The most important thing that happened in the US in 2016 was smartphones hitting market saturation in the US.
You might argue that the election of Donald Trump as President was more important, but I'm going to argue that this market saturation is actually a lot of what led to him being elected.
Go west young man
Come my tan-faced children,
Follow well in order, get your weapons ready,
Have you your pistols? have you your sharp-edged axes?
Pioneers! O pioneers!
For we cannot tarry here,
We must march my darlings, we must bear the brunt of danger,
We the youthful sinewy races, all the rest on us depend,
Pioneers! O pioneers!
O you youths, Western youths,
So impatient, full of action, full of manly pride and friendship,
Plain I see you Western youths, see you tramping with the foremost,
Pioneers! O pioneers!
Have the elder races halted?
Do they droop and end their lesson, wearied over there beyond the seas?
We take up the task eternal, and the burden and the lesson,
Pioneers! O pioneers!
All the past we leave behind,
We debouch upon a newer mightier world, varied world,
Fresh and strong the world we seize, world of labor and the march,
Pioneers! O pioneers!
-- Walt Whitman
There has never been a time in American history when there wasn't a frontier. Buried deep in the American cultural fabric, deeper than protestant work ethic and yankee ingenuity, is the idea not just of growth but of growth into the unknown. The foundation of the American project was relentless expansion of territory (which separates America from Europe) performed by yeomen and petite bourgeois (which separates North America from South America, which operated on a system of large patrón landowners who were vassal to the european metropole). That expansion produced coal from appalachia for steam engines, then buffalo hides from the plains for machine belts, iron from colorado for steel bridges, and copper from alaska for electrical wiring. After the land ran out then over the ocean we had the take-over of the Phillipenes and Hawaii. By the time that was over consumerism was in full swing and expansion came into a parade of new product categories that everybody had to have: tractors, cars, fridges, washing machines, record players, televisions, transistor radios, plastic tupperware, computers, CD-ROMs, the internet, broadband, smartphones... and then what?
The end of history
And then what? Silicon Valley not only exists culturally on the premise that growth into green-field new markets is always possible, but the stock option compensation sturcture they use is a ponzi scheme if it isn't. They Must find a Next Thing. And they've been trying, as you may have noticed. They tried self-driving cars, and that turned out to be too difficult (the only player left in that game is google/waymo, and their product only works for city driving and is partly remote-controlled by workers in India instead of being fully autonomous). They tried VR goggles, but no one knew what to do with them and besides they give people headaches. They tried NFTs, where no one had a coherent idea of what they were useful for. The covid pandemic did drive a huge amount of demand for telecommuting, which gave the industry a reprieve, but that demand dropped off after every knowledge worker already had a work-from-home setup.
I wrote a post on my old blog in early 2016 (so right when smartphones were approaching the limit) trying to call what the potential next things would be. Two of those things were complete whiffs (cheminformatics and global development). The one other thing on that list that I haven't mentioned yet is LLMs.
In the post I talked about "sequence-to-sequence learning", because that's what they were called at the time (transformers wouldn't be invented for another year). To quote myself:
I think improved machine translation is going to be the single best result of the recent progress in deep neural networks (more than computer vision, which is where the most action has been so far).
The year of the bull
Transformers do great at translation, speech recognition, text to speech, predictive text/autocompletion. They can even write prose to complete an input prompt, and it's usually even mostly coherent. The prose-writing is neat in a "monkey trained to play checkers" sort of way, and I think that's where most of the initial public reaction came from. Look, a talking dog! I mean computer!
That initial reaction generates enough razzle-dazzle that if you're quick enough you can sell AI with the other hand. It's taken five years but I think through the tech diffusion process, as people actually try to apply this stuff to real problems, the inherent limitations of the technology are becoming more widely understood.
In the meantime the stock market valuations of the companies doing the razzling and dazzling grew to the point that they're now over a third of the S&P 500 (you know, the one that your retirement fund is in). What happens when people have heard enough of this spectacle, this tightrope walk?
The superman
When Zarathustra had thus spoken, one of the people called out: “We have now heard enough of the [tightrope walker]; it is time now for us to see him!” And all the people laughed at Zarathustra. But the [tightrope walker], who thought the words applied to him, began his performance.
Zarathustra, however, looked at the people and wondered. Then he spake thus:
Man is a rope stretched between the animal and the Superman — a rope over an abyss.
A dangerous crossing, a dangerous wayfaring, a dangerous looking-back, a dangerous trembling and halting.
[...]
I love those who do not first seek a reason beyond the stars for going down and being sacrifices, but sacrifice themselves to the earth, that the earth of the Superman may hereafter arrive.
I love him who liveth in order to know, and seeketh to know in order that the Superman may hereafter live. Thus seeketh he his own down-going.
I love him who laboureth and inventeth, that he may build the house for the Superman, and prepare for him earth, animal, and plant: for thus seeketh he his own down-going.
[...]
I love him whose soul is so overfull that he forgetteth himself, and all things are in him: thus all things become his down-going.
I love him who is of a free spirit and a free heart: thus is his head only the bowels of his heart; his heart, however, causeth his down-going.
I love all who are like heavy drops falling one by one out of the dark cloud that lowereth over man: they herald the coming of the lightning, and succumb as heralds.
Lo, I am a herald of the lightning, and a heavy drop out of the cloud: the lightning, however, is the SUPERMAN.
While all this has been going on the other strand of culture has been the rise of nationalism. Nationalism is bipartisan now. China is a threat to our collective honor, apparently. The Biden administration decided the way to respond to trump was to be just as authoritarian, but to do it through the proper channels.
The relationship between the military-industrial complex and silicon valley has always been intimate. Steve Blank presents the history of that relationship really well so I'm not going to rehash that here, but suffice to say large companies in silicon valley not wanting to be war profiteers is something that really only existed in the 2000s.
Palantir is the most famous of the current crop of silicon valley defense contractors. Their business is to collect various sources of data into a single database, use that data to draw statistical inferences, and then present those inferences to operators. The operations those operators perform may be of the kinetic variety. The data sources include license plate readers in the US.
The mag7 companies are just as involved, though. AWS has a datacenter specifically for use by spies and spooks. So does Microsoft.
When AI demand dries up and they have huge datacenter overcapacity that they need customers for, wouldn't it be convenient if there was suddenly huge demand for AI Cyberweapons (for the Warfighter, in the Battlespace, Kinetic)?
This explains a lot of why so many people in top positions in silicon valley pivoted to supporting Trump in 2024. For the public to accept war mobilization they need not just a credible enemy (and for most people outside of DC China isn't credible enough), you also need a figurehead that people can rally behind. For millions of Americans Trump is that figurehead.
Combine that with a large cohort of people who genuinely believe that any day now an AI godhead will emerge. A superman. “Commerce is our goal here at Tyrell. ‘More human than human’ is our motto.”, to quote Blade Runner. You don't have to spend much time on LessWrong to find people willing to fall into the abyss to bring it about.
The question is whether the public is willing to fall into the abyss with them.