Samir Varma on the forthcoming tech breakthroughs

From my email:

We are now starting to get a hint of the future transformative technologies that you guessed were on their way in “The Great Stagnation”. You had not speculated on what they might be, but there are faint hints on what is likely to happen.

I believe this article is one leg: extremely fast air travel. The second leg is the Hyperloop and similar: extremely fast ground travel. The third leg is synthetic biology (e.g: The fourth leg is quantum computing, which is finally starting to show that it might work. And the fifth, and final leg, is fusion energy, which looks eerily like it will actually come to fruition this time.
Put those 5 together and you have the makings of a new economy, with a huge burst of growth to come for many decades. These are just faint hints, of course, but they’re starting to get increasingly clear.


I don't believe transportation will get much faster.

Transportation is actually the most believable on that list of impossibles.

I don't believe that faster transportation will be transformative. Nor will it be affordable, or compatible with the coming climate austerity.

Agreed. Faster trains won't be viable economically in NA for all the same reasons Amtrak loses money every year and the same reasons Asia and Europe have high speed rail but NA doesn't. Faster air travel could happen on a very small scale but that's old news, we already had the Concorde and it failed as a business for predictable reasons.

As for the others: quantum computing, sure, it'll change some stuff at the margins. Synthetic biology, sure, this is likely to be big and probably terrifying. As for cold fusion, LOL, thanks for including this to tip us all off to the fact that this email is satire. Next year in Jerusalem, as they say. Thy Kingdom Come.

As for Hyperloop specifically: when I was eight I drew a picture of my sweet-ass fantasy house with two swimming pools, a moat to keep out zombies, a laser gun turret on the roof for shooting enemy spaceships, and a zipline to take me directly from my bedroom to school. Somehow this idea failed to attract tens of millions of dollars of venture capital. If only my name had been Elon Musk.

American made planes fall from the sky killing thousands. Such is life in Trump's America. Brazil's Embraer makes better planes than sad Boeing.

Talk about recency bias...

...........and Brazil bias.

Not at all. But evidently a Brazilian company wouldn't ever be involved in such a scandal. In America, greed is good.

TR said: "...a Brazilian company wouldn't ever be involved in such a scandal..."

I thought Petrobras paid nearly a billion dollars late last year to settle a corruption scandal. Operation car wash?

Evidently, there were deviations and mistakes, but most of them were the work of corrupt politicians under the orders of corrupt, leftist king-maker, former President Mr. Silva, who is behind bars (has Trump fulfilled his vow of "locking her up"). I don't believe a private company would ever do something like that. President Captain Bolsonaro has ordered a very strict federal investigation on corrupt deals.

So Brazilian companies " wouldn't ever be involved in such a scandal. " but actual Brazilians would be. Duly noted.

Left-wing corrupt politicians working for international communists. I can not imagine Brazilian entrepreneurs commiting such crimes. In Brazil, greed is not considered good.

A dam built by Vale in Brumadinho burst killing nearly 300, and injuring hundreds more. Last year, Vale admitted in an internal document that the dam was weak and that its collapse would cause a large loss of life.

Sure, corrupt politicians etc. If I nod and pretend to agree with you, will you give up?

There are incidents, everyone knows incidents happen and authorities can make bad judments, but it is not a systematic practice in Brazil as it is, undoubtedly, in the USA.

Wasn't Embraer making a deal with Boeing? And didn't they warn of little or no profits for the next two years? Such is life in President Bolsonaro's Brazil.

Embraer closed a deal with Boeing, but I am sure Brazilian executives will never compromisse their moral values. I can not imagine a Brazilian company behaving the Boeing has behaved. More than one hundred human lives were lost because Boeing was too greedy to recall a defective model. Such is life in Trump's America. Sad.

Greed is a human thing, not a nationality thing. Check this out and let me know where you find Brazil. Lower ranked than India, China, and Bosnia Herzegovina. SMH

"Greed is a human thing, not a nationality thing. "
So are cruelty, crassness, violence, etc. Some peoples value those things, other peoples fight them. Brazilians don't revere greed as Americans do. Quite the opposite. President Captain Bolsonaro is a former militaryman who was awared a medal for saving a soldier under his command whereas American president Donald Trump is a famous conman.

Brazilians do not revere greedy the way Americans do. Quite the opposite, indeed.

The takeaway is that hypersonic long-distance transportation may be more practical than supersonic. Although one key here is "long-distance," as subsonic seems fast enough for journeys up to 3,000 miles or so.

But, it's also possible that hypersonic may be limited to long-range weapons, as their flight path would be far less predictable than that of a post-boost ballistic missile. And weapons are not subject to the same cost and safety constraints as commercial travel.

Hyperloop seems improbable as it's too costly, and far too many passenger-safety issues. To make the economics work at all you'd have to have a very large number of travellers. And you'd have to have a way to get people out of the tunnel quickly, if/when something goes wrong.

So, the US needs more people to make HSR possible like in Japan, China, Europe, but that's impossible because we are already too full?

And paying workers to build capital costs too much and high costs kill jobs and destroys wealth?

The GOP promoted building high speed transport starting with the Civil War and 50 years later the US had so many competing rail lines, conspiring to charge high prices to rural businesses and farmers the GOP invented anti trust laws to centrally plan pricing and competition (ICC). Then too much debt drove bankruptcy when the economy contracted circa 1910, leading a total government takeover of the railroads in WWI so government central planners and government cash could eliminate wasteful competition, plan out paying workers to build new capital, plan operations, and run the railroads for about 4 years, then return the railroads back to shareholders. As the private sector followed though on the government plan, railroads boomed, with trains running at high speed between major centers of commerce, with much better service.

The early part of the 20th century was the heyday of commuter rail building, mostly driven by housing and property developers. Two examples: LA was built out by land developers subsidizing trolleys out into the desert where they built housing and real estate for retail and business. Once the land was sold, the subsidizes ended, and the trolleys ended up the responsibility of government.

In Florida, railroads were improved and extended to enable northerners to commute to Florida during the winter. This created the east coast which has Mar a largo at the center, and then turned swamp into Miami, and then put large populations in danger on the Keys. Once the Northerners seeking profit as land developers died off or went bankrupt, the movement of people became government's problem.

Eg, when a hurricane destroyed the railroad line to the Keys, government had to serve the Keys, and built a road bridge.

If you look at big city metros, they are illogical hodgepodge of incompatible lines reflecting the private builders who were driven by selling land for development. Private developers "petitioned" for charters which generally gave them land beyond the right of way to build rail lines to their land. The railroads eventually ended up bankrupt, and demanded subsidies or purchase by government.

And it was Congress that built passenger air travel. First by subsidizing regular passenger routes with money losing Air Mail service via the Post Office. Then via the ICC centrally plannning routes and fares to prevent the kind of unregulated competition that bankrupted competing railroads (built with government subsidies provided by competing towns). But especially with Federal government funding massive amounts of R&D by funding almost everyone who wanted to build a local factory building military planes and had some Congressmen in their pocket.

Then starting in the 80s, govrnment oversaw eliminating the wasteful competition as the GOP renounced their old ways of big government, picking winners and losers over and over, primarily in bankruptcy court where an unelected technocrat redistributed wealth and killed jobs, but also in Congress laying out how industry should be structured, like in the mandate which led to the creation of ULA. Not to mention government training almost all airline pilots and mechanics in exchange for 4 to 6 years of servitude.

Many parts of the world has as much or more resources as the US, but can't compete with the US because they lack all the transportation options in the US. And many parts of the US have deindustrialized due to government neglecting or intentionally destroying transportation services in and too them. Not only are Midwest towns dying, towns in California are dying: no longer transportation, no longer an economy, no longer jobs, no longer anyone to live in houses or go to schools, churches, etc.

The GOP never seems to care about people dying any more. Tey oppose requiring workers be paid to keepmpeople safe. After all, paying workers costs too much, and high costs kill jobs. That is the core of Trump's economics. Stop requiring businesses to pay workers to ensure planes don't crash, coal mining doesn't kill workers or bystanders, products don't kill or harm customers. Government regulations don't require much higher profits (money not paid to workers) that cost customers more, so cutting costs of regulations is always about killing jobs.

Trump wants to get rid of environmental regulations so he can harm property owners, but he will not be willing to pay to keep the workers building and maintaining his wall safe. Instead, they will be working in harsh conditions with no ready access to such things as medical care, or suitable housing and community.

I call the economics you reason by "free lunch economics". Paying workers to build capital costs too much and destroys wealth by building so much capital monopoly/scarcity profits vanish as capital prices fall below the labor cost that built the capital (aka depreciation is acceptable aka entropy).

Extremely fast air travel is extremely costly due to thermodynamics and the limiting factor of getting to and from airports 9plus making your suitcases). That was what really killed the Concorde and why modern airliners are slower than those of ‘60’s.
Hyper loop is absurd. Tunneling is very costly. It is a fixed immutable infrastructure. An airport can send aircrafts to any destination and those destinations change at will and whim.
The other two might account to something.

Tunneling was meant to be a hack around private property and built cities. Without it you are back to surface land acquisition, a hard problem.

The Concorde was reasonably fuel-efficient when cruising at Mach 2. But, it was horribly inefficient when flying at subsonic speeds.

This inefficiency limited its range and that, combined with universal bans against supersonic flight over land, meant there was little practical advantage to it: New York to London (or Paris) in 3.5 hours instead of seven just isn't all that valuable.

Whereas there might be a market for supersonic travel from Chicago to Bangalore, but, that was well beyond the capability of Concorde.

The challenge in designing a practical supersonic transport is to tame those sonic booms so as to allow over-land transport, and to make the plane reasonably efficient at lower speeds.

One of the counter-intuitive things about supersonic travel is that drag, and therefore fuel consumption, is lower at Mach 3 than at Mach 2, and lower at Mach 2 than at Mach 1.2; therefore, higher cruise speeds are better, as is a quick transition from subsonic to supersonic cruise speed.

Yet engines and wings that are optimised for supersonic flight will be inefficient (if they can be made to work at all) at subsonic speeds. Concorde "solved" this my optimising for supersonic cruise, but, that resulted in high fuel consumption in all other phases of flight and that limited its range.

The latest thing is variable or adaptive cycle engines which adjust things like the bypass ratio to the flight regime. They can be efficient as a subsonic accelerator or supersonic cruiser, at least up to a few Mach.

They will likely debut in the next generation of fighter, and while it looks like we may have new supersonic passenger aircraft before that, eventually we'll see them in that application too.

Extremely fast air travel is costly due to thermodynamics. This what killed the Concorde. Modern airliners are slower than in the ‘60’s.
Hyperloop is absurd. Tunneling is slow and expensive. It is a fixed immutable infrastructure. An airport can send planes in any direction at any distance for the same cost of runways.
The other two might amount to something.

Huh? The hyperloop is absurd because tunneling is slow and expensive? Yet every major city has tunnels.

"The other two might amount to something."
There is no place on planet earth that has fusion energy. Quantum computing can barely factor the number 15 and that's with a ton of errors.

Hyperloop tunnels have to cross countryside. Very long tunnels are not very common. The longest railroad tunnel is barely 60km, the longest water tunnel, the Delaware Aqueduct, is 137km, and water tunnels require a lot less precision and have no real need to worry about stuff like grade as long as they are under pressure. Hydrologic
Systems can deal with a massive amount of leakage yet a hyperloop tunnel needs to maintain negative pressure over a huge volume with a massive surface area.

I am not saying it can’t be done but 350km tunnel doubled over, a return is necessary to the proposed designs, would be a technical marvel in itself.

You can "tunnel" cheaply, but you need to open dig and cover. Hard to do that "under" anything already built.

Air travel I could see happening notwithstanding the comments above. The demand is there, no question.

Transportation, I see hyperloop, subway, rail as decidedly NOT the future. Too expensive to build things.

The future in city transport is going to be using the existing infrastructure smarter. Some mix of dedicated bus lanes, bike/scooter lanes, congestion pricing, parking villages and buses into city centers. All of that can be done without self driving cars, but if you can get self driving within 10 years, you can probably get even more out of the existing infrastructure.

And all paid for in Bitcoin?

We can use our Bitcoins to invest in Fusion Energy stocks!

There's a few promising startups all attempting to make supersonic flight available in the mid 2020's, some of them have generated a lot of interest from traditional aerospace companies. The Concorde was built in the 60's with sliderules. Modern engines and airframes are modeled with extremely powerful and accurate computer models.

Additionally, SpaceX's next rocket should be able to do suborbital hops between any two cities on Earth in ~half an hour, for fares which are competitive to existing airlines.

For Hyperloop, I'm hugely pessimistic. No one has shown how making hundreds or thousands of contiguous miles of pressurized tubes economical or even possible.

"Additionally, SpaceX's next rocket should be able to do suborbital hops between any two cities on Earth in ~half an hour, for fares which are competitive to existing airlines."
Landing where? Which nation on earth won't regulate this? And what do you mean by competitive?

Well, you've got spaceports in FL, CA, NM.

French Guyana, Kazakhstan, I assume China and India have them. Wouldn't be tough to place one in Australia.

French Guyana and Kazakhstan. OK.

On concrete pads. Presumably at places like airports because it will be fairly loud. Did you see the Falcon Heavy boosters land yesterday? It looks pretty benign compared to an airliner hitting the ground while its still doing 150 mph.

To other commentors upstream who worry about fast air travel being energetically expensive, this is the answer. We might see people buffeting about in the atmosphere at Mach 2 or 3 in the next few years, but we won't seen Mach 25. Mach 25 will be done ballistically, and primarily outside the atmosphere.

UBS called it a $20B annual market. It could wipe out long haul air travel. Though I'm not nearly as optimistic about the time-frame. Imagine the effort to get one of these things type certified?

"On concrete pads." Duh. Really? What I meant is neighbors wouldn't allow El Toro Air Station to be turned into a civilian airport. Do we really think landing rockets is going to be an easy sell? Also, look at the reaction to two 737 Maxes crashing.
I think we agree about the challenges when you speak of certification.

I guess I'm saying that if an airport works for big jets it will work for type certified rockets. Noise is the big issue.

Getting to type certification is going to require a lot of work for regulators who will need to develop new procedures. The upside is rockets are much simpler than jets and have less to go wrong. The downside is they're unfamiliar. My guess is type certification will follow a long history of lots of safe space launch, which is only becoming possible because prices are dropping so much. It won't be based on theory.

Are you just expressing skepticism that they can be made reliable?

No. My skepticism is entirely about regulation rather than technology. Making the rockets is viable, I'm sure. It's the infrastructure I have doubts about.

Understood. I think that's a reasonable concern.

If I was trying to attack the problem, I'd want to have many thousand flawless space launches under my belt first. And I'd start with military transport before I tried to displace commercial airliners.

Far more likely, the launch/landing pads will be big floating barges offshore big cities. Or big tethered platforms, like oil wells. Passengers might take a helicopter or fast boat to the launch pad. Even if there's some commute out to the actual launch pad, the speed of the rocket more than makes up for it. Obviously rocket flight wouldn't be time or dollar efficient for Boston to NYC, but for NYC to Melbourne, it makes a lot more sense. Instead of an airliner pushing itself through the atmosphere for 10,000 miles, a rocket only is in the atmosphere for a few miles at the beginning and end of the journey. The rest of the time it's outside the atmosphere and can build up ridiculous speeds.

Short-term, yes, but that's largely because there's still a considerable risk that they'll crash. It's not a mature technology - it's been stagnant for 50 years.

By definition, crashing is extremely unlikely for a type certified craft. So if we accept that the reliability is coming, we can dispense with the requirement that launches and landings occur on remote ranges.

Noise is still a big issue, particularly at the scale of rockets really optimized for transport of bulk cargo to space. One might reasonably want passenger rockets to be smaller. Not least because they can fly a route much more frequently than slow airliners and therefore optimize at lower passenger capacity, and more frequent flights, per route.

If I had to bet, rockets won't be carrying passengers point-to-point at scale while they're still risky enough that they need to be launched off-shore. That "spaceflight participant" stuff won't be considered okay for tourists and business people for ordinary travel.

I think SpaceX's illustrations of that sort of thing are just to get us used to thinking of it as plausible, not because that's what they specifically intend to do. I also don't think this sort of thing happens if it's just SpaceX.

Speaking of tech, did anyone else catch the SpaceX Falcon Heavy launch earlier today. The landing of the boosters made me tear up a little. Elon Musk is truly Iron Man.

+1, all three parts of the rocket successfully returned to Earth

It makes point-to-point rocket travel look like a pretty reasonable idea.

Speaking of which, the US Air Mobility Command better be paying attention to all this.

SpaceX's Starship, which already has a prototype undergoing hop tests, can easily carry an M1 Abrams and could potentially hop back and forth between CONUS and Doha or Ramstein every hour on the hour, rather than every day or so for a lumbering C-5. And, it doesn't loiter at low altitude near hostiles - it's barely in the atmosphere for the very beginning and end of its flight.

The cost of the trip would exceed the value of the tank. M1's are around $10 million.

Starship, which is really the first generation product that might be capable of such a thing, is supposed to cost single digit millions per flight. So I think your price estimate is too high. Fuel is only like 500k. Really if you want to get the costs down further you'll want to dispense with the first stage and its operational complexity, but that's a more advanced vehicle.

The C-5 would cost about $1m to do the same, based on my very rough multiplication of the cost per hour times the hours of flight. I have no idea if that's a complete accounting; it strikes me as low. But it's not a massively different number, and it's what we use today.

Obviously if you can move tanks by ship, you do. Flying them makes sense in emergencies only. But moving stuff fast in an emergency is exactly what these new technologies could be good at, and it's been a pressing need for the military for ages.

The payload that the rocket can haul is really high >200 tons to low earth orbit (so somewhere near that for suborbital hops), about the same as a C-5 Galaxy (the largest military airlift plane in the US).

The ticket price for SpaceX's suborbital rocket is expected to be competitive to current longhaul airlines.

There's no way point-to-point passenger rockets will be approved any time in the forseeable future. You will be putting all your passengers on top of the equivalent of a twenty story building filled with explosives. Rocket fuel is not at all like JET-A fuel - it's ready to explode into a giant fireball at the slightest leak or failure. Have a look at what happened to the falcon 9 when it exploded on the pad during a fueling accident. SpaceX's Starship would have several orders of magnitude more fuel.

I think you would have to establish safety across many years and hundreds if not thousands of flights before anyone would allow passengers to fly routinely.

In addition, a ballistic rocket is basically an ICBM. How many countries are going to be happy with the notion of a giant missile being fired at them, near large cities, daily? What would happen if a 20 story building filled with explosives fell in the middle of Los Angeles?

SpaceX put the point-to-point launch idea in a promo video to show the versatility of their rocket. I think they know that we won't be seeing anything like that for decades, if ever.

Meh. Rocket fuel basically is kerosene. It's RP-1, commonly. That's just a tight spec of kerosene. Jet fuel is also just a tight spec of kerosene. Commercial airliners also explode when they crash. They just don't crash much. This isn't a material difference. You need to solve the crashing problem, not think about different fuels.

Clearly you'd need thousands of flights to demonstrate that level of safety. But reusability means you're going to get that in a few years.

That said, type certification is a lengthy process, so a decade or two is probably a reasonable number. I also expect version 1.0 of SpaceX's reusable product is not going to be the vehicle we see actually do this. FWIW, that vehicle runs on methane, which is again, nothing special.

A rocket is no more an ICBM than an airliner is a bomber. Those are just scary words that don't mean much in this context.

SpaceX Starship doesn't use kerosene - it uses methalox, or a combination of methane and liquid oxygen. For that matter, you conveniently left off the part where a kerosene rocket is kerosene AND liquid oxygen. It's the presence of the dense oxidizer that turns the kerosene or methane into an explosive, just as it's the oxidizer in fertilizer that turns diesel fuel into an explosive.

I'm very familiar with type certification. If you want to carry passengers for hire, type certification demands all kinds of passive safety that just isn't built into rockets. If they want to certify rockets to carry paying passengers on anything other than an experimental basis, they'll have to write the certification rule book from scratch. That alone will probably take a decade at the speed the FAA works, not to mention the extensive coordination between governments that would be required.

We might see this type of transport some day. I don't expect to be alive to see it. In my opinion, these rockets would have to have thousands of incident-free flights before anyone is willing to consider them for public transportation.

I stated in the post you're replying to that Starship's fuel is methane, which is again, nothing special.

But yes, just like for any other type certified aircraft, you'd expect thousands of incident free flights before certification. And yes, certification itself takes a long time. And yes, it's built around the assumption that the vehicle is a plane built by one of the majors.
None of those things is controversial.

With the exception of solid rockets and hypergolics, the oxidizer just isn't that important an addition to the problem. It's pretty hard to get fuel and oxidizer to mix in a way such that they will detonate, and there's plenty of oxidizer in the atmosphere anyway. What you see in most rocket crashes is the vehicle breaking apart and the fuel burning up. If your airplane comes apart at speed, you will have just as bad a day.

SpaceX’s product has two big problems for the point-to-point passenger market.

First, it stages, and I doubt staging will ever be approved for routine passenger travel, any more than in-flight refueling would ever be approved. They just won’t let two unconnected objects fly close to one another with 100 people onboard. I view that as a reasonable objection should a regulator raise it.

Second, flying at its potential cadence, it’s an order of magnitude more capacity than an A380, which is ridiculously big. It’s designed to be an efficient cargo ship, not a passenger plane. Rockets can in theory fly much more frequently because it takes so much less time to fly a route. Therefore the rocket wants to be a lot smaller than the plane it’s replacing to match the required capacity of long-haul routes. So a 30 passenger rocket flying ten round trips a day would replace a 300 passenger aircraft flying one round trip a day. That’s more the size that the first point-to-point products will aim at, and it dramatically reduces safety concerns related to the total energy in the vehicle.

Then again, FAA has become very risk-averse.

Can your ballistic Mach-25 vehicle get to breathable atmosphere within 90 seconds if pressurization fails? If not, no certification for you.

Why 90 seconds? If a Concorde has a breach large enough to evacuate it in 90 seconds, everybody is going to die. I wouldn't expect a requirement not imposed on commercial aircraft to be imposed here.

Further, rocket flights are inherently short. Slow developing problems, like a leak, are less of a problem not more.

Impacts like bird strikes are also less of a problem because unlike an airplane, all the low altitude stuff occurs at low speeds. You aren't going to run into a bird 300 miles up.

It will eventually be seen as inherently safer than airplanes.

Strange to omit AI. Never mind general AI and singularities, which are surely fair game at this level of blue-sky speculation... we've barely scratched the surface of what today's machine learning will unlock in the short to medium term. China is not squeamish and other nations dare not fall behind.

When everyone has AI - or super fact statistical analysis, all that matters is the data. And getting data to behave is actually still quite hard. I'm quite blase on the potential here. What do you see?

"all that matters is the data. ... What do you see?"

Aren't you ignoring the obvious. Self driving cars, robotic house maids, gardeners and waiters, etc. Highly automated manufacturing. Partial Von Neumann machines driving a space industry....

Self driving cars will be/are a regulatory problem, robotic house maids will force us to organize our house so they can be cleaned by them (see cost of Boston Dynamics relative to the cost of immigrant labor), and automating manufacturing is a long way off from being dynamic enough to justify investment beyond what already exists. And your last statement sounds like a Wired article, which is a great place for science fiction.

More civilly, these things may come to pass, but I don't see them accelerating development in the way the other items might.

"Strange to omit AI."

It's an economist over 55 thing.

I think Samir needs to explain why he thinks fusion energy will work this time. Is he a scientist? I googled the name but found many Samir's. I doubt he's a lawyer.

+1. People need to understand that even if fusion reaches sustained net energy production it will still be several times orders of magnitude away from being economically competitive with other zero emissions energy sources.

Thorium power is far more short term viable than fusion. 5 years away at most as it’s relatively simple engineering problems. Mostly the problems are ‘how to build 500 similar units to get costs down’. The science is already understood.

given the way that productivity is calculated will any if these innovations show up in the statistics? Taker quicker airline travel for instance, this could actually reduce measured productivity if it takes more people to run the rockets than a 737.

#1 - Air travel: a faster plane reduces the flying time. But we don't travel from airstrip to airstrip which accounts for flying time, we travel from door to door.

For a hour flight you have to be at the airport 1 hour before because security, drop luggage or spend 15-20 min moving inside the airport to get the gate, then another 15-20m while all the inefficient humans fight for the overhead containers, put jackets on them, and request to be seated with family members after not doing it at the right time: when getting the damned boarding ticket . Then comes the actual flying. When you land, sometimes the plane goes to a gate =) or sometimes parks in the middle of nowhere and a bus comes and you spend another 10-15 min on ground just moving inside the airport. Then, something as simple as getting your luggage back means another 10-15 min.

The point is a 1 hour flight can become a 3-4 hour ordeal. If some engineers cut flying time by half, bravo......the whole ordeal was reduced from 3 hours to 2.5 hours. There are 3 periods to minimize before flying time: a) boarding time, b) displacement inside airports, c) recover the luggage.

In the year of the Lord 2019, only 1 door is used to get in/out small planes (i.e. Boeing 737) . I've only seen a few airport use 2 or more doors/tunnels to get people into a B747 or a A380. Just build a second tunnel for small planes to use the rear door, it's the low hanging fruit.

Yeah, I've wondered about the two doors thing for a long time. You lose a couple seats you can sell each flight, but boarding time gets cut in half. I would think that's a win for shorter flights, where it takes 20 minutes to load the plane for a 60 minute flight.

One of the nice things about Burbank airport until several years ago was that, being old and relatively tiny, it didn't have jetways at all. Instead they had those stairways on wheels that they'd roll up to the airplane's door so the passengers could emerge and walk down the stairs and walk into the airport (maybe pausing at the top as if they were the Beatles or the President of the USA landing at their destination).

Some airlines (IIRC Alaska and Southwest) and some planes (IIRC 737s) used their rear doors for deplaning (and in the case of Southwest, boarding). So deplaning and sometimes boarding would be almost twice as fast because both doors were being used.

But then the airport or the FAA or somebody told the airlines they couldn't do that any more for some sort of safety reason.

And I started using LAX more often than Burbank because the fares were lower and I could fly nonstop to major destinations. LAX was inconveniently across town for me, but I learned that if I drove there before 6 or 6:30am, the LA freeway system is the fastest transportation system on earth. Even a helicopter would be hard-pressed to match the travel time, given the need to drive to the heliport, check in, do flight checks, etc. compared to just hopping in my car and zooming down the freeways.

Self driving cars are a much bigger deal than faster airplanes. Why not take longer car trips if I can get work done while traveling or just watch some movies.

Self-driving cars are as bad as I predicted, for the reason I gave.

Self-driving requires real, human level, intelligence. Neural nets and pattern matching has hit its wall (or pedestrian).

From the article you linked to: "Even though Hackett walked back immediate hopes for autonomous vehicles, he made it clear that he believed in the technology's potential. "When we bring this thing to market, it's going to be really powerful," Hackett added."

That is total CYA for earlier optimism.

The point, repeated throughout the body if the article, is that everyone is scaling back.

If you understand current AI (as statistics and pattern matching, rather than deep knowledge representation and manipulation) you know why.

They didn't have an obvious path forward.

The second article is bad. The writer uses two examples to say the hyped predictions of the past were wrong. He quotes a Gardian headline from 2015: "Self-driving cars: from 2020 you will become a permanent backseat driver" which nobody thought except apparently The Guardian and the article itself didn't imply that. Next up, Business Insider with the title: "10 million self-driving cars will be on the road by 2020." Again, in the article there is a quote from one study that claims 10 million cars in 2020 will have *some* driverless feature.

After that, the writer extensively quotes a social scientist about AVs... Oh and AV skeptic Gary Marcus, a psychologist, gets a quote as well.

The first article does not say everyone is scaling back as you claim. Uber is mentioned and in a linked article the CEO said: "What is clear is that in a 10-year timeframe there will be a mix of both (self-driving and human-controlled cars)."

No idea your skill level, but if you want a foil with strong reputation in applied AI and Robotics, why not Rodney Brooks? He's quoted there as well.

Yeah, I've known Brook's "arguments" for a couple of years, which seem mostly hand waving. He's also said ridiculous things about the risk of GPS collapse.

Well, you and anyone else can bet against him. I don't think that's smart money, but maybe that's just me .. and anyone else who really understood that video.

Right... just you and anyone who agrees with Brooks understands A.I. and the hurdles. I'm at 45:00 and Brooks assumes A.I. health robotic care assistants will need to be helping people get kidney dialysis treatments *decades* in the future. At 53:00. Brooks takes a long time to say "A.I. isn't perfect yet."

Well, if I may interpret for you .. Brooks is addressing AI promises, including for AI elder-care robotics. Newspapers can be glib about that, and breakthroughs along the way .. as they are for all AI including driverless cars.

You can't expect driverless cars in "decades," but not elder-care robotics in the same timeframe, because it is the same problem.

These things need a robust general intelligence, something we (as Rodney repeats) don't know how to build.

No, it is not the same problem at all. Brooks at 62 is just clueless.

Anyone wanting the crux, skip to 34:00

Sorry, not sorry. Almost a dad joke

Perhaps the insane obsession with faster travel is caused, at least in part, by the ease with which it is visualized. We already have travel of various kinds so increasing its velocity is just adding to a working paradigm. On the other hand, a completely new answer to the problem of "saving" twenty minutes in a trip from St. Louis to Chicago is still in the world of science fiction. Also, isn't the fixation on time really evidence of a growing cultural egotism coupled to the individualism created by a state that assumes the roles of clan and family? Why are people in such a hurry?

Disagree somewhat about flying time. As someone who is both tall and stocky, being crammed into those cattle cars they call a coach seat these days is no fun at all. Shortening the amount of time I have to sit crammed in there is worth a good deal to me, and I don't even fly that often.

So while commuters are stuck on the freeway for hours, wasting their time, consuming resources, and filling the air with carbon emissions, the answer lies in the Hyperloop? Tech can't build efficient transit, but it can build flying cars and rockets to Mars. Tech can collect, store, and mine data and has created vast wealth doing it. But with that wealth, tech is increasingly drawn to what are vanity projects. Recall a few years ago tech was going to build a self-driving car. Yes, a car. Tech learned that building a reliable car is hard and quickly abandoned that effort, shifting their resources instead to converting the windows of a self-driving car built by somebody else into a billboard. But give tech its due, for tech has given us something highly practical, an efficient cab company. It's not exactly cutting edge technology, but it's efficient. Alas, the business loses billions per year and has no strategy for reversing the losses. But there's hope: using self-driving cars as cabs and converting the windows into billboards. Selling advertising is what tech excels at.

I will acknowledge that I don't understand core platform contribution margin:

#2 hyperloop: high speed is sexy, it arouses engineers, but is it convenient and comfortable? Our cars travel at puny 70 mph and we love them because we can travel with the comfort of a living room from door to door.

The problem with cars is congestion. The Hyperloop logic to avoid congestion is to put humans on tunnels and leave the beautiful highways to semi-trailers moving cargo. What if the cargo is sent to the tunnels and leave the highways for passenger cars? Some people is thinking about moving cargo underground 24/7 with autonomous vehicles at 35 km/h.

#3: synthetic biology is just biology, no one talks about new physics or modern math.

#4: the greatest challenge will be coding. our previous knowledge may be useless in the new hardware.

#5: let it rest until a significant improvement is done.

"... no one talks about new physics or modern math."

Physicist/computer scientist Stephan Wolfram wrote "A New Kind Of Science" in 2002. A second edition came out in 2012 with the title "A Not As New As Before Kind Of Science" so there's that.

I am a nonstagnationist. I believe the critical error stagnationists always make is to compare their moment in time to past decades, or centuries.

Step back and tremendous changes have happened, and technological progress has not slowed down at all.

The 80's prediction that we would master elections and then genes still looks good, and fifty years to do it isn't so bad, in the scale of human history.

(Demanding particular gee-whiz things from the future is part of the stagnationist problem, and not a credible measure of success. It leads them to discard a lot of wonderful things as not what they asked for.)

lol, no coffee yet this morning.

Electrons, not elections.

Imagine, declaring stagnation exactly as smartphones spread across the planet.

I get the sense that there was perhaps a hyperlink (maybe more than one?) embedded in the original email? If so, mind including Tyler? Congrats on the foresight, "guess" or not.

>they’re STARTING to get INCREASINGLY clear

>STARTING to show that it MIGHT work

Hedge much?

This may be small potatoes, but I remember reading about a couple of different technologies a while back that were supposed to replace SSD drives with non-volatile RAM that would be as low-latency / high-bandwidth as current RAM but retain its state without power. I wouldn't mind that coming to market.

I wouldn't mind either. I'm not sure end users would notice the difference, but it opens up whole new computing paradigms based on universal object persistence. Which of course brings both opportunities and management issues.

Still it would be a gold rush for the great new operating system.

"Instant on" booting would be nice. I/O heavy applications (databases, video processing) would see a huge boost.

Quantum computing is overrated. It'd mean the cryptography would need to change, and not much else. It won't make your copy of Minecraft run any faster.

And yet QC seems the least subject to regulatory and activist slowdown. I can easily imagine a crowd of snowflakes sitting in the path of the hyperloop construction for the sake of the poor birdies and prairie dogs.

The computers that will matter are not the ones on your lap or under your desk. It will be the data centers that run AI routines and process sensor data. Quantum computers might well speed this up, and the effects might really matter.

STILL racing into an irrevocable future?

Why are we not there already?

(Why didn't we get there decades ago?)

What specific threats to our tech tyrannies and to our lying and spying Tech Sector are posed by the advent of Technogenic Climate Change?

More interesting is that serious people fantasise about mega-engines at no cost to their credibility, whereas it is more costly to have unfalsifiable fantasies about major world religions, for instance.

Nuclear fission, which is orders of magnitude less technically complex than fusion, is not economically competitive in a large number of countries at the moment. What on earth is there to suggest that fusion will ever be economically viable, even if anyone manages to overcome the nightmare technical issues? To my knowledge there's no material on earth that could withstand the levels of neutron bombardment from a commercial fusion plant without starting to crumble away within a few years. Most of the work that's being done is still on the plasma physics, because that's the sexy part, but materials is where the bigger issue will be.

Nuclear power is not competitive for purely regulatory/activist reasons, not because the technology can't be built cost-effectively. The standard trick of activists is to wait until construction has started and billions of dollars committed, then hit the site will constant injunctions, reviews, and regulations to slow it down. Heavy up-front capital investment makes nuclear power plants uniquely vulnerable to this kind of activism. The solution is simple: Pre-certify existing designs that have proven safe, limit the time period for legal review, etc.

The bottom line is that countries that have extensive nuclear power also have the cheapest power. Countries trying to generate power with renewables have the most expensive power.

The new technology that isn't even mentioned is the development of new Small Modular reactors. These are built in a factory, shipped to a site on a flatbed truck and buried in the ground. Heat from the reactor powers steam turbines, which can provide power for thousands of homes. After ten years or so, you dig up the reactor, put it back on a truck, and send it back to the factory for spent fuel removal and refurbishment. No on-site storage required, no meltdown risk. No environmental damage. The amount of energy that would require a gigantic solar farm or wind farm can be provided with the surface footprint of a small industrial building.

If we were actually serious about global warming other than as a political tool, virtue signaling issue or means to raise new taxes, we'd be accelerating development of SMRs and embedding them everywhere.

Here's a simple question. As I understand it, nuclear waste needs to be kept safe forever.

Isn't the net present value of forever always infinity?

Are we so "serious about global warming" that we are ready to pay infinity?

You are not correct. There are many types of nuclear waste. The stuff you are talking about, that needs to be kept safe for thousands of years, is the spent fuel from light-water reactors. That's actually a tiny amount of waste - all the spent fuel generated from reactors since the start of the nuclear era would fit in a football stadium. The other waste is things like contaminated water, gloves and lab coats, tools, and other low-level waste that isn't particularly dangerous.

Back to the high-level waste - modern reactors like the CANDU actually eat that kind of waste for fuel, and the waste they produce has the really long-lived stuff removed. The waste from a modern reactor is only dangerous for a few hundred years, not thousands. That is eminently manageable.

There's an awful lot of fear mongering out there about nuclear power. Sadly, much of it is coming from the same people who think global warming is an imminent threat to the entire planet. They don't understand that nuclear fission power is the ONLY solution we have that could get us to a non-GHG energy infrastructure in less than 50 years. And we are told we have only a decade or so before we're really boned, so everyone should be singing the praises of nuclear power.

It seems like generations hundreds, or thousands, of years in the future pay a cost either way. It's not like warming or nukes are all win.

Whereas renewables are all potentially recyclable? Or at least shelf stable.

Renewables use a huge amount of resources in their construction, because they are harvesting energy sources that are not very dense. If we were going to power the world with wind and solar, we would have to increase our mining operations by a factor of about five to provide all the raw materials required. That will require more fossil fuel burning, more waste, and lots of damage to the environment as we clear the thousands of square miles of land required for these huge installations.

But the larger point is that it simply isn't feasible. You can never get to more than maybe 50% of all your power needs from wind and solar, because there isn't enough of it available when you need it. For example, where I live in Canada we get about 1/4 the solar energy in the winter as we do in the summer. So if we build enough solar power to cover 100% of our summer consumption, we'll only produce 25% of our consumption in the winter. You would have to over-build the infrastructure by a factor of four to get 100% renewable power, and even that doesn't account for long periods of clouds, snow storms, etc. So you still need some form of baseload power that can load-follow, and you need enough of it to prevent blackouts and brownouts when the sun doesn't provide energy for long periods. That's why renewables are so expensive at the grid level - you basically have to maintain the old power systems along with the new.

Finally, solar and wind require huge engineering, which means you can't build them fast. Ten years is a ludicrously short time to convert the world to renewable power, or even to 20% renewable power. Germany has been building out solar farms and rooftop solar for over 20 years, and solar accounts for 6.8% of their power.

If we are going to demand that all power in the future be renewable, without nuclear in the mix, be prepared for a future where we still get the majority of our power from fossil fuels 50 years from now.

Here's another simple question: As I understand it, chemical waste (for example coal ash full of heavy metals) needs to be kept safe forever. Why is the charge of "this mess lasts forever, therefore it is unacceptable" only levied at nuclear waste?

Fly ash from coal power stations is disposed of in 3 main ways:

1. It is added to concrete.
2. It is stabilized by growing plants on it.
3. You say you are stabilizing it but let the wind keep blowing it way and rain keep washing it away until the problem is solved.

People tend to get upset if you try any of these methods with radioactive waste and number 2 can be difficult if the radiation is a bit intense.

Activists aren't behind the cost overruns on the two EPR reactors which are currently waaaay behind schedule. You can blame the Japanese steel makers who screwed up the steel for the pressure vessels for that. And activists aren't anything to do with the changing grid economics now that large amounts of renewables cause prices to fluctuate so much.

The reason nobody talks about SMRS is that they aren't actually an extant technology, and are a very long way away from being so. The are lots of nice on paper designs and, well, that's it. They would have lots of their own issues. Much less thermally efficient for one, due to their size. And a number of the molten salt designs will face the usual problems that stop molten salt reactors from working well.

For the avoidance of doubt here, I used to work in the industry and have actually spent plenty of time stood on top of commercial reactors, so I know what I'm talking about. The real nuclear industry is a very different place than the one that YouTube and Tyler would have you believe.

Research thorium. Environmental concerned are vastly lower and we will be able to take advantage of knowledge gains from building 1000s of small plants.

I was shocked fusion made this list over thorium

Out of these five technologies, two are essential to an unlimited future: synthetic biology and fusion power. Faster airliners are nice (no more 12-14 hour jaunts across the Pacific), but not as key as these first two.

Synthetic biology is key for the next industrial revolution as well as to end the tyranny of aging and death once and for all. The desirability of fusion power should be self-evident to anyone with a room temperature and above IQ.

Hyperloop is a fantasy that will never work as proposed. We have no solution for the metal expansion of the tubes or the buckling problems from differential heating. We don't know how to safely contain a vacuum in a tube hundreds of miles long. We don't have vacuum pumps that can evacuate such a tube quickly or without huge expenditures of energy. There is no proposed solution to the various safety issues that arise from putting people inside a solid welded tube hundreds of miles long, surrounded by a vacuum. Furthermore, by its nature Hyperloop can only carry a small number of people at a time. It is not a mass transit solution.

What is more concerning is that the various companies involved don't seem to be trying to solve these show-stopping problems, but rather spend their effort designing fancy-looking but hollow pods, sponsoring design competitions for futuristic terminals, etc. In short, things that don't advance the engineering, but do provide for a good show. I suspect that these companies have a business model built more around rent-seeking and raising funds from investors than building a working transportation system. Hang on to your wallet.

We've known how to build supersonic aircraft for a long time. They just don't make economic sense. Maybe we'll figure out how to make one that's economical, but all that does is speed up one part of a chain of time consumption involved in air travel. I currently have to arrive at the airport two hours before my flight. I have to catch a cab to the airport, which is generally way out of the city. So I've used up 2.5 hours. Now my flight across the continent is maybe 5 hours, of which an hour is spent leaving the gate, taking off, climbing to cruising altitude, descending from cruising altitude, landing, and getting to the gate. Another hour to get my baggage, hail a cab, and get to my hotel. A total of 8.5 hours. Now, if I fly in a plane that can go Mach 2, I might cut the cruise portion of the flight from four hours to two. So now my 8.5 hours is 6.5 hours. Not a huge improvement. Certainly nothing that is going to change the world. We could improve efficiency more by just getting rid of the current security theater we engage in.

Fusion may be of use some day, but fission is here today. Fusion won't change the game - especially when the activists find out that neutron bombardment of the reactor will make it radioactive over time and need to be cleaned up/disposed. Also, they're talking about 'prototype' fusion plants that can actually generate power by 2040 - conveniently about 20 years away. 20 years is a magic number in engineering - it's close enough that people can say it's 'just around the corner', but far enough away that we can hand-wave away serious problems as something 'to be solved before then'. This is why fusion power is 20 years away - and has been for the past 40 years.

Here's a technology that should have been on the list: Reusable rockets. Falcon Heavy just launched for $90 million dollars. The nearest rocket to its size, the Delta IV heavy, can lift about half as much and costs $350 million. And the only reason Falcon Heavy still costs that much is because it has to throw away the upper stage. When the next big rocket from SpaceX flies, it will be 100% reusable, and theoretically could launch three times the payload of Falcon Heavy for under $10 million dollars.

These kinds of prices open up space to universities, medium sized companies, exploration firms, etc. There are asteroids out there that contain more metal than all the metal refined on Earth to date, and it's all just sitting there waiting to be brought back. Solar Power Satellites may become feasible at low launch costs. Tourism in orbiting hotels could easily become a thing. And like most breakthroughs there will be new uses and profits to be made that we haven't even considered. But the fact is, we are on the doorstep of the era of cheap and easy access to space, and that could be the biggest game-changer of all. The current satellite and space market size is built around the assumption that it costs several thousand dollars to launch a single kilogram into orbit. What does that market look like when it sinks in that you can now launch that kilo into orbit for $100?

"There are asteroids out there that contain more metal than all the metal refined on Earth to date, and it's all just sitting there waiting to be brought back."
This will greatly upset all the miners that voted for Trump. They'll need bigger red hats to contain their seething rage at their jobs being destroyed by science and technology loving blue states.

"I currently have to arrive at the airport two hours before my flight. I have to catch a cab to the airport, which is generally way out of the city. So I've used up 2.5 hours. Now my flight across the continent is maybe 5 hours, of which an hour is spent leaving the gate, taking off, climbing to cruising altitude, descending from cruising altitude, landing, and getting to the gate. Another hour to get my baggage, hail a cab, and get to my hotel. A total of 8.5 hours. Now, if I fly in a plane that can go Mach 2, I might cut the cruise portion of the flight from four hours to two. So now my 8.5 hours is 6.5 hours. Not a huge improvement."

Computer-piloted vertical takeoff and landing systems eliminate most of your time on the ground. The "airports" at arrival and departure can simply be large parking lots (which you get to and from in computer-driven cars). There is no delay circling around congested airports or taxiing to and from runways.

For any airline flight less than maybe 1500 miles, subsonic omputer-driven VTOL aircraft will be far faster door-to-door than supersonic aircraft taking off and landing from conventional airports.

My guess is that only a small fraction (less than 1/3rd?) of the passengers traveling globally travel more than 1500 miles on a given trip. I know that less than 1/3rd of the airline flights I've ever taken have been over 1500 miles.

Good post Dan. I think the supersonic aircraft time saving problem ( only modest savings ) that you point out can be solved.
Supersonic self driving transportation to the airport using an hyper loop powered by fusion power ;-)

Computer-driven surface and air transportation will be a much bigger deal than supersonic planes or the hyperloop.

Computer-driven surface transportation of people and goods allows speeds in excess of 100 mph even on highways within cities. It completely eliminates traffic congestion. It will cut the cost per passenger-mile traveled by more than 50%.

2) Computer-driven air transportation allows vertical takeoff and landing aircraft that can make essentially any large parking lot an airport. (No need for runways or even a terminal.)

From what I've heard, the whole "building a spaceplane" thing with Reaction Engines isn't really happening - they just build heat exchangers.

We'll see with Fusion Power. It has been getting closer to be useful over the decades, but there's a still a huge gap between "maybe we'll be able to hold a net-positive fusion plasma in action for longer than a few seconds" and "this produces enough power at a reasonable enough cost to be competitive with renewables and other energy sources."

Comments for this post are closed