It was McCain 2000 – another successful insurgent candidacy, at least relative to what it was up against – taken to the nth degree, and it almost worked!

While Warren and Sanders were making untold numbers of calls and door knocks in Iowa, Pete was chilling with the guys on Pod Save America and various Sunday morning shows. On election night in Iowa, he overperformed his polls, the other two underperformed, and he won, gaining another wave of media attention that nearly won him New Hampshire, too – and might have succeeded had it not been for a vicious takedown by Amy Klobuchar in the debate a few days before the primary.

Pete eventually ran aground in Nevada because his base was far too white, but the power of earned media that he – like Trump – had exploited was still there for the taking. And instead of Bernie tapping into it with some overtures at unity, it was Joe Biden, who generated a bunch of good news cycles from his pivotal endorsement by James Clyburn and his blowout win in South Carolina. Despite having essentially no presence in any Super Tuesday state, he won 10 out of 14 them – far outdoing Warren and Sanders, who between them had invested extensively throughout the country.

The earned media over that Feb. 29 – March 1 weekend did it for him. And it should make the left rethink its strategy and probably focus more on controlling media narratives going forward, as uncool as that seems.