It is possible that Great Stagnation in biomedical sciences is over, but outside of that, it is as bad as it was in the recent 50 years. A lot of optimism seems to come from broad misunderstanding of so called "AI".

Actually I am surprised that Tyler repeats all media hype about "AI". Sure, stupid journalists, who can not solve quadratic equation, do not have a clue that their "AI" is just semi-parametric statistics, but Tyler is supposed to know better. 90% of current AI/ML algorithms are just image recognition algorithms, which do highly-efficient data mining. We are as far from true AI as we were in 1960s and it is still not clear whether it is possible at all. If it is possible, then it will be based on symbolic logic, which has nothing to do at all with everything, called "AI" by media now.

Notice that apart of biomedicine and "AI", all Tyler`s examples are about ventures of Musk. So, apart of his idiosyncratic case, there is nothing left. Nuclear fusion has been 20 years in the future for the last 60 years.

" but outside of that, it is as bad as it was in the recent 50 years. A lot of optimism seems to come from broad misunderstanding of so called "AI"."

In 2003, I wrote on a forum that machine translation would significantly improve by 2015 using statistical methods to the point where the job loss in translation would be noticeable and a translator who had worked on AI in graduate school replied that was "patently ridiculous" since a computer can't understand language, which is obviously beside the point.

Google Translate was still almost five years away and look at the quality gains with not just Google Translate from 2008 but specialized machine translation systems.

I completely agree that machine translation will replace most of human translation in the near future. But again, this has nothing to do with AI. If you present improvements in machine translation as having algorithms, which can perform this tasks efficiently, I am totally with you. Framing them as AI or even ML obscures the fact that these are simple data mining algorithms, applicable to a single task and unable by construction to do anything else.

We get stuck on the problem at hand.
The simplest Eth2.0 proof of stake node is you and I, each being both counter party and proof of stake verification. You and I have agreed to let our bots execute the provable contract, and that means bots chatting in secret. This is a bigger bugaboo than we currently assume, it is the hidden wedge in AI, and a stinky subject.

There are other connections.
In the case above, each hand held bot holds a private key and a counter party key. Two humans agree that the private keys remain private, across the board. The new Apple multi-core has a heural net capability with the ability to generate a difficult, but repeatable private key. And one core can be isolate with a snip of a few gates, such this it is always an unknowable core operating from a proven micro code. It is a few bits of logic code to make the unknowable core. The technologies here are pushing very hard on key government institutions, and the institutional stall leaves openings for virus problems. That stall is a huge cost to fiat banks. That stall is making it difficult to operate legal drones. And it is quite a huge step.

However, as any graduate student will tell you, a big part of the challenge of data analysis is getting ahold of a half-way interesting dataset in the first place. The internet, smart-phones and cheap storage allow for the harvesting of data from individuals and companies on a scale that has never been possible before. Couple that with the faster processing speed that allows you to run non-trivial algorithms on huge quantities of data that almost certainly was not possible 30 years ago.

Sure, and this will create a lot of value. I agree that image recognition algorithms as well as "big data" will have major impact on many industries. Just do not call them AI and do not pretend that they are applicable outside of a small set of very narrow tasks.

In the 80s and into the 90s, there was strong AI (HAL9000 types) and weak AI (Alphafold types) but somehow the later became AI whereas the former became AGI.

Ai in the 1960s couldn't do word2vec, google translate, GPT3 or now in France to make earnest attempts at modelling intrinsic motivations and goal-directed behaviors through the compositionality of language. (Stanislas Dehaene is great reading)

Once you get the AI to the edge of understanding words then it's the final step to full blown consciousness. This is the final decade of the human.

AI did not exist in 1960s and does not exist now. The fact that some algorithms can perform one statistically simple task (like translation) at a level, acceptable for most applications, does not make them AI. Not even close. If you forgot, AI means Artificial Intelligence.

Algorithms, currently called AI, are not on the verge of understanding words. In fact they can not "understand" anything, since they are mechanically unable to do anything, resembling thinking.

And all the talk about "earnest attempts at modelling intrinsic motivations and goal-directed behaviors through the compositionality of language" is obvious nonsense. Sure, there are some people working on it, and of course they will try to sell their work. But it does not mean they are getting anywhere.

It’s not just image recognition anymore my friend. The generative models (GANs) are showing strong results ex-post in finding the latent variables (key features) we ourselves find. Using words to report on its own internal states is next. “Getting anywhere” would mean the end of the game, consciousness. Like our own evolution of life on Earth words came basically at the very end. And then pop.

“What I cannot create, I do not understand.” — Richard Feynman

A provable contract, sample:
'After first payment the product is delivered with each party paying half the delivery charges".
This is provable because future prices are congestion priced in a market, the delivery market. Each party thus can estimate current market costs before start of provable contract. Risk is taken off the back of bots, making all these contract automated as long as fair congestion prices gate the exits.

">> as we might simply retreat into our former sloth and complacency once we are mostly vaccinated

I don't know about this. I have never been more complacent and sloth-like in my life as during this pandemic. Everything I could possibly do to progress the world has been blocked.

My impression is that the pandemic caused the need for various decisions to be made, and due to the ideology of the times, many of the decisions are really stupid. For example, what should we do about college admissions testing when we can't hold it in person with pencil and paper tests due to the virus? Our elites thunk and thunk and many of them decided to permanently get rid of testing in the name of Diversity Inclusion Equity (DIE).

Your elites maybe. The British elites tried this first instead.

www.theguardian.com/education/2020/aug/13/england-a-level-downgrades-hit-pupils-from-disadvantaged-areas-hardest
Pupils from disadvantaged backgrounds have been worst hit by the controversial standardisation process used to award A-level grades in England this year, while pupils at private schools benefited the most.

Private schools increased the proportion of students achieving top grades – A* and A – twice as much as pupils at comprehensives, official data showed.

Analysis published by the exam regulator Ofqual showed that disadvantaged pupils were most affected by its statistical model. The pattern in England appears similar to but less dramatic than in Scotland, where pupils and schools in disadvantaged areas were marked down the most harshly by the statistical model used to replace exams.

Pupils in lower socioeconomic backgrounds in England were most likely to have the grades proposed by their teachers overruled, while those in wealthier areas were less likely to be downgraded, according to the analysis.

For students from disadvantaged backgrounds on the cusp of attending higher education, more than one in 10 of those assessed as receiving C grades by their teachers had their final result lowered by at least one grade, compared with 8% for those from non-disadvantaged backgrounds.

Yeah, "GPA only" is definitely a scam for profit driven schools. It works so perfectly hand-in-glove. Pay us lots of money and we'll give you excellent education as evidenced by your high grades.

Except that is not what they were doing. The model basically said that teachers in State schools puff their students. They give them high marks in their coursework and predict their students will do well. They do not.

The private sector is more realistic. After all few students at Eton will stab their teacher if they fail.

We know a kid. Cs and Ds at the public magnet school, then As at the expensive private school. He says "because the teachers are better" but to be honest he isn't any brighter at our kitchen table.

Incentives matter.

One of the things that could be done with that is to take the sample set of algorithm predicted grades (what we originally to be used) and then teacher predicted grades (which were used after public outcry instead, despite the reality of the massive inflation effects from this), and then compare to a set of performance outcomes on university admission.

Predict it will never be done or published unless it finds the "right" result though (e.g. "Tories BAD!").

fwiw, I am in favor of simple, non-essay, tests.

An interesting 'intellectual poke' into the qualities of ambition and initiative:

How many people first viewed the Pandemic as the sabbatical they always wanted, and with no immediate dire worries of hunger or homelessness or even loss of position at work; searched out opportunity, hungry for knowledge and exploration of that which lack of time prohibited -- only to find the lure of Netflix and unread novels and political online-bar-brawls their only true adventure.

Now look at the quality of yourself:
Do you aspire and push only when the deadlines and pressures and boss-screams beckon? -or- do you squeeze every spare hour to advance your passions and investigate your blind-spots?

That is the measure of a person's contribution to ending the Great Stagnation -- which name I despise utterly -- better described as: A New World of Time Clock Watchers - the End of Task-based Work. -or- Flex-Time: how to leave work and goals to the very last minute...

Stagnation, assumes a mire or other atmosphere of stickiness and lack of momentum.. but no, we are on a hill with endless potential speed at every side, we just require that spark to an easy momentum (easy access to information, colleagues, and tech) which accelerates innovation and accomplishment. But to what end? meaningless promotion? incremental raise? barely acknowledged citation or industry award? not enough. There is no 'larger goal'. Stagnation? something to run from. Climate Change? fear to run from. Pandemic illness? low risk to cower/ cover from. Have we not learned of the decreasing power of overcoming negative and unending vague fears? The answer: Post-Scarcity. It is the smart, economic-minded's goal that answers.. let's make XX great again. Prosperity! More! Wealth! All can get behind that. Now let’s find out how to measure, plan, and push for it.

The M1 chip is not a fundamental advance in computer science: it is an important step forwards for Apple's very profitable long term strategy - see e.g. https://www.theregister.com/2020/12/07/apple_apple_apple_wip/

Vulture Central is not the sort of place that Cowen visit. Even its motto about biting feeding hands is too disturbing for him.

Interesting link, especially the end. "The M1 chip is Apple daring to make ever more money by being more Apple-y than ever. It shows with absolute clarity that Apple has no ambitions beyond that, no plans to change its world, no new ideas or new ambitions. Extreme efficiency is it. Which is fine, it’s clearly an excellent way to be very rich, and if that’s all you want from life then congratulations."

Being very rich is precisely how Cowen measures progress, at least if MR reflects his viewpoint. Leading to the last sentence being a cautionary tale about the revolution of the M1 chip. "It’s not much of a future, though. Extreme efficiency in an ecosystem invariably leads to extinction when that ecosystem changes. Enjoy the beautiful plumage of those benchmarks, but watch the sky for asteroids coming out of the clouds."

Is always tempting to see things as culminations, but the M1 chip is obviously one in a long series of incremental improvements.

American chauvinists might worry that it relies much on Taiwanese production ability and capacity.

https://en.m.wikipedia.org/wiki/TSMC

M1 is impressive, but it's just a good execution of reasonable engineering trade-offs, on TSMC's new manufacturing node. Well in line with what the silicon industry is capable of. I haven't read any odes to joy on this blog for AMD's Ryzen 2 chips, which already provide ~15 hours of generic office workloads. Ryzen 3 chips improved on that even further (15W total power draw). Manufacturing AMD's 5000U chips at 5nm node would likely result in comparable, if not better, power draw and battery life as M1.

"t's just a good execution of reasonable engineering trade-offs"
Exactly. Apple has basically just built a large smart phone and called it a computer... with processing chips optimized for the particular hardware of that machine - which Apple can do because it controls both the hardware and software in those devices.
There is no particular scientific or computing step forward involved in this at all.

What do you think the smartphone of ten or twenty years from now might be like? The rate of change from the early handhelds has been amazing, but could we be near a plateau and stagnation on smartphones? I don't really want them to do even more spying on me, for example.

What do you think the smartphone of ten or twenty years from now might be like?

My best guess would be 'pretty much like the one I have in my pocket'. My first Android smart phone was a Galaxy S2 -- released almost 10 years ago (in early 2011). My current phone is faster and has a better camera (particularly better camera software--the sensor itself is not much different) and a larger, higher-res screen, but effectively, it does all the same things as that 10-year-old model did. No significant new functions have been added.

"Exactly. Apple has basically just built a large smart phone and called it a computer... with processing chips optimized for the particular hardware of that machine - which Apple can do because it controls both the hardware and software in those devices. There is no particular scientific or computing step forward involved in this at all."

One could make this same argument about all desktops, laptops, and smart phones for the past decade. However, in the case of the M1, I disagree. In moving to their own silicon, Apple has added a neural core processor for machine learning that has no comparable functionality in an existing Intel chip (and not likely to for several years). The ability to incorporate silicon functions (e.g. a neural processor) specific to Apple for which Intel has no interest or particular skill is a significant advance IMHO. If you listen to the RISC-V folks, being able to add your own silicon functionality (vis a vis Intel or ARM) is one of their major value propositions. This is just the first generation of Apple silicon. Who knows what new capabilities the second generation will hold.

Apple has been putting the pieces into place for a very long time--15 to 20 years. During that time they've been accumulating very esoteric engineering disciplines. In 2010, they bought Intrinsity, which was an Austin-based company that specialized in making design tools that hyper-optimized internal chip layouts. That was the first real signal they were serious about growing their ambitions. There is indeed some very big innovation happening inside at the silicon level, and it manifests in the huge wins you see over Intel.

A long time ago, Apple made a VM to handle the emulation of 68K on x86. That proved very successful for apple, and showed they could effectively emulate another processor with almost no penalty for the user across their entire ecosystem. They've done it again with the M1.

Microsoft, for whatever reason, wasn't able to do the same and they forked their app story and ended up with a category of apps running on ARM, and a category of apps running on x86. And a big mess.

What Apple has done is very unique and indeed innovative. And you won't see it from other vendors for a long time (or ever) because they are all so silo'd.

The M1 is a nice chip, but the ARM architecture is seeing a lot of innovation across the board. See the Fugaku supercomputer, powered by the Fujitsu A64FX microprocessor.

The interesting thing about ARM is that it does allow per-design instruction set expansions, for anyone willing to tie software and hardware in that way.

Yes, am not sure what people expect, New laws of physics at very turn ? The M1 is certainly an advance.
It's easy to get too blase about these things. Innovations include everything from the small improvements to the large ones.
You could have said the same thing about any improvement along Moore's curve ( Big deal) but we are now much further along.

People who don't see these improvements should stay with the iPhone 1, but I am guessing they don't

A lot of Apple haters and/or people that don't understand CPU tech commenting. It's a much bigger deal for computing than 'a smartphone turned into a PC'. Intel didn't invent the transistor, but they revolutionized it's use. Unfortunately, Intel had a near monopoly on the standard for decades and have clung onto it and not invested elsewhere - classic stagnation recipe - and we're seeing others such as Apple, AMD, and nVIDIA fill the void. Jobs actually went to Intel to develop an ARM cell phone chip for the original iPhone and they turned him down because they didn't think the R&D investment would payoff in a niche product! Hence why Apple started developing their internal expertise on this in 2008ish. It's taken them a >10 yrs R&D and acquisitions to get here.

Most folks don't realize x86 architecture has essentially stagnated over that same period in large part because of power draw (plus associated heat) and needless instruction complexity. This had resulted in ever more watts to eek out tiny increases in raw processing speed. To compensate, Intel and AMD shrunk die sizes - but each shrink gets progressively more challenging. They then got the idea to add more cores and encouraging splitting of processes - which helps but doesn't generate the same performance gains because optimizing software for multiple cores is very hard. They've now reached very diminished returns here too. Don't get me wrong - x86 will be around for some time to come if nothing else to support legacy systems and software, and AMD has definitely extended x86 life with its recent die shrinks to 5nm - which has allowed them to materially boost clock speeds for the first time in ages - so much so that they are now ahead of Intel even at the high end, but the outcome will be the same. ARM will be the future for the next generation of PC and high end computing. Others will be in catch up mode to Apple for quiet sometime, but this will significantly change the game.

Don't forget graviton and neutrino detection.

I admit this is the first time I've had to somewhat realistically consider the potential that death may one day be outlawed.

There are some accounting errors with renewable energy. It's been decreasing health costs and environmental damage for years, but because externalities weren't priced in people are thinking it's only contributing to economic growth now when it has actually been doing that for years. While estimates vary, I'd say a typical kilowatt-hour of clean energy is worth at least 3 cents more than a kilowatt-hour from coal when the cost of damage from coal emissions are included.

Not just emissions, the process of mining coal itself is incredibly deadly, specially in countries like China, and then you have incidents like this:

https://en.wikipedia.org/wiki/Kingston_Fossil_Plant_coal_fly_ash_slurry_spill

But that is not relevant to Australia where coal mining is incredibly safe. One of the safest countries in the world for coal mining - and, unusually, very safe deep mining not just surface mines.

So if you want to compare Chinese coal with Chinese renewables, consider the health costs of dumping the toxic waste from cell production directly into local rivers. As is the norm in China.

We made up for careful coal mining with killer coal power stations. Hazelwood power station was so full of asbestos I expected them to put a Chernobyl reactor 4 style concrete shroud over it after it was closed down.

At under 20 cents a watt for solar modules, I'm surprised they can afford much toxic waste. At $70 a panel you wouldn't think they could be using a whole heap of chemicals to make each one, would you?

A 350 watt panel in Australia can easily generate over 13,000 kilowatt-hours over 25 years. To generate that much from coal would require over 3 tonnes of it and emit around 13 tonnes of CO2. The solar panel clearly comes out ahead on least amount of toxins released into the environment.

Yes and if only solar panels would work at night and on cloudy days. Yes I have solar panels on my house. I produce 4 times the aggregrate energy I use in total but I draw on the grid every single night and cloudy day.

You expect stagnation even as there is no stagnation. Mulltiple home battery storage systems exist which can be used to make you a night-time electricity supplier. The market is small but growing rapidly and more than one company will offer a new technology like flow batteries to compete with Tesla Powerwall which was mostly better marketing of a niche product from half a dozen UPS makers.

These products were expensive, custom, and thus expected to grow slowly as a market. Is Elon a genius for thinking economy of scale and standards would drive exponential demand growth? Tesla can't meet Powerwall demand for a profitable product. Nor can competitors.

In Australia we pay people money to supply electricity to meet demand. If you get the incentives right you can have electrical power 23 or even 24 hours a day.

I am consistently surprised that any of these rosy futures never includes the social- as if we will maintain the same structures and institutions but miraculously become better people to where even anarcho-market-communism would work splendidly if we were all just nicer (or had better incentives).

Our best ideas about government do not scale, invite corruption, and in the past few cycles have had Nobel Prize winning; suffer-the-little-children-through-drone-strikes and reality tv star at the helm. You can't even make this shit up.

And of course the means to induce the little minds away from their twitch streams and OnlyFans accounts is to increase the "stature" of scientists (which sounds suspiciously like added responsibility for no more pay or at least some cult level social engineering), because truly what motivates people is high prestige riding around in a dumpy econobox.

Krugman may be an idiot with every fiber of his being, but at least he is pointed in the right direction.

I don't see it for this generation. Yes we have some good progress in science but the translation to the market at least in Europe will be difficult. If I look at German Twitter as a basis I see hard working or entrepreneural activity is more or less hated one. 1/4th of the graduates want to work in a government job. Risk Aversity is on an all time high and socialism and Uniformity are marching through society and universities at least in Germany.

All that is not ideal to have high productivity and to translate inventions into products.

When future generations at least in North America are forced to take on huge loans to go to college and yet another even bigger mortgage to keep a roof over their head, their tolerance for risk-taking is very rationally diminished.

The only people who've been taking out huge mortgages for homes are those stuck in coastal 'zoned zones'. But Covid and WFH seem to be addressing that problem. The same may prove true for higher ed. After the pandemic, will high-priced universities be able to bring everybody back into the dorms and classrooms and restore their former pricing structures?

There is some evidence from Duflo, Banerjee et al's study of microfinance that only about 20% of the population is really entrepreneurial anyways. I'd speculate it's a form of neurodiversity, like mathematical/programming aptitude, and it's pointless to expect 100% of the population to exhibit what is a very minority trait. The best we can do is remove the obstacles to letting these productive individuals create jobs for the rest of the population. Removing overbearing regulation helps, but the single biggest factor is the glut of cheap capital.

'I see hard working or entrepreneural activity is more or less hated one'

So the hatred against BioNTech is palpable then? How about CureVac?

How is the stagnation over when most of the world is stagnating at home thanks to the pandemic?

Or maybe there is no there there

I'd count a major increase in the storage of batteries an end to the Great Stagnation. It's not just cars - enough energy density in batteries and you can have electric planes, actual flying cars, etc. It's a multiplier technology.

If they work the solid state batteries could be a real game-changer. People may buy battery cars as they are a better deal, not for any green reason.

Not sure there is a future for battery airplanes. Maybe short-hop stuff.

But everything else...

I thought flywheels had a future, but they never seem to make it.

The cost of battery storage has actually been on a very smooth learning curve throughout this supposed stagnation.

https://rameznaam.com/2014/09/30/the-learning-curve-for-energy-storage/

+1. Personal take: It's perhaps somewhat defensible for TC to claim stagnation due to relative lack of fundamentally new applied technologies in society, whether that's false or true we can test, but it is at least an argument.

However certainly you can't take such a stance and then go "Oh well, now stagnation is over because of the M1 chip, or faster vaccine cycles, or better batteries" when neither are discontinuous new applications, nor really a faster break to trend.

Lots of this sudden new techno-optimism simply seems like a mood phenomenon where technical classes suddenly feel optimistic for completely unknown reasons (objectively there is no real change to trend on big things that matter like ecology, world poverty, international power competition, surveillance, etc) and this has bled through to commentators like TC. Or this same class of persons don't actually feel any more optimistic but suddenly have decided that depicting the future as a hellscape is bad for the people they've put in power and somewhat cynically want to reorient the conversation to techno-optimism now its expedient to do so (a climate of "Don't worry, be happy" techno-optimism is deemed to be good for getting "Back to Brunch").

Will be a glorious round of large grants funding Progress Studies throughout the land.

I wonder if the pandemic reminded pessimistic Westerners that there can be happiness return to scientific progress.

But once the pandemic is pushed aside and normal life resumes, this happiness return will be less obvious.

I'd like to see a great stagnation book about political, moral and religious/existential progress. I think that's where, for us relatively affluent westerners, the returns to progress will seem most promising post pandemic.

Yes, but this is the wrong place to argue it. Tyler is now in Paul Krugman's boom mode and interested only in getting his share of Biden's rewards.

Where's the "War on Christmas" this year now that a Democrat is elected? They can't be all getting stabby with each other at "Stop the Steal" rallies can they?

in minneapolis where shootings have doubled this year, the woke have determined christmas lights are "harmful to the community"
cornpops sees "defund the police" only as a branding/political problem and dont ever ever talk about the increase in violent crime because its bad "optics"

What do these people call the parallel dimension in which they exist?

How does it feel to lose erections, Dick? Keep on losing because you are good at it. Loser.

Bless your heart.

On 6 Feb 2020, I had bilateral inguinal hernia surgeries. Now, it's stronger and approximately an inch-and-a-half longer.

Make America Poor Again, duh, Build It Back Better!

If Not "The End of the Great Stagnation," What to call it?
The lower (economics) half of Americans now own only 1% of America's assets/wealth, down from 3% in 1989 (Reagan) and have never been more in debt.
Weekly initial unemployment claims 10 Dec 2020 rose to 835,000, 730,000 was expected. Lockdowns Rule!
An eviction crisis is coming. Approximately 12,000,000 renters each owe more than $5,000 in past due rent. Just under one-third of US households are renters: US home ownership is at 67.4% 3Q2020.
Estimated US Federal tax receipts will be down $900 billion over next two years.
Already-Bankrupt blue states' tax receipts will be down 25% to 50%.
Black Lives Matter! Almost half of black businesses wiped out by lockdowns. Millions of minority students crushed by lockdowns.
Seven million more American families driven to dire poverty.
UN World Food Program Official warns that 270,000,000 people are "Marching Toward Starvation In 2021."

China Joe Didn't Win. We Were Robbed.

Or, how I learned to stop worrying and love the China Occupation Government.

DeLong might be cackling in his lair imagining that finally he will be recognized for the towering economist he really is, but Tyler? His chances with the Biden crowd are about as good as Trump's chances to override the Electoral College (an attempt that glaringly demonstrates his utter contempt for the Constitution he swore an oath to uphold).

China Joe Didn't Win. We Know What They Did @ 4AM Last November 4.

Embrace The Suck.

Georgia is clearly a hot bed of DemoRats, just like Arizona.

Who cares that Republicans are running those states? Because we all know that Republicans threw the election to Biden.

Embrace the delusion.

>Tyler is now in Paul Krugman's boom mode and interested only in getting his share of Biden's rewards.

This EB-Ch guy is very, very perceptive.

Thank you for all you contribute to this comment section. You make it worthwhile.

Past performance is no guaranty of future results.

When Krugman and assorted infallibly ignorant academics say, "Buy!" Sell - it worked for me.

always looking of an edge.

Tyler is a god but like everyone here in the comments he misses the point on this one. Elon did point this out at Springer a couple of days ago about when he expects AI (pronounced DeepMind) to reach human level intelligence. He said 2025. Kurzweil of course says “2029 but probably sooner.” With innovation like the protein folding coming every month, every day you can say goodbye to the Great Stagnation in all fields except ironically space travel, where ordinary humans travel slow no matter the intelligence. Hopefully the new mind can find a workaround.

Space travel will probably follow the precedent set by aviation, in which the innovations come from safety and cost reductions rather than increased speed. Currently, the cost of putting something into low earth orbit is about $10,000 per pound. SpaceX's advertises launches at $2,500 per pound. If it can make a profit, that is a real improvement over the status quo. And if it can do so while also not blowing up 1 in 100 payloads, even better.

Regardless of whether there was ever a Great Stagnation (slowing of technological progress relative to trend), the AI trend seems fairly true. https://www.gwern.net/newsletter/2020/05 - When? Estimates of Moore’s law-like progress curves decades ago by pioneers like Hans Moravec indicated that it would take until the 2010s for the sufficiently-cheap compute for tiny insect-level prototype systems to be available, and the 2020s for the first sub-human systems to become feasible, and these forecasts are holding up.

Not sure I would trust Elon's or Kurzweil's timelines. Their track record on hitting schedules is not good. Elon eventually delivers at least. Can't say the same about Kurzweil.

Well Google, the best out there in AI, hired Kurzweil with the stated purpose of him getting the machine to understand natural language. That’s some serious credibility whatever you may say about the accuracy of his predictions which I say are spot on so far.

Kurzweil said life expectancy would exceed a hundred years by 2019.

Risk taking is the oxygen of all development.
When bank regulators introduced risk weighted bank capital requirements foolishly based on that what was perceived as risky is more dangerous to our bank system than what’s perceived as safe, they denied our economies that oxygen
Result? The buildup of hugely dangerous bank exposures to what’s perceived as safe and stagnation.
http://teawithft.blogspot.com/2009/10/please-free-us-from-imprudent-risk.html

Of course, it was (the obsession with) the internet that caused the great stagnation. To now claim that the internet will lead us out of the great stagnation takes a flexible mind, and indeed Cowen has a flexible mind. We have been held hostage by the internet for at least a decade and a half, while it absorbed mass amounts of both investment capital and human capital. Easy money for the hostage takers but little reward for the greater economy. The hostage takers had many accomplices, just as Trump has many accomplices as he tries to destroy American democracy: Trump is not alone in his quest, and neither were the hostage takers, the Fed helping feed the internet frenzy. But just as Trump and his accomplices continue with their quest, so do the hostage takers, this past week's high flying IPOs by Airbnb and DoorDash proof that the hostage takers will not go quietly just as Trump and his accomplices will not go quietly. If Airbnb and DoorDash mark the end of the great stagnation, then so do the years 1958 to 1962 in China. One with a flexible mind must believe it takes several steps backward to proceed forward to see the silver lining in the dark cloud that hovers over America. And it takes great men to lead the way, the innovators who have not been given their due, who have borne the brunt of the criticism for the great stagnation they believed was a figment of the imagination, who will now experience redemption from none other than the man who cast a pall over their efforts in 2011. Airbnb and DoorDash, leading the way for America's great leap forward. It's 1958 all over again.

Aha. Now I understand the tantrum directed at FDA.

You were trying to win a bet.

TC's list of advances doesn't seem to line up with his claim later in the post that figuring out effective use of the Internet would be the critical item to end stagnation. Other than instant collaboration and access to information, the Internet itself has little to do with the high tech advances behind batteries and electric cars, space and rocketry, CRISPR, and so on.

Sure, the Internet has had gigantic impact throughout society, and that integration continues to grow rapidly. It is awesome to behold. Is it really the engine behind ending stagnation? Does near-immediate delivery of goods to my door through the wonders of the modern supply chain and Internet really matter to growth? Is it all those cool engineering jobs that count?

Indeed. Fair question worth drilling into.

I am also curious if the lockdowns and work at home unleashed any breakthroughs that had been stymied until key folks were locked in with nothing better to do and less meetings to attend.

This definitely happened at my company. Would love to see this story come out if it was more widespread.

There was no great stagnation, but ..

In terms of your questions, it is directly a problem of computation to allow rockets to land on their feet. More than that, audacious confidence in computation.

You mentioned crispr, that too is a direct result of computation. There is no way in hell the Human Genome Project could have been run with test tubes and pencils and paper.

We get complex genetic answers in hours because the machinery and the computation are good. All developed in a smooth learning curve since 1990.

That's the reason the coronavirus could be sequenced so fast.

Rapid computing no doubt is huge.

I suppose we can give the internet credit for that.

To a technologist they are pretty much synonymous, racks of ever denser computation and storage, connected by ever higher bandwidth infrastructure.

"Computation" is not the same as "the Internet." There is no question that computing power and related technology such as storage capacity and density have been absolutely essential to so many things. However, that is not the same as TC's claim about smart Internet usage as being the essence of ending stagnation. The computing power also underlies the ability of the Internet to succeed and be pervasive.

Okay.

I reacted the way I did because so many stagnationists do group it all together, "it's stagnation if you don't count computers and the internet!"

In retrospect, I wonder if the massive outperformance of tech companies relative to other stocks over the last couple of years which went against many investing rules of thumb before (small over large, value over growth, look at CAPE, international diversification, etc.) was a harbinger of this new level of tech growth.

Cowen has emphasized stagnation in total factor productivity when discussing the great stagnation. The graph at the link shows TFP increased from 1955 to 1965, was stagnant from 1966 to 1984, increased from 1984 to 2005, again stagnant from 2005 to 2013 and increased from 2013 through 2016 where the FRED Graph ends.

https://fred.stlouisfed.org/series/RTFPNAUSA632NRUG

The section on tfp was strangely devoid of links. The following does not seem to agree with what Tyler said:
https://www.frbsf.org/economic-research/indicators-data/total-factor-productivity-tfp/

When Eth2.0 runs on our smart phones.
When that happens then me and any online seller can engage in a complex transaction, guaranteed to completion with provable exits having congestion pricing. No third party really needed. When that happens then we get real time search engine price takers. Goods around the world will suddenly have less inventory volatility a the bots automatically balance using flexible contract payments over short periods.

It is like having a humongous Google protein folding computer running in real time, maintaining the best economic 'folding pattern'.

What might an end to the Great media Stagnation consist of?
- a 50 million american dollar defamation lawsuit against the
cnn.hemorroids

Looks like in the debate between Brynjolfsson and Cowen,
Brynjolfsson wins.
https://techonomy.com/conf/11-tucson/jobs/can-technology-be-societys-economic-engine-a-debate/
Time to move on. Being contrarian may get you headlines, and a book, but over time, it all becomes clear and indisputable.

"I also wrote, in 2011, that as the great stagnation approaches its end, we will all be deeply upset, and long for the earlier times. That too is by no means obviously wrong.

Jeremiads are rarely accepted in their time. However hopeful or optimistically framed.

3 necessary conditions to fail to see Great Stagnation:

1. Being an American who has never lived outside the US for at least half a year.
2. Being not an independent thinker.
3. Having complete lack of imagination and never reading science fiction.

1. Americans and everyone in the world really are having a hard time living outside their borders this year.
2. No thinker is truly independent.
3. Nobody truly lacks an imagination though I agree we can all read more science fiction.
4. There is no Great Stagnation. There has always been a technological progression and it is not good to exaggerate one way or the other for effect.

Are the “next leap forward” innovations Gaussian or fat tailed?

We seem to struggle reading even Gaussian small samples - way to quick to declare certainty. Fat tails?

AirBnB just had an amazing IPO. There is NO Great Stagnation because I can get a room using my phone. That has never happened before in the history of the human race. I'm completely in awe.

You know hotels have apps, right? I can book a room, check in, open my room, and check out all on my phone without ever talking to a human being. That is not AirBnB's unique bailiwick.

Well, one thing that is not stagnating is
Covid deaths in Sweden:

"Finland and Norway have offered medical assistance to Sweden, as their neighbour faces an increasingly severe second wave of coronavirus that has stretched clinical staff and intensive care capacity to the limit in some areas....Sweden has reported 1,400 Covid deaths in the past month compared with about 100 in Norway and 80 in Finland, each of which have half its population. Sweden reports deaths differently to the rest of Europe, which makes knowing their exact level difficult, but analysts at Nordea, the Nordic lender, believe they could be higher than their peak in April.

The region of Stockholm on Wednesday warned that 99 per cent of its intensive care beds were full, while healthcare unions have warned of large numbers of workers quitting as infections, hospitalisations and deaths continue to rise. About 3,600 health personnel have quit in the Stockholm region since the start of the pandemic, according to state broadcaster SVT, around 900 more than over the same period last year."
https://www.ft.com/content/81299a7a-ec53-41fa-bf33-7ea40f4bca07

Have you ever noticed that this, and other hot topics of debate, disappear when new, contrary, information reveals itself?

A debate is never over if there is new information; it just appears over when it is contrary to the original claims, and then it disappears.

Just like the Great Stagnation.

George Orwell had something to say about the rewriting, or disappearance, of history.

Covid-19 deaths per capita:

1. Belgium
2. Peru
3. Spain
4. Italy

7. U.K.
8. Slovenia
9. U.S.

23. Sweden

"You've heard it countless times: Sweden kept its COVID-19 curve flat without lockdowns or other major restrictions, and without damaging its economy. The U.S. and other countries should follow its lead, we're told.

Baloney.

Let's take a look at those claims, along with how Sweden really did in comparison with the U.S. and other nations:

"It kept its death rate down." Not really, no matter how you look at it. Case fatality rate? 6.1%, versus 2.8% in the U.S. (the White House likes this statistic because it makes us look relatively good). But that's a poor metric for international comparisons, since it's driven not only by cause-of-death identification, but also by policies on testing that vary markedly by country. How about COVID-19 deaths per million population? It's 582 in Sweden, not much better than the 596 in Italy, despite that country's early horror, and only somewhat better than the U.S.'s 650. Sweden has the world's 14th worst rate of COVID deaths (out of 214) on the basis of population. Sweden's neighbors, meanwhile, imposed more restrictions on people and businesses and their death rates have been far lower."

Here is the link: https://www.medpagetoday.com/infectiousdisease/covid19/89017

Your numbers are out of date. Check Worldometers today and sort on deaths per M. Sweden is 25th . If you remove Andorra , Montenegro, San Marino ( small countries) , it's 22.
Deaths per M:
Sweden : 742
Italy: 1068
Spain: 1018
US: 920
France : 884

Let's look at Sweden compared to its nearest neighbors, because covid started at different times in Europe (eg. Italy Spain)

So, let's look at deaths per million from your source:

Sweden 742
Norway 71
Finland 82
Denmark 16
Germany 265

Cat, Do you dispute that you should be looking at deaths per million with respect to nearest neighbor, and, if so, why?

Also, Cat, if you maintain that Belarus health care system is comparable more to Norway or Sweden, please say so. Some of these countries not only had an infection earlier, when it got out of control, but they also have non comparable health systems. You know that. Whereas the Nordic have similar systems and similar start times.

@Bill, I merely pointed out that your numbers were old. Yes Sweden doesn't look good compared to their immediate neighbors. If they had never said " herd immunity" I don't think anyone would care.

They fared badly early on with their older seniors because they didn't treat them. ( palliative care only). This was a mistake. I believe had they done that they would look like Switzerland.

Herd immunity even at 25%, takes a while to get there , at least 1000 deaths per M, the US isn't even there yet. Sweden is probably at most 17% past infected as a whole, I am not sure

The reason people care about Sweden anyway is ideological . They have no real concern one way or the other. Some people want lockdowns and restrictions and some people don’t. It’s become an ideological issue. State control vs individual freedom.

Sweden used voluntary restrictions. Many countries had a mix of approaches and fared worse. That's the case of Argentina and Buenos Aires, months of lockdowns/ half assed lockdowns and still a disaster.

Promoted from below the fold so people can see the reply to Cat:

You not only pointed out the numbers were old, but put in a list of numbers that really are not representative, for the reasons you identified.
I was also responding to Todd as well.

I am troubled by the words lockdown, and, as we live in communities, we don't have the luxury of "individual freedom" when we have all signed on to a social contract, one which allows us to vote for our representatives, who are entrusted to act in a way that maximizes total welfare. And, if they don't the get voted out.

We both agree that we do not need to politicize public health, for our own health and those with whom we live.

Todd K
2020-12-13 17:32:36
#
You think lockdowns, which have caused mass mental suffering as well as people not able to get critical care, is maximizing total welfare?

Also, Todd, Sweden has since November been imposing controls: "

On Monday, it announced that public gatherings of more than eight people were no longer allowed, marking a shift in tone for the Scandinavian country that has so far largely relied on voluntary measures and guidance during the pandemic.

Prime Minister Stefan Lofven announced the new limit on gatherings, lowered from a previous limit of 50 people (or 300 people for some cultural or sporting events), signaling a more aggressive approach to containing the spread of infections.

“This is the new norm for the entire society,” Lofven told a news conference, Reuters reported. “Don’t go to gyms, don’t go to libraries, don’t host dinners. Cancel.” https://www.cnbc.com/2020/11/17/sweden-toughens-up-coronavirus-rules-as-infections-and-deaths-rise.html

That is not remotely a lockdown nor a U turn in policy. People can still go to libraries in gym in Sweden as well as host dinners.

Todd, Did you read the materials. They are being told not to do those things . They also die at a level not seen by their neighbors. For what? Look at the neighbors, look at the numbers.

Tyler Cowen in this mode is just a bad pundit trying to cover his tracks. Such guys are a dime a dozen. Tyler only distinguishes himself in that he’s notably pompous while doing so.

The thing I like about this blog is how it constantly reminds me how much work there is still to be done.

The end of the Great Stagnation will not be found in technological advancements or some measurement of financial accounting. It will only be realized from within the human heart.

And from within George Church's laboratory.

The human imagination.

Tyler and 99% of conservatives consider anything requiring more than a hundred cooperating people impossible.

If not built by people with imagination willing to organ by the millions, the US would not have the railroad system which was state of the art 1930, the highways which were state of the art 1970, the telecommunications system which was state of the art 2000, etc.

Conservatives have blocked collective action so the US lags various nations with vision to use collective action to lead in many technology areas.

I worked on fiber optic futures in the 80s, mentored by a former Bell Lab engineer. By 1990, he had convinced me dark fiber everywhere was not only feasible, but inevitable.

The stagnationists marshaled their forces in the 90s so they could kill fiber to the home, because the Comcast monopoly profits had to be protected.

Meanwhile Korea, China, Vietnam, et al are building fiber optic to every address, while the US bases everything kon high speed internet which is denied half the population. Given schools depend on high speed internet, the public schools are now responsible for building high enough speed internet everywhere.

Maybe the Great Stagnation won't end until
economists have less influence on public policy.
Has anything done more harm to productivity
than the Milton Friedman doctrine that a firm's
main responsibility is to its shareholders.

yep. the viral pandemic has done beaucoup more to harm productivity than friedmans premise

I only scanned the ADS link briefly and saw no mention whatsoever of two huge, global technical problems that remain to be addressed ASAP:

1) while the chances of a coronal mass ejection directly hitting the Earth-Moon system (with or without numerous Z-class flares accompanying) on any one day, in any one week, in any one month, in any one year, or in any one decade are always remote or even less likely, the plain fact that a large and significant and HUGE CME event could strike the Earth-Moon system at any time (and with no more than three hours warning) would undo practically every technical achievement that Progress Studies cheerleaders are pleased to crow about. EVERYTHING would be instantly undone, because all electronic communications systems and all electronic data storage systems would be perfectly wiped out and thereafter perfectly useless.
--so who is even studying (much less devising hardware or software) what would permit a crippled humanity and a trashed tech infrastructure to recover from such a momentous event? (The Carrington Event occurred as far back as roughly the mid-19th century: so our planetary system is statistically already decades overdue for an event comparable, either larger or smaller, and the outcomes this time would far exceed telegraph poles spontaneously combusting).

2) my quick scan of the ADS summary TC linked to showed no mention at all of the immense challenges lurking on all of our horizons concerning the inexorable advent of Technogenic Climate Change. 2020 has given the briefest of pauses to (further) technogenic compromises to climatological and meteorological systems due to the minimization of human and technological activities on the usual scale: but plastics continue to degrade into befouling microplastics in our oceans and seas by tons and tons, glaciers and ice sheets continue to melt with reckless abandon in both polar regions and atop mountain chains in all hemispheres--and now the Chinese want to begin micromanaging climatological and microclimate systems to insure rainfall on Chinese topographies, something that their AI and quantum computing aspirations may lead them towards, should no CME event occur in the interim.

These are only TWO huge global technical challenges that no cheerleader of Progress Studies seems to be addressing, at least not Tyler Cowen.

From my understanding of geomagetic storms, the most significant threats of a Carrington-level event would be power grid and satellites. A smaller event did quite a bit of damage in Quebec in the late 80's, and I think that they've built in some protections to the power grid since then. Essentially, the grid would have to be shut off for a while, but you can avoid most of the permanent damage. Satellites are a little trickier, and could still be a very large loss. There's definitely quite a bit of research in better prediction and modeling, but I'm not as sure about the efforts on ameliorating damage.

I still think the biggest technological advance of the past 10 years has been the removal of wires from headphones.

The Chinese fusion thing is not really novel. They're getting the experiment running, which promises data, but it's no more advanced than ITER, and profitable fusion is still at least 20 years away.

The Economist article suggested that our ‘intangible investments’ might strike a blow against overpriced housing, education and health. The pandemic has clearly demonstrated end runs that have opened opportunities in all three: online working, online classes and online medical teamwork. But how does that impact the number of job openings, and will we ever see a market for high-paid labor? Might stagnation recovery spell increased productivity in some sectors but not for labor?

What is the The Great Stagnation?
(do I need to buy a book to find out?)
(...how very 'American' of you...)

So...when are we going to get a workhour reduction with all of these new gains in productive capacity? Our current 40-hour-a-week regimen (ha! 50-60 is more the norm these days in the US) was the pipe dream of the Welsh industrialist and labor reformer Robert Owen in the early 1800s and the subject of severe state repression in the labor struggles in the United States throughout the 19th and 20th--just so working folks could have a decent worklife balance. It was not until the 1930s that it was enshrined into law in the US--it had in Europe far earlier. And consider the technological progress that has occurred in nearly the last 100 years. Consider just word processing to say nothing of communications. If I make a mistake on a document, I don't need to pull out the White-Out. To say nothing of the ease and speed of communication.

At some point, these "innovations"--I saw an advertisement a while back for a fridge you could send Snapchats on from Samsung--are a mess of pottage. I'll turn my own damn lights off--I don't need Alexa for that.

Make Thursday the new Friday. Three day weekends, every weekend.

Practical fusion power.

Its good as your other posts : D, regards for posting . "Reason is the substance of the universe. The design of the world is absolutely rational." by Georg Wilhelm Friedrich Hegel.

Gee, and I thought it was caused by government anti-business policies.

Well, Cowen said he expected the Great Stagnation would be over by the 2040s so great that it is happening twenty years early. Of course, almost no science major in 2011 thought there was a great stagnation.

I'd like to see a review by a balanced panel of academics outside econ on whether there was ever a stagnation in their area. It seems wrong in computing and bio examples TC gives, while potentially being true in energy, but that might just be because I don't know much about energy, while I feel (perhaps in delusion) that I do know a enough about about computing and bio to think there's no step change (acceleration).

I agree. For energy we had fracking take off in 2008 and the US flooding the world with cheap oil and natural gas. That seems significant

Fracking is an old technology.

"While there are plenty of misconceptions that surround this modern technology (it poisonous substances the drinking water, or it creates cancer), the largest is that it's actually a recent innovation.
The Real History Of Fracking

Surprisingly, fracking can be traced back to 1862. It was during the battle of Fredericksburg VA., where Colonel Edward A.L. Roberts discovered something incredible when firing explosive artillery into a narrow canal that was blocking the battleground. The breakthrough was then described as 'superincumbent fluid tamping.'

On April 26, 1865, Edward Roberts obtained his very first patent, for an “Improvement” in exploding torpedoes in artesian wells. In November of 1866, Edward Roberts was awarded patient number 59,936, known as the “Exploding Torpedo.”

This removal method was executed by packing a torpedo in an iron case that contained 15-20 pounds of powder. The case was then dropped into an oil well, at a spot nearest to the oil. From there, they would blow up the torpedo by linking the top of the covering with wire to the surface area and then loading the borehole with water.

This creation boosted oil production by 1200 percent from certain wells within a week. Additionally, this new type of extraction led to the founding of Roberts Petroleum Torpedo Company, which charged $100-$200 dollars per rocket, plus a royalty of 1/15 of the profits generated from the product.
The Birth Of Modern Fracking

The very first improvement on fracking didn't occur until the 1930s, when drillers used a non-explosive liquid substitute called acid, instead of nitroglycerin. This innovation made wells much more resistant to closing, boosting productivity significantly.

Although fracking was technically born in the 1860s, the birth of modern hydraulic fracturing began almost 90 years later. In 1947, Floyd Farris of Stanolind Oil and Gas began to study the relationship between oil and gas production output, and the amount of pressurized treatment being used on each well.

This research resulted in the first experiment involving hydraulic fracturing, which occurred in Kansas in 1947. In this trial, 1,000 gallons of gelled gasoline and sand were injected into a gas-producing limestone formation with a deepness of 2,400 feet. This was then followed by an injection of a gel breaker. While this experiment failed to produce a significant production increase, it did mark the beginning of what is now known as fracking.

Despite the failure in the Kansa experiment, the research proceeded. On March 17, 1949, Halliburton performed two commercial experiments; one in Oklahoma and one more in Texas. These outcomes were far more successful.

oilprice.com/Energy/Crude-Oil/The-Real-History-Of-Fracking.html

" ...while potentially being true in energy, but that might just be because I don't know much about energy"

This is what I thought when Peter Theil lectured Eric Schmidt in a debate in 2012 when Theil said: ".... we've had no progress on energy [from 1973 to 2012]. You've had sort of a catastrophic failure in energy innovation, and it's been basically offset by computer innovation." Theil just didn't understand what has happened with energy over those 40 years.

Bet he wished he stuck a few dollars into Vestas around 1993 or so.

Though being the world's largest wind turbine company might not attract someone like Thiel.

Fracking might be old, but it did indeed take off only in the last decade or two as Keith asserted. Part of that is new tech, part of that is regulatory. But it all culminates in reduced CO2 intensity and cleaner air.

Why not talk about tar sands too? That technology is at least mainly from the 1980s, and it too took off when oil prices were high enough.

Fracking only makes economic sense when you run out of the cheap oil and gas to drill for. Just like tar sands.

The real change was the combination of fracking with horizontal drilling, enabling horizontal laterals thousands of feet long through the productive zones, rather than one vertical wellbore.

Fracking technology can be directly applied to geothermal, a very clean technology with the highest capacity factor of all electricity generation.
I think this would be a significant advance.

"Fracking is an old technology."

One the one hand yes, around 1980 I worked for a consulting firm's proposal that utilized research into secondary and tertiary extraction methods. Those techniques have been around a long time (and helped convince me that while oil can become more scarce, we're not going to "run out of oil") but were not cost effective for widespread use.

Higher prices were obviously a key driver of the expansion in the use of fracking and other extraction techniques. What I don't know is what role technological change played. We now have better ability to remotely examine underground geology, I presume that we have better technology to drill wells accurately and to deal with adverse unexpected conditions, etc.

I.e. for sure we have progressed from the days of dropping exploding torpedoes down wells. What I don't know is how important that technological progress was, compared to good old fashioned price increases making fracking and other techniques economically relevant.

Cowen also predicted in 2011 that there would be no significant medical breakthroughs before 2030. I'm quite sure as I was then that the 2020s would show that this was wrong.

For the sake of deeming a prediction true or false we have to define "significant" in a formal way. One people might be saved by even incremental improvement in cancer treatment while others might not consider gene therapies and organ transplant since they cannot be applied widely. There were many cool developments in the last 10 years but how significant are they on the grand scheme of things ?

Those of us steeped in science and technology of all kinds never saw a Great Stagnation even in those years where this blog and others have claimed so (mRNA vaccines weren't invented this year like the media makes you believe). I understand you want to sell books and appear important to your very important friends but the rest of us put our heads down and continued to march the ball downfield. Notice that most of TC's list are things that were done or worked on years ago. A better listing of actual science discoveries by the year (as this blog recommends, read more Wikipedia):

https://en.wikipedia.org/wiki/List_of_years_in_science#2000s

BioNTech, to give one very concrete example, started out working on an individualized mRNA cure for cancer back in 2008, before the Great Stagnation was ever published.

And 12 years later, how is that working out for them? Cured cancer have they?

Keeping in mind the Szilard letter was written on Jul 12, 1939. Six years before Hiroshima. Between the first government committee and Hiroshima was four years. So how are they doing?

Kennedy pledged a man on the Moon in May 1961. So they have had 150% of the Apollo program.

Biontech has not cured cancer. However, the company took less than a year to develop and manufacture a vaccine for SARS-CoV2.

Considering the results of that success, it is likely its ambitions to develop cost effective cancer treatments that are individually targeted has just taken a giant step forward now that a large investment in the necessary manufacturing supply chain has occurred. Considering that between the discovery of pencillin 1928 and its mass production starting around 1944, Biontech still has 4 years to go in making a major contribution to treating cancer.

It took Pfizer two days to develop theirs. This is clearly not rocket science.

Although the RNA technology looks good.

However you are not being fair to the penicillin story. It was discovered in 1928 and first used to cure patients in 1930. Then for some reason it sat on the shelf until Florey's team tried again in 1939. Their first success was treating a policeman in 1941 even though he died. They took some to America in that year and over the next three years Pfizer came up with a method of mass production.

So it was not twelve years. It was just several 2.3 years steps.

You mean Moderna. Second, working something out on a screen and actually going into mass production are two very different things. Was the penicillin example not clear enough?

The mysterious reason why billions of pounds/dollars were not thrown at penicillin production until 1939 seems to be the sort of thing hard to forget. It was called WWII.

And the effort involved considerably more than Pfizer. Wiki is definitely your friend - "On March 14, 1942, the first patient was treated for streptococcal sepsis with US-made penicillin produced by Merck & Co. Half of the total supply produced at the time was used on that one patient, Anne Miller. By June 1942, just enough US penicillin was available to treat ten patients. In July 1943, the War Production Board drew up a plan for the mass distribution of penicillin stocks to Allied troops fighting in Europe. The results of fermentation research on corn steep liquor at the Northern Regional Research Laboratory at Peoria, Illinois, allowed the United States to produce 2.3 million doses in time for the invasion of Normandy in the spring of 1944. After a worldwide search in 1943, a mouldy cantaloupe in a Peoria, Illinois market was found to contain the best strain of mould for production using the corn steep liquor process. Pfizer scientist Jasper H. Kane suggested using a deep-tank fermentation method for producing large quantities of pharmaceutical-grade penicillin. Large-scale production resulted from the development of a deep-tank fermentation plant by chemical engineer Margaret Hutchinson Rousseau. As a direct result of the war and the War Production Board, by June 1945, over 646 billion units per year were being produced."

Though considering you possess a certain aversion to crediting women involved in science, maybe wiki is not your friend.

And Moderna's own turn key GMP-certified manufacturing facility able to produce a million doses a day magically appears.

+1. This blog is weak on science because it is aimed at other economists who lack the background. It badly shows.

All blogs are weak on science.

In any given modern population, no matter how you define the population - residents of Shanghai, residents of San Jose, graduates of Yale, graduates of MIT, Nobel prize winners to include winners of the Swedish banker prize, rich Bitcoin bros, supremely gifted interpeters of Finnegans Wake ---- once you expand to about 500-1,000 your accepted definition of the population in any of those groups, the proportional number of people who understand what science is (as discussed by Medawar and Popper and Chandrasekhar, if you are looking for cheap paperbacks from the 80s on Amazon, and as discussed elsewhere at length, you can look it up, I am not your research assistant) is going to be less than one in a hundred, every single time.

There are some exceptions - more than half the people who are doing up-to-date research on the Riemann hypothesis probably understand science, and, to be kind to biologists, probably about half of the hundred people in the world who spend the most time working on what is misleadingly called "protein folding" understand science, and I guess, if you think about barbers and Bertrand Russell a lot, in that fanboy way lots of people do, the class of "people who understand science" is going to include mostly, at a minimum, people who understand science.

I have read lots of really good science blogs, and none of them, except the ones with extremely limited focus, are really all that good.

But what do I know?

to be fair, this afternoon on the subway, a homeless person said to me "you look like one of those rich Bitcoin bros" and I said, "so do you, I am gonna think some more about this" .... and I also said "Today is the day the Lord has made, let us rejoice and be glad" ---- which I say alot, although I only say it to homeless people when they seem like they are not gonna get offended, which this guy certainly was not

maybe I should start a blog on protein folding, I am actually very interested in that specific subject.

SOME OF MY BEST FRIENDS ARE PROTEINS, sort of.

I think my first blog entry shall describe the difference between protons and proteins. I think that would be a useful place to start, here and now in 2020 ( a better year than 2019 in so many ways, by the way).

some of the good ones, written by non-economists - Natalie Wolchover on anything, Chad Orzel (or Matt Strassler??? - they are both good), particularly on imaging of extra-galactic black holes, Wilczek (just for the pictures) , Motl when he is describing developments in HEP from anytime before 2007, and me on the BPD studies of self-aware tracings of interconnected low-energy but not zero-energy real-world PDEs and the resulting implications for the triviality/non-triviality distinction after exponential but finite reiterations (nothing to do with quaternions or big O, more like functional deltas (that phrase would be clearer in the original Latin - but my volumes of Liebniz and Newton are miles away, not in the next room) mapped out with at most unitary attributes from the previous iteration of the relevant derivative - (obviously anything more than a unitary attribute would be non-trivial after the first reiteration!!!, as you would know if you were paying attention). (Thanks for reading).

"Notice that most of TC's list are things that were done or worked on years ago. A better listing of actual science discoveries by the year (as this blog recommends, read more Wikipedia):"

Right, stagnation is very much in the eye of the beholder.

To his credit, Tyler does offer an objective, reasonably verifiable measure of stagnation vs progress: total factor productivity. It's not the only way to measure progress but it's better than trying to categorize discoveries as major vs minor, or trying to categorize their contribution as happening now vs having been made years ago.

I'm not aware of the shifts or data that cause Tyler to confidently state that TFP is rising. But if indeed it does rise, or show a change from the so-called years of stagnation, that's at least one bit of evidence.

"ongoing advances in space exploration, most of all from SpaceX"

You forgot that China put their flag on the moon a few weeks back. They can do space cheaper than America's bloated public sector and expensive private sector. A huge cost advantage means cheaper and faster experimentation. India is another one to watch here.

"You forgot that China put their flag on the moon a few weeks back. They can do space cheaper than America's bloated public sector and expensive private sector. A huge cost advantage means cheaper and faster experimentation."

China is not cheaper. However China is spending orders of magnitude more while the US spending is less over the past half century.

And US government policy makers chose to replicate shift jet bomber to jet passenger planes, forcing the creation of ULA to bid fixed price. Which opened the way for SpaceX which government providing easily half the risk capital plus building industry standards to promote multiple private competitors.

ULA was not alone as Europe, Russia,..., were providing fixed price services, but not really innovating because they saw a limited market.

Elon Musk saw a rocket market of hundreds of rocket launches per day.

When you expect stagnation, which US conservatives do, why would anything but stagnation get the investment.

Tyler has been pretty much completely ignoring what the Chinese have been doing on the Moon.

"Technological stagnation" sells to one group, mostly non-technical people, while turning off another group, mostly technologists.

Interesting?

I meant this as a reply to Biologist, above.

On the subject of China and the Moon .. something something "Always has been."