Posts Tagged ‘computing’
“The clustering of technological innovation in time and space helps explain both the uneven growth among nations and the rise and decline of hegemonic powers”*…
As scholars like Robert Gordon and Tyler Cowan have begun to call out a slowing of progress and growth in the U.S., others are beginning to wonder if “innovation clusters” like Silicon Valley are still advantageous. For example, Brian J. Asquith…
In 2011, the economist Tyler Cowen published The Great Stagnation, a short treatise with a provocative hypothesis. Cowen challenged his audience to look beyond the gleam of the internet and personal computing, arguing that these innovations masked a more troubling reality. Cowen contended that, since the 1970s, there has been a marked stagnation in critical economic indicators: median family income, total factor productivity growth, and average annual GDP growth have all plateaued…
In the years since the publication of the Great Stagnation hypothesis, others have stepped forward to offer support for this theory. Robert Gordon’s 2017 The Rise and Fall of American Growth chronicles in engrossing detail the beginnings of the Second Industrial Revolution in the United States, starting around 1870, the acceleration of growth spanning the 1920–70 period, and then a general slowdown and stagnation since about 1970. Gordon’s key finding is that, while the growth rate of average total factor productivity from 1920 to 1970 was 1.9 percent, it was just 0.6 percent from 1970 to 2014, where 1970 represents a secular trend break for reasons still not entirely understood. Cowen’s and Gordon’s insights have since been further corroborated by numerous research papers. Research productivity across a variety of measures (researchers per paper, R&D spending needed to maintain existing growth rates, etc.) has been on the decline across the developed world. Languishing productivity growth extends beyond research-intensive industries. In sectors such as construction, the value added per worker was 40 percent lower in 2020 than it was in 1970. The trend is mirrored in firm productivity growth, where a small number of superstar firms see exceptionally strong growth and the rest of the distribution increasingly lags behind.
A 2020 article by Nicholas Bloom and three coauthors in the American Economic Review cut right to the chase by asking, “Are Ideas Getting Harder to Find?,” and answered its own question in the affirmative.6 Depending on the data source, the authors find that while the number of researchers has grown sharply, output per researcher has declined sharply, leading aggregate research productivity to decline by 5 percent per year.
This stagnation should elicit greater surprise and concern because it persists despite advanced economies adhering to the established economics prescription intended to boost growth and innovation rates: (1) promote mass higher education, (2) identify particularly bright young people via standardized testing and direct them to research‑intensive universities, and (3) pipe basic research grants through the university system to foster locally-driven research and development networks that supercharge productivity…
…
… the tech cluster phenomenon stands out because there is a fundamental discrepancy between how the clusters function in practice versus their theoretical contributions to greater growth rates. The emergence of tech clusters has been celebrated by many leading economists because of a range of findings that innovative people become more productive (by various metrics) when they work in the same location as other talented people in the same field. In this telling, the essence of innovation can be boiled down to three things: co-location, co-location, co-location. No other urban form seems to facilitate innovation like a cluster of interconnected researchers and firms.
This line of reasoning yields a straightforward syllogism: technology clusters enhance individual innovation and productivity. The local nature of innovation notwithstanding, technologies developed within these clusters can be adopted and enjoyed globally. Thus, while not everyone can live in a tech cluster, individuals worldwide benefit from new advances and innovations generated there, and some of the outsized economic gains the clusters produce can then be redistributed to people outside of the clusters to smooth over any lingering inequalities. Therefore, any policy that weakens these tech clusters leads to a diminished rate of innovation and leaves humanity as a whole poorer.
Yet the fact that the emergence of the tech clusters has also coincided with Cowen’s Great Stagnation raises certain questions. Are there shortcomings in the empirical evidence on the effects of the tech clusters? Does technology really diffuse across the rest of the economy as many economists assume? Do the tech clusters inherently prioritize welfare-enhancing technologies? Is there some role for federal or state action to improve the situation? Clusters are not unique to the postwar period: Detroit famously achieved a large agglomeration economy based on automobiles in the early twentieth century, and several authors have drawn parallels between the ascents of Detroit and Silicon Valley. What makes today’s tech clusters distinct from past ones? The fact that the tech clusters have not yielded the same society-enhancing benefits that they once promised should invite further scrutiny…
How could this be? What can we do about it? Eminently worth reading in full: “Superstars or Black Holes: Are Tech Clusters Causing Stagnation?” (possible soft paywall), from @basquith827.
See also: Brad DeLong, on comments from Eric Schmidt: “That an externality market failure is partly counterbalanced and offset by a behavioral-irrationality-herd-mania cognitive failure is a fact about the world. But it does not mean that we should not be thinking and working very hard to build a better system—or that those who profit mightily from herd mania on the part of others should feel good about themselves.”
###
As we contemplate co-location, we might recall that it was on this date in 1956 that a denizen of one of America’s leading tech/innovation hubs, Jay Forrester at MIT [see here and here], was awarded a patent for his coincident current magnetic core memory (Patent No. 2,736,880). Forrester’s invention, a “multicoordinate digital information storage device,” became the standard memory device for digital computers until supplanted by solid state (semiconductor) RAM in the mid-1970s.
“Man is not disturbed by events, but by the view he takes of them”*…
From Stripe Partners, a framework for rethinking the way we talk about the AI future…
AI is both a new technology and a new type of technology. It is the first technology that learns and that has the potential to outstrip its makers’ capabilities and develop independently.
As Large Language Models bring to life the realities of AI’s potential to operate at unprecedented, ‘human’ levels of sophistication, projections about its future have gained urgency. The dominant framework being applied to identify AI’s potential futures is 165 years old: Charles Darwin’s theory of evolution.
Darwin’s evolutionary framework is rendered most clearly in Dan Hendycks work for the Center for AI Safety which posits a future where natural selection could cause the most influential future AI agents to have selfish tendencies that might see AI’s favour their own agendas over the safety of humankind.
The choice of Natural Selection as a framework makes sense given AI’s emerging status as a quasi-sentient, highly adaptive technology that can learn and grow. The choice is a response to the limitations inherent in existing models for technological adoption which treat technologies as inert tools that only come to life when used by people.
The risk in applying this lens to AI is that it goes too far in assigning independent agency to AI. Estimates on the timing of the emergence of ‘Artificial General Intelligence’ vary, but spending some time with the current crop of Generative AI platforms confirms the view that AI’s with intelligences that are closer to humans are some way off. In the interim using natural selection as a lens to understand AI positions humans as further out of the developmental loop than is actually the case. Competitive forces whether market or military will shape AI’s development, but these will not be the only forces at play and direct interaction with humans will be the principal driver for AI’s progress in the near term.
A year ago we wrote about the opportunity to reframe the impact of AI on organisations through the lens of Actor Network Theory (ANT). More than a singular theory, ANT describes an approach to studying social and technological systems developed by Bruno Latour, Michel Callon, Madeleine Akrich and John Law in the early 1980s.
ANT posits that the social and natural world is best understood as dynamic networks of humans and nonhuman actors… In our 2023 piece we suggested that ANT, with its focus on framing society and human-technology interactions in terms of dynamic networks where every actor whether human or machine impacts the network, was a useful way of exploring the ways in which AI will impact people, and people will impact AI.
A year on the value of ANT as a framework for exploring AI’s future has become clearer. The critical point when comparing an ANT frame to an evolutionary one is the way in which the ANT framing highlights how AI will progress with and through people’s interactions with it. When viewed as an actor in a network, not a technology in isolation, AI will never be separate from human interventions…
A provocative argument, well worth reading in full: “Why the debate about the future of AI needs less Darwin and more Latour,” from @stripepartners.
Apposite: “Whose risks? Whose benefits?” from Mandy Brown.
* Epictetus
###
As we reframe, we might recall that it was on this date in 1946 that an ancestor of today’s AIs, the ENIAC (Electronic Numerical Integrator And Computer), was first demonstrated in operation. (It was announced to the public the following day.) The first general-purpose computer (Turing-complete, digital, and capable of being programmed and re-programmed to solve different problems), ENIAC was begun in 1943, as part of the U.S’s war effort (as a classified military project known as “Project PX“); it was conceived and designed by John Mauchly and Presper Eckert of the University of Pennsylvania, where it was built. The finished machine, composed of 17,468 electronic vacuum tubes, 7,200 crystal diodes, 1,500 relays, 70,000 resistors, 10,000 capacitors and around 5 million hand-soldered joints, weighed more than 27 tons and occupied a 30 x 50 foot room– in its time the largest single electronic apparatus in the world. ENIAC’s basic clock speed was 100,000 cycles per second (or Hertz). Today’s home computers have clock speeds of 3,500,000,000 cycles per second or more.
“Simplicity, carried to the extreme, becomes elegance”*…
Jordana Cepelewicz on a very different approach to computing…
In 1936, the British mathematician Alan Turing came up with an idea for a universal computer. It was a simple device: an infinite strip of tape covered in zeros and ones, together with a machine that could move back and forth along the tape, changing zeros to ones and vice versa according to some set of rules. He showed that such a device could be used to perform any computation.
Turing did not intend for his idea to be practical for solving problems. Rather, it offered an invaluable way to explore the nature of computation and its limits. In the decades since that seminal idea, mathematicians have racked up a list of even less practical computing schemes. Games like Minesweeper or Magic: The Gathering could, in principle, be used as general-purpose computers. So could so-called cellular automata like John Conway’s Game of Life, a set of rules for evolving black and white squares on a two-dimensional grid.
In September 2023, Inna Zakharevich of Cornell University and Thomas Hull of Franklin & Marshall College showed that anything that can be computed can be computed by folding paper. They proved that origami is “Turing complete” — meaning that, like a Turing machine, it can solve any tractable computational problem, given enough time…
Read on for more on how folding paper can, in principle, be used to perform any possible computation: “How to Build an Origami Computer” from @jordanacep in @QuantaMagazine.
* Jon Franklin
###
As we contemplate calculation, we might send entropic birthday greeting to Rolf Landauer; he was born on this date in 1927. A physicist, we made important contributions made important contributions in several areas of the thermodynamics of information processing, condensed matter physics, and the conductivity of disordered media… most of which important to the development of computing (of the electronic variety).
He is best known for his discovery and formulation of what’s known as Landauer’s principle: that in any logically irreversible operation that manipulates information, such as erasing a bit of memory, entropy increases and an associated amount of energy is dissipated as heat– a “thermodynamic cost of forgetting,” relevant to chip design (how closely packed elements can be on a chip and still handle the heat), reversible computing, quantum information, and quantum computing… but not an issue for origami.)
“The world of reality has its limits; the world of imagination is boundless”*…
… and the digital world? Maybe, as Rob Beschizza reports, somewhere in between…
Alex set out to debunk the given wisdom that the maximum dimensions of a PDF are 381 km2, which is smaller than Germany. She presents her conclusions in an article titled “Making a PDF that’s larger than Germany,” so you know from the outset she succeeded. It’s a fascinating example of the disalignment of specifications, implementations, and reality. You can make one by hacking the postscript, and while Adobe Acrobat won’t like it, other apps will…
Borges would be delighted…
“On exactitude in PDFs“
Just how big was Alex [Chan] able to make her PDF?…
… unlike Acrobat, the Preview app doesn’t have an upper limit on what we can put in MediaBox. It’s perfectly happy for me to write a width which is a 1 followed by twelve 0s…
If you’re curious, that width is approximately the distance between the Earth and the Moon. I’d have to get my ruler to check, but I’m pretty sure that’s larger than Germany.
I could keep going. And I did. Eventually I ended up with a PDF that Preview claimed is larger than the entire universe – approximately 37 trillion light years square. Admittedly it’s mostly empty space, but so is the universe. If you’d like to play with that PDF, you can get it here.
Please don’t try to print it.
“Making a PDF that’s larger than Germany“
* Jean-Jacques Rousseau
###
As we scale up, we might spare a thought for Émile Borel; he died on this date in 1956. A mathematician (and politician who served as French Minister of the Navy), he is remembered for his foundational work in measure theory and probability. He published a number of research papers on game theory and was the first to define games of strategy.
But Borel may be best remembered for a thought experiment he introduced in one of his books, proposing that a[n immortal] monkey hitting keys at random on a typewriter keyboard will – with absolute certainty – eventually type every book in France’s Bibliothèque Nationale de France. This is now popularly known as the infinite monkey theorem.
“Even a fool who keeps silent is considered wise; when he closes his lips, he is deemed intelligent.”*…
A substantial– and important– look at a troubling current aflow in the world of technology today: Emily Gorcenski on the millenarianism and manifest destiny of AI and techno-futurism…
… Early Christian missionaries traveled the pagan lands looking for heathens to convert. Evangelical movements almost definitionally involve spreading the word of Jesus Christ as a core element of their faith. The missionary holds the key that unlocks eternal life and the only cost is conversion: the more souls saved, the holier the work. The idea of going out into the world to spread the good word and convert them to our product/language/platform is a deep tradition in the technology industry. We even hire people specifically to do that. We call them technology evangelists.
Successful evangelism has two key requirements. First, it must offer the promised land, the hope of a better life, of eternal salvation. Second, it must have a willing mark, someone desperate enough (perhaps through coercion) to be included in that vision of eternity, better if they can believe strongly enough to become acolytes themselves. This formed the basis of the crypto community: Ponzi schemes sustain only as long as there are new willing participants and when those participants realize that their own continued success is contingent on still more conversions, the incentive to act in their own best interest is strong. It worked for a while to keep the crypto bubble alive. Where this failed was in every other aspect of web3.
…
There’s a joke in the data science world that goes something like this: What’s the difference between statistics, machine learning, and AI? The size of your marketing budget. It’s strange, actually, that we still call it “artificial intelligence” to this day. Artificial intelligence is a dream from the 40s mired in the failures of the ’60s and ’70s. By the late 1980s, despite the previous spectacular failures to materialize any useful artificial intelligence, futurists had moved on to artificial life.
Nobody much is talking about artificial life these days. That idea failed, too, and those failures have likewise failed to deter us. We are now talking about creating “cybernetic superintelligence.” We’re talking about creating an AI that will usher a period of boundless prosperity for humankind. We’re talking about the imminence of our salvation.
The last generation of futurists envisioned themselves as gods working to create life. We’re no longer talking about just life. We’re talking about making artificial gods.
…
I’m certainly not the first person to shine a light on the eschatological character of today’s AI conversation. Sigal Samuel did it a few months back in far fewer words than I’ve used here, though perhaps glossing over some of the political aspects I’ve brought in. She cites Noble and Kurzweil in many of the same ways. I’m not even the first person to coin the term “techno-eschatology.” The parallels between the Singularity Hypothesis and the second coming of Christ are plentiful and not hard to see.
…
… The issue is not that Altman or Bankman-Fried or Andreesen or Kurzweil or any of the other technophiles discussed so far are “literally Hitler.” The issue is that high technology shares all the hallmarks of a millenarian cult and the breathless evangelism about the power and opportunity of AI is indistinguishable from cult recruitment. And moreover, that its cultism meshes perfectly with the American evangelical far-right. Technologists believe they are creating a revolution when in reality they are playing right into the hands of a manipulative, mainstream political force. We saw it in 2016 and we learned nothing from that lesson.
Doomsday cults can never admit when they are wrong. Instead, they double down. We failed to make artificial intelligence, so we pivoted to artificial life. We failed to make artificial life, so now we’re trying to program the messiah. Two months before the Metaverse went belly-up, McKinsey valued it at up to $5 trillion dollars by 2030. And it was without a hint of irony or self-reflection that they pivoted and valued GenAI at up to $4.4 trillion annually. There’s not even a hint of common sense in this analysis.
This post won’t convince anyone on the inside of the harms they are experiencing nor the harms they are causing. That’s not been my intent. You can’t remove someone from a cult if they’re not ready to leave. And the eye-popping data science salaries don’t really incentivize someone to get out. No. My intent was to give some clarity and explanatory insight to those who haven’t fallen under the Singularity’s spell. It’s a hope that if—when—the GenAI bubble bursts, we can maybe immunize ourselves against whatever follows it. And it’s a plea to get people to understand that America has never stopped believing in its manifest destiny.
David Nye described 19th and 20th century American perception technology using the same concept of the sublime that philosophers used to describe Niagara Falls. Americans once beheld with divine wonder the locomotive and the skyscraper, the atom bomb and the Saturn V rocket. I wonder if we’ll behold AI with that same reverence. I pray that we will not. Our real earthly resources are wearing thin. Computing has surpassed aviation in terms of its carbon threat. The earth contains only so many rare earth elements. We may face Armageddon. There will be no Singularity to save us. We have the power to reject our manifest destinies…
Eminently worth reading in full: “Making God,” from @EmilyGorcenski (a relay to mastodon and BlueSky).
See also: “Effective Obfuscation,” from Molly White (@molly0xFFF) and this thread from Emily Bender (@emilymbender).
* Proverbs 17:28
###
As we resist recruitment, we might spare a thought for Ada Lovelace (or, more properly, Augusta Ada King, Countess of Lovelace, née Byron); she died on this date in 1852. A mathematician and writer, she is chiefly remembered for her work on Charles Babbage‘s proposed mechanical general-purpose computer, the Analytical Engine— for which she authored what can reasonably be considered the first “computer program.” She was the first to recognize that the machine had applications beyond pure calculation, and so is one of the “parents” of the modern computer.
You must be logged in to post a comment.