Henry Farrell has critical words for the notion that technological progress is both inevitable and inevitably good:
Marc Andreessen’s recent “tech optimist manifesto” is one of the most significant statements of Silicon Valley ideology. As I’ve written elsewhere, it’s actually less a political manifesto than an apostolic credo for the Religion of Progress. The words “we believe” appear no less than 113 times in the text, not counting synonyms.
....The religion of the engineers is the hopium of Silicon Valley elites. It’s less a complex theology than an eschatological soporific, a prosperity gospel for venture capitalists, founders and wannabes. It tells its votaries that profits and progress point in exactly the same direction, and that by doing well they will most certainly do good. It should barely need pointing out that the actual problems and promise of technology lie in the current political struggles that this vision of the future waves away.
I haven't read the "Techno-Optimist Manifesto" and I don't want to. However, I've read a few bits of it and I wonder how much of the backlash is due to its actual content vs. its triumphalist language? The latter is why I don't want to read it, but I suspect that if it were toned down considerably I'd more-or-less agree with what it says.
The "Religion of Progress," as Henry puts it, is hardly new or unique to Silicon Valley. It's two or three centuries old and it really did mark a sea change in human affairs. Before the Enlightenment, people assumed that their lives would be mostly the same from birth to death. After that, we started to see change on a short enough timescale that it was noticeable during a lifetime. And after that, we adopted the notion not just that change happened, but that it was something we should actively seek out and encourage. This is the Religion of Progress, and you can date it back to Francis Bacon in its embryonic form and to around the mid-18th century in its more widespread form.
Henry knows all this better than me. It's pretty conventional stuff. So why the contempt for the modern version of it? Technological progress is obviously not all good—climate change, anyone?—but on the whole the prosperity of the world really does depend on it. If you stop technological progress you'll also stop economic growth.
The Silicon Valley version of this is mostly bound up in software, but so what? That just makes it closer to the Platonic ideal of progress: an advance in thinking that eventually produces material benefits. Artificial intelligence is closer still.
The better that AI gets, the more progress it will produce. What's more, truly advanced AI is far more likely to produce progress with fewer harmful effects. Why? Because it's smarter, more knowledgeable, and able to think more broadly. So even if Marc Andreessen expresses himself in the most annoying way possible, we should all be hoping like hell that he's right.
Looks like the last post got eaten. It's basically coming from a set of very left-wing writers and commentators, who have an outsized presence on social media and online media in general because the VC boom in media startups heavily favored strident, opinionated commentary in an attempt to get engagement. These folks don't like Silicon Valley for any number of reasons - its wealth, its influence over their industry, the right-wing politics of some figures in it, and so forth.
Oh, it's simple why they don't like Silicon Valley.
STEM represents an alternative source of power to the sounding-clever professions, and frequently bursts the pretensions of the sounding-clever professions. What's the joy in having an English Degree from Harvard if there's a chance that some random MIT graduate might correct anything you say, and in a way that you simply cannot rebut?
That's the personal side. On the political side, Silicon Valley and Tech generally are an affront to the epistemic learned helplessness championed by the left. If you're taught from day one that you can achieve nothing, that all your problems are caused by someone else, and that the only way forward is the growth of the state; well that's a useful Overton window for certain political parties.
Tech is a continual reminder of what BS this is, of how plenty of people in America (and plenty from, uh, "oppressed" classes) manage to do just fine, achieving things in spite of, not via, the state.
It's getting worse because of Elite Overproduction.
So let me get this straight. You haven't read it. You don't want to. You aren't going to. But you wrote about 350 words defending it and saying we should all hope Marc Andreesen is right.
I have read the manifesto. It's appalling. It's self-contradictory, it betrays shocking and unforgivable ignorance of basic political economy, and its squarely on the most reactionary side of the culture wars, which is tries to cloak in obfuscatory, coded language. (If a tech-libertarian is saying "I don't believe in answers that can't be questioned," nine times out of ten what they actually mean is "I think you all are being very mean and anti-intellectual by shouting down those polite race realists.")
Then we've built a fucked-up system that needs massive re-engineering, because designing something around "we must grow limitlessly, forever, or the wheels come off" is grotesque and appalling.
We're nowhere near AGI. People keep saying we are. There's been jack shit in the way of progress. People have lit mountains of money on fire and burned shocking amounts of electricity in order to product fancy algorithms that are worse than most search engines when it comes to actually providing one with useful knowledge.
Folks seem very convinced if we keep setting more money on fire and burning more and more resources and throwing more and more data onto the pile, it will transmute into their new AI god. I don't know why anyone should take that seriously.
+100
"... designing something around "we must grow limitlessly, forever, or the wheels come off" is grotesque and appalling."
Why?
Because we live in a finite, bounded world that in turn exists in a finite, bounded universe? Therefore ordering your society around "we must grow limitlessly" means you're making the horrible decision to shank a ton of people at some point when the wheels come off?
The people who sucked oceans of carbon out of the planet's crust and spewed them into the air were real big on "growth now, consequences tomorrow."
Nwabudike Morgan is a cautionary tale, not an example to follow.
"Because we live in a finite, bounded world that in turn exists in a finite, bounded universe? "
This "bound universe" is large enough that we are not going to exhaust it any time soon. In million years from now, maybe we will need to think about it, but not now.
"The people who sucked oceans of carbon .."
That is not the only posible way of progress. There are other ways.
And that "carbon sucking" was (and currently still is) absolutely essential for modern life. Now it needs to stop, but we wouldn't get that far without it, and we won't be able to stop it without progress.
"Nwabudike Morgan is a cautionary tale, not an example to follow."
If that supposed to say something interesting, I missed it.
"And that 'carbon sucking' was (and currently still is) absolutely essential for modern life. Now it needs to stop, but we wouldn't get that far without it, and we won't be able to stop it without progress."
An often-overlooked fact is that simply using fossil fuels _slower_ does not solve the problem (as long as we are using them on roughly civilization-level and not geologic-level timescales). Slowing down economic growth or more particularly technological development does not prevent the apocalypse, it only pushes it back by some nominal number of years. The only options to prevent a climate collapse are to decarbonize by going pre-industrial or decarbonize by going post-fossil; even if you wanted the former, you can't get the rest of humanity to do it, so the only other option is the latter. That means technological development.
And I say that as somebody who works in the field of decarbonization.
+1
You’re the problem.
The world may be finite, but the solar system is, compared to humanity, functionally infinite in resources.
We can and should continue to grow the economy. We should both count the unpaid family work now, pay for things necessary for everyone to live that we skip today, and allow people to buy, trade and share art and expression.
The reason art is more valuable today than ever - when it's being made in a greater amount that ever - is literally a near limitless expansion of our economy in itself.
Yeah, why NOT? It works just fine for cancer!
I agree with all of this. (Also remember, "AI" is as much a product of the fossil fuel economy as the semi-tractor.)
As someone who has worked in SV for basically my entire career, let me add, Andreesen is not really talking about technological progress by itself. If his screed were about any number of things promoting the advancement of tech, any tech, that would be one thing. But he isn't.
He wants particular sorts of people with particular priors to do great things, which he believes can only happen if they're unshackled to, I dunno, blow up more rockets faster or not get have to listen to the little people get lippy over using the N-word. It is a less interesting (but thankfully far shorter) Ann Rand screed calibrated to his current venture portfolio.
More fundamentally, these people want to be worshipped. They really have come to believe that vast wealth indicates vast wisdom, even then that wisdom is half-baked comic book history meets middle-age masculinity crisis.
Frankly, his rant is boring and vapid. The only reason we're talking about it is because a rich reactionary wrote an ode to his investments and cronies. And that's all you really need to know.
Listening to monomaniacal tech bros lecture us on history and the shape of society (Sparta!) is to learn that skill in one field does not translate to other areas.
AGI is just moving the goalposts while people lose their jobs to AI algorithms.
Every time a new technology comes along people lose jobs but in the long term more jobs and growth are are created. My grandfather designed the first assembly line for packing coffee for the Yuban company but he later felt really guilty about putting all those young women out of a job. Our economy and job market have grown enormously since that time and automation was a big part of that.
It’s possible that this time will be different but I doubt it. There are plenty of creative people who will figure out new ways to use AI for their new businesses.
That is the perfect, one-sentence comeback for this stuff, Crissa. Well played!
Nowhere does he defend it.
Well, technological progress (including medical) coupled with increasing population…
I’m paid $185 per hour to complete the task using an Apple laptop. I absolutely didn’t think it was conceivable,(Qx) but my dependable buddy convinced me to give this straight forward chance a go after she made $26,547 in just 4 weeks working on it. Visit the following page to find out additional
instructions———>>> cyberfortifypro555.pages.
“The better that AI gets, the more progress it will produce. What's more, truly advanced AI is far more likely to produce progress with fewer harmful effects.”
Uh huh. Except for the tens of millions or more that it displaces out of their jobs, leaving them struggling to maintain their households. People can’t easily exist without a steady stream of money in the modern world, and the modern world is structured so that 90% or more of all income and wealth is hoovered up by the upper classes.
“Progress” has a tendency to leave some folks behind. Creative destruction etc. We long ago agreed we were ok with it.
Let us suppose that AI is actually close. By all indications, when humanity does create intelligent artificial beings, their primary purpose will be to make money for their creators. Not to improve the human lifespan, not to explore the cosmos, not to increase happiness or to find ways to slow the greenhouse effect: although these things might happen incidentally. But they will be owned by their creators and those creators are obviously and openly more interested in profiting from AI than they are in improving the world for everyone, even people in countries which do not provide 24-hour electricity to many of their citizens.
When the people who control the instrumentalities of progress are overwhelmingly white, male, middle-age, upper-class, tech "geniuses" we might reasonably ask whether that's the best group to be making all the decisions, secretly, accountable mainly to their investors and not accountable at all to the life-forms they might deliberately create.
When humanity in general is denied the benefits of progress in the name of massive profit for a tiny portion of humanity, whether that portion sells insulin or artificial intelligence doesn't matter all that much.
Great. More advanced bullshit and hallucinations. As if the world didn't already have enough of that.
And before you start talking about how AI is smarter and more knowledgeable and able to think broadly, first lets see some evidence that it knows anything at all and is able to think period. I'll give it a pass on thinking smartly.
The truth is currently neither of those is true. LLMs quite literally know nothing at all. They are an encoding of statements that often go together. The meaning of those statements is irrelevant. Which leads directly to the thinking part. LLMs cannot create new thoughts that aren't hallucinations because they know nothing other than correlations between statements.
I'll wager they will be in about the same place in 5 years as they are today because expressing thoughts that have already been had and written down is not creativity by any definition.
What are new thoughts, then?
What is prosperity? I think it's a free life devoid of war and conflict.
Nerds, progress, technology, it always wins in the end. Slow, but inevitable.
But we always have to stand up and make sure it applies to real people in the here and now.
"The religion of the engineers is the hopium of Silicon Valley elites. It’s less a complex theology than an eschatological soporific..."
OMG, what loaded language! And big words!
I don't think "escatologicial" means what he thinks it means.
"Hopium" is silly. I agree. But: Silly language does not prove his arguments wrong. Pointing out is an easy teacher trick, not a valid argument.
As someone else noted, there's a certain desperate quality to Andreessen's manifesto, because his particular VC fund went all in on crypto. That turned out to have no value to anyone but criminals. Now it's all in on AI. Apparently Sundar Pachai, CEO of Alphabet/Google, just said "AI will be bigger than electricity, and maybe bigger than fire".
Yeah right. There's the red flag for the hype-bubble right there. It'll be big for Google because it gives their customers the content of other people's web pages without clicking on them, thus stealing any advertising revenue for itself. Will it be a big deal for anyone else? I already see Quora being dumbed-down by LLM answers, and DeviantArt being flooded by mediocre images.
Anyway, AI and the cloud are actually the only new techs with promise in some time from the big boys. Google's search performance is getting worse, Apple is selling the same phone as 10 years ago, Microsoft, OMG, Facebook failed on the Metaverse, and Netflix has broken its promise of the world's movies at your fingertips. These guys are screaming about their genius at a time when they're not delivering.
"new techs": rocket launches are a hell of a lot cheaper than they used to be.
And you kind of left Amazon out of your list of failures...
"truly advanced AI is far more likely to produce progress with fewer harmful effects. Why? Because it's smarter, more knowledgeable, and able to think more broadly."
Advanced AI already exists. Humans communicating and coordinating is artificial intelligence, especially at the level of hundreds of thousands to millions of people. It's unlikely that we will progress with fewer harmful effects than we currently have, and we are unlikely to think more broadly than we currently do.
Well, I have a social science experiment I'm conducting. The thesis is, "You can't have a discussion of Silicon Valley without someone engaging in masculinity shaming".
This comment section does not contradict my thesis. Not that I sign on to Marc Andreeson, who did one very smart, very valuable thing, and for such was blessed with a lot of money, which he has more or less pissed away.
In my entire career, I have had as one of my criteria for taking a job be that the work I was helping with would make things better for quite a few people, and not just the 10 percent.
I have specifically turned down working on hedge fund type software precisely because of this. I have no interest in making rich people richer. They can help themselves just fine, they don't need me.
I say this not to pat myself on the back, but to note that it all falls under the banner of Progress. One needs to be careful in a way Andreessen does not address.
There is a process here in Silicon Valley that is still a very good one. It has brought many great things to a broad swath of people and that process - never mind people like Marc Andreessen - will continue to do so. The process is "try stuff that we think we can get to work and seems like it will help. Take some risks making the try. If it didn't work, ditch it and try again."
That is too risky for most people's taste, and especially for most investor's tastes. It happens here. It might happen elsewhere, I'm not sure.
That's not a belief, or a manifesto. It's an observation.