A Visit to Bengaluru

Posted on September 20, 2026 by jlredford

I recently spent a week in Bengaluru, Karanataka, India on a business trip. I went to help a customer make use of a new product from our company. The trip cost about $7000 in airfare and hotel. If the customer does adopt it, we expect that they’ll buy about $20M a year. That’s $10,000 for each of the 2000 working hours in a year. If my trip advances the adoption by a single hour, the trip will have been worth it.

This is not a big product or a lot of money by American standards, but I mention it because it put me in a sealed First World bubble inside the chaotic and lively world of this huge city. I stayed in a glossy hotel, met in a conference room full of video gear, and was driven everywhere in an air-conditioned and sound-insulated cab. The acoustic insulation matters – there are signs on the streets saying “The solution to traffic is patience, not honking”, but no one paid them any attention.

Here’s what typical businesses look like along the street:

A store selling plumbing supplies, a store dealing in silver, and above them an office doing data analytics. English is an official language, and everything is labeled in it, but the beautiful swirling script is the local language, Kannada, a Dravidian language three times as old as modern English.

Here’s where I was working – gardens, grass, and no crowds. It feels like a different planet, and there are high walls and guards around it.

At one point my hosts took me out to a local vegetarian fast food place. It was only a block away, but they insisted on driving. One had a nice Indian-made EV called a Mahindra that looked just like a Land Rover. I insisted on walking back, but they were worried about delicate American ankles. The sidewalk was all broken up by wear and tree roots, and scooters would swerve onto it periodically. We survived, much to their relief. The food was excellent, but somewhat mysterious:

Bengaluru is the tech capital of India, and that goes back a surprisingly long way. The last Muslim ruler of the area, the Mysorean Sultan Tippu, fought several battles around here against the British East India Company in the 1790s. The British did NOT have a tech advantage over the locals – they were both using muskets – but the Sultan had something they had never seen – rockets. These were tubes of soft iron packed with gunpowder, with a bamboo guidance stick in the back and often a blade in the front. They could be fired in an arc like artillery and explode on landing, or could be fired horizontally and pierce through soldiers and horses:

Mysorean rockets at Tippu Summer Palace from quint.com

They wreaked havoc among the British and sepoy troops, but unfortunately for the Mysoreans, one of the British commanders was Arthur Wellesley, later the Duke of Wellington and defeater of Napoleon. In their last battle in 1799 Tippu was killed, and Wellesley discovered his body. He became governor of the area and crushed the remaining resistance. The rockets were adapted by William Congreve to British use, and became a standard weapon of the 19th century, most famously in the attack on Fort McHenry in 1814 that inspired “The Star-Spangled Banner”.

The British liked the area around Bangalore because of its mild climate – it’s at 1000 meters. It was actually cooler when I was there than it was in Boston. It stays between 10o and 30o C year-round. They built huge greenhouses like the Crystal Palace in the great park the Lalbagh Botanical Garden. It has bandstands, topiary, and groves of bamboo full of wild monkeys. They also built huge forts in the center of town. They’re still there, but now occupied by the Indian military and aerospace operations. They’re grim places with concrete walls with “No Drone Zone” signs everywhere and topped with barbwire. I did see an actual “Trespassers Will Be Shot” sign. The Indian Space Research Organization, ISRO, is also there and is doing phenomenal work these days, having just landed on the Moon.

Getting here takes twenty hours on planes each way, and the time zone change, 10.5 hours, is really hard. It’s an exhausting trip, but I must admit that I’m glad the company dime lets me get here. It’s a part of the world I would never get to on my own. It’s utterly different from home, in spite of First World intrusions, and that’s why we travel.

Posted in Uncategorized | Tagged , | Leave a comment

High-Tech Faddishness

Posted on May 10, 2026 by jlredford

I happen to have been at several chip conferences recently, and it’s depressing to see how quickly high-tech moves from one fad to another. AI is the current fad, of course, and was a dominant theme at all of them. I’ve seen real value in it for chip design and other tasks with a lot of rote work in them, but it’ll never make back the trillions (yes really) that have already been invested. When the similar dot-com bubble burst, there was at least some value in all the fiber optics that were laid, but that won’t be true this time. The chips that people are spending so much money on today will be obsolete in a year, and not worth the electricity to run them. New software schemes can offer far more improvement than even the hardware improvements, so the chips will be doubly useless if they can’t run the latest code.

Anyway, AI is just the latest in a long series of fads. The next one is already building – humanoid robots. The current demonstrations are cringey, where the robots are obviously human-piloted, or are failing at simple tasks. Their backers promise that they’ll be useful Real Soon Now, but it’s hard to see how they’ll be more efficient than robots customized for particular tasks.. The last fad was for eVTOLs, drones big enough to carry people, an extremely terrible idea given the existence of weather. Before that it was crypto, and hydrogen-powered cars and trucks, and small modular nuclear reactors, and low-payload rocket launchers, and self-driving cars, and the Internet of Things. There is generally some real value here (although none in crypto beyond crime), but never what is promised.

The bigger fads, though, are more concerning. After George Floyd’s murder by police in 2020 and the subsequent riots, there was a real corporate push for diversity. This wasn’t simply white versus brown because high-tech is already majority non-white – it’s dominated by South Asians (mainly India and Pakistan) and East Asians (mainly China, Japan, and Korea). The real shortages were for Blacks and Hispanics, and especially women. I see only 10-20% women at these conferences. That’s less than there were when I started my career in the 80s, incredibly. So about 2/3 of the US population is under-represented.

This is obviously bad. These companies are ultimately selling to consumers, and 2/3 of the consumers aren’t working at them. They aren’t going to build the right things. Look at the widespread public hostility to AI – its creators completely blew their branding. They’re also missing a huge pool of talent. There’s nothing magic about these fields – it’s material that anyone can master if they work hard and think straight. You don’t even need advanced math except at the very leading edge.

Two years ago this lack of diversity was a common subject at these conferences, but now it’s nowhere. This is partly due to the hostile attitude of the Trump administration, but also because of the fickleness of tech leadership. A new shiny object came along and they all surged toward it.

The other serious issue that they’ve now dropped is climate change. Here again the Trumpists are partly to blame, but as Philip K Dick once said “Reality is that which when you stop believing in it, does not go away.” Physics doesn’t care about your political feelings. There are huge opportunities in the renewable energy transition, but they are nowhere at these shows. The chip opportunities are in EVs and power management more than in computing. They’re also in corners of the field like the high voltage DC transmission lines that are now re-wiring the world. There’s also big value in high-temperature electronics for nuclear and deep geothermal. Outside of electronics, it’s a golden era for chemistry as people try to make better batteries, and try electrical ways to make the fundamental materials of the modern world – steel and concrete.

So there’s lots of corporate value in both diversity and climate initiatives. Unfortunately, the wild success of Nvidia makes corporations think that there’s lots more in AI. They can’t walk and chew gum at the same time, so the bigger issues get neglected. No wonder tech is now considered the domain of scamsters and greedheads. Kowtowing to Trump does not help.

Posted in Uncategorized | Tagged , , | Leave a comment

Jiahu: An Actual High-Tech Atlantis

Posted on January 8, 2026 by jlredford

I recently learned about an extraordinary archeological discovery, that of an ancient high-tech culture that really was drowned in a cataclysm.  It just didn’t happen to be the one Plato was thinking of when he described Atlantis.   He was probably referring to the Minoan towns destroyed by the eruption of Thera (modern Santorini) in about 1600 BCE, but this was a place vastly older and farther away – the Neolithic Chinese town of Jiahu:

Excavation of Jiahu, showing skeletons. Image from a travelogue in iNews

The site is in central China, Henan province, and is 9000 (!) years old. It’s on the plains between the Yellow and Yangtze rivers. It was discovered in the early 1960s, and has been under intermittent excavation ever since. It’s now a tourist attraction for the town of Wuyang.

So what makes this high-tech? At a time when almost all societies were still nomadic hunter-gatherers, this place had:

  • Agriculture – the first evidence of rice cultivation in China, and also millet.
  • Alcohol – pots were found with fermented rice, honey, grapes and hawthorn residues. Dr. Patrick McGovern at U Penn figured out the recipe and Dogfish Brewing in Delaware made it as Chateau Jiahu. This is the oldest alcohol recipe known.
  • Music – About 30 bone flutes, generally in a pentatonic scale. They were made from the wing bones of the red crown crane, an Asian bird admired to this day for its elegance but now endangered. They’ve been able to date the flutes and see a progression in quality and tonality. Tech advanced even then. More here from a curator of their exhibit in the Henan Museum: Jiahu Bone Flute of Wuyang.
  • Writing – There are tortoise shells with symbols carved on them. No one can read them, but they are grouped and the pictograms do recur. Understandable writing only came many millennia later.
  • Domestic animals – pigs, poultry, dogs, and even some cattle. A pond was used to raise carp. Food was plentiful and good.
  • Defense – the village was surrounded by a moat. The human skeletons found do not show signs of violence. They lived to a decent age too – maybe 40 on average. They appear to have been an egalitarian culture – there are not many differences in grave goods among them.

Just imagine being a primitive hunter slogging across the plains and coming upon a place like this. A place where people lived long and peaceably, ate well, and spent their time drinking, making music and crafting written prayers to the gods. They doubtless would not let you in, but it must have looked like paradise.

Playing a restored bone flute,

The village appears to have had between 250 and 800 people in it at a time. It lasted for 1300 years until a flood from a nearby river drowned it in 5700 BCE. It was evacuated in an orderly manner – no tools or household effects were left behind. The flood and its sediment is undoubtedly why it was preserved, rather than reworked by millennia of later settlement.

It didn’t have the trappings of the Tomb Raider Atlantis – lasers, god statues with glowing eyes, levitating pyramids – but it sure looked sophisticated. There is as much time between us and the Great Pyramid, 4500 years, as there is between the Great Pyramid and Jiahu. There was a lot going on very early that we know little about.

Posted in Uncategorized | Tagged , | Leave a comment

Actually Made in America

Posted on January 2, 2026 by jlredford

Most of the stuff in your daily life, of course, is made in the US. That’s groceries, and gasoline, gas and electricity, and construction materials like lumber, bricks, glass, and drywall.

Yet your clothing almost certainly isn’t, nor are most manufactured goods. Even for major goods like cars and planes, a big fraction is made elsewhere. E.g. for the Boeing 737 MAX, a bit under half is foreign-made, and that includes the engines from LEAP in France.

This is unsustainable. It’s been great for the 1%, but everyone else needs work. It’s going to be tough to bring it back, though, and I recently came across two examples of why:

American Roots

This is a clothing manufacturer near Portland Maine. I have one of their excellent hoodies:

It was founded by a couple, Ben and Whitney Waxman, who had been in politics. They made a point of sourcing every single component in the US: the fleece and cotton materials, the drawstrings, the thread, the zipper, and the printing. Most of the equipment for this is gone, as is the expertise in running it. They found a few old-timers to re-create it, but that was difficult and expensive.

They were also determined to do right by their workers, and they’re unionized. They have training programs to make people efficient, and offer flexible work hours. That turned out to be critical to hiring women.

The trouble is that they can’t find many native Americans to do this tedious work for $18 an hour. Their workers are almost entirely immigrants. They come from a wild range of places, from Turkey to Central Africa. It’s a good deal for them, as this is a lot of money compared to their home countries, and their bosses are not the usual crooks, but it’s not that great by US standards.

Their difficulties are well-described in “Making It In America” by Rachel Slade (2024). She follows the Waxmans from 2015 as they search for suppliers, battle anti-immigrant sentiment, and deal with many setbacks.They were hit hard in 2019 by bad quality inputs, and then by the pandemic. They switched to making masks, but got hit again when Chinese imports started up again.

Their saving grace was that they largely sold to unions and non-profits. Any time an organization needed logo-branded apparel for a convention or merch, they could supply it in bulk. They do now sell retail, and that’s where I got mine, and are thinking of opening their own stores. They currently have about 75 workers and have revenue in the low millions, although that’s hard to determine. That’s not that much after ten years, but I wish them all the luck.

The Smarter Scrubber

I came across this in a Youtube video on the excellent engineering channel SmarterEveryDay by Destin Sandlin:

A friend of his, John Youngblood, went through the common experience of inventors – he found a problem with the existing way things were done, and devised a better solution. In his case, the problem was with barbecue grill scrubbers. They’re made as wire brushes to scrape off the baked-on grease on a grill. Unfortunately, the steel bristles break off and can get in one’s food. If you swallow one, it’s bad news as it perforates your digestive tract.

So instead, he thought to make the scrubber with chain mail. Now the wires are tightly bound and won’t break. The head can also be run through a dishwasher to clean it. This is the kind of Aha! moment that every engineer hopes to feel. Sandlin loved the concept and thought to try an experiment – could this be made entirely in the US? He had been dismayed when his home state of Alabama was completely unable to make PPE during the pandemic, except for one old guy who knew how to make injection molds.

He also wanted to make it in the US because Amazon encourages Chinese manufacturers to rip off American products. They tell them exactly what is successful and put the knockoffs on the same search page. Because the quality is worse and they have no development expenses, they can undercut the inventors. Amazon then ignores patent infringement claims. Don’t ever buy anything from them.

Sandlin does a nice job describing the whole design process for the scrubber. He had to try a lot of different pads and schemes for attaching the mail to the head and patterns for the chain mail. Finding a US supplier was hard and they couldn’t get large quantities. The simplest and most standard part, a 1″ 1/4-20 stainless steel bolt, turned out to be the hardest of all. Chinese suppliers sell if for 9 cents, and the best he could do was a Massachusetts supplier for 35 cents.

What really depressed him was discovering that he couldn’t even make the molds for the head parts in the US. All that skilled work is now in China too. The US bosses fired the expensive mold makers in favor of cheaper Chinese ones just like they fired US workers. He ended up doing it himself, with a lot of help from old experts. That was enjoyable, but again not sustainable. One ray of light is that there are new ways to make smooth metal molds with 3D printing. Yet there’s still a lot of lost know-how in what the molds actually work.

So here’s the final product:

Smarter Scrubber from JJ George

$75! And on back-order. That’s a lot compared to a standard brush, but it won’t, you know, kill you.

Sandlin put a lot of effort and money into this minor product, far more than any ordinary venture would. It would have been a lot easier and cheaper to do in a place with a working industrial infrastructure. He could do it because he has 12 million Youtube subscribers.

Yet he’s also doing it to make a point. It’s going to take a lot to undo the damage the oligarchs have done to the US economy. Trump blathers about bringing manufacturing back, but that’s all a pitch for the rubes – his actual actions have been entirely counter-productive. It’ll take clear and concrete actions like Sandlin’s and the Waxmans’, and therefore proof to politicians about how to really help.

Posted in Uncategorized | Tagged | Leave a comment

Give ’em a Short, Sharp Shock and They Will Totally Do It Again

Posted on November 25, 2025 by jlredford

We’re watching the tremendous Ken Burns PBS series “The American Revolution”. It stresses that the British tried to hit the Americans suddenly and hit them hard. They would disperse this disorderly rabble with a whiff of grapeshot. They burned everything they could reach along the coasts, and held New York, Charleston and Savannah for the duration of the war. They saw that all the colonies were at each others’ throats and thought they would fall in line once they saw the danger of their treason.

The Royal Navy attacks Long Island on the way to taking New York, 1776

Well, that didn’t work. The opposite happened – the colonials all realized that the British were their true enemies and put aside their differences, at least briefly.

This strategy has failed against the US again and again. The Confederates thought that the greedy and mercantile Yankees would sue for peace once they took Fort Sumter rather than risk their profits in the Civil War. The Japanese Empire thought the same thing when they destroyed almost all of the US Pacific fleet at Pearl Harbor. That display of valor would surely cow the fat and dishonorable Americans.

It’s also what Trump and Musk thought this year. Within days of getting into office they were destroying federal agencies. It was wildly illegal, but they wanted to overawe the opposition with a demonstration of power. Musk seized the federal IT resources while Trump seized all the police power he could. He set up concentration camps and has masked goons grabbing people off the street. He threatened universities and news organizations and extorted them for hundreds of millions. He talked blithely of third terms and military occupation of blue cities.

That too has united the opposition. The October 18th No Kings protest was the largest in US history, and occurred in every state. The November 4th elections were the first since his ascension and were wipeouts for the GOP. The legal system has ruled against him now dozens of times, even though the corrupted Supreme Court still rules in his favor. It’s got to go.

The Democrats have set aside their endless squabbling over appeals to the working class versus appeals to marginalized groups to focus on the real threat. Even the GOP is cracking. They backed him as an outsider candidate who was not as tarred by their massively unpopular policies as Mitt Romney was. They liked railing against neoliberals who lose wars and export jobs, but now find themselves backing a pedophile who is stealing everything not nailed down.

The British obviously should have set the colonies against one another. They could have offered representation in Parliament to Virginia and New York while isolating restive New England. The Loyalists in both places would have approved, and probably sent troops. The British couldn’t do that at the time because they were too proud and upset. Yet the Revolution was such a disaster for them that they learned that lesson well. They applied that kind of divide-and-conquer strategy in India and ruled the sub-continent in spite of being out-numbered a hundred to one.

Trump and Musk are also too blinded by pride and hatred to be smart about their takeover. They’re seeing the pushback now. Let’s hope they continue to fail, but that the conflict doesn’t become as brutal as the Revolution!

Posted in Uncategorized | Leave a comment

Little Free Libraries – Boon or Blessing?

Posted on November 10, 2025 by jlredford

I was out taking my constitutional the other evening when I noticed a new little free library. There are five of them within just three blocks of my house:

Hmm! I see from the Little Free Library site that about 200,000 of them have been installed around the US since the program began in 2012. People register their sites with them for a fee so they appear on their finder app.

They say that on average one book is borrowed from each library each day, which would be about 70 million a year. That’s a visible percentage of the ~800 million book copies sold each year, and is even a few percent of the ~1.5 billion books lent by the 9,000 public libraries in the country. My town of Arlington may have a lot of LFLs just because it’s especially literate – our own public library hit one million items checked out in FY25, which is ~25 per person in the town.

The site sells kits for them, formerly of wood but now of a more durable composite for about $440. That seems like a lot, and as you can see from the above a lot of people make their own. I especially like the one from George Hart, an old college housemate and now free-lance mathematician and sculptor:

Taken from “The Joy of Zonal Polyhedra”, 2021

He came to this style through his work on the geometric construction toy the Zometool. It has been used in architecture and is a favorite at Burning Man.

Now, one reason these libraries are popular could be because people are transitioning from paper books to digital ones. Books do occupy a lot of home space, although they do furnish a room. A lot of the items I see in these libraries are old textbooks and school workbooks. Those are entirely digital these days. Yet e-books and audio books are still a small part of the publishing world, accounting for only about 20% of revenue. Print book unit sales continue to rise and had a surge during the pandemic.

Yet books can’t be kept forever, and you hate to just throw them away. Most are still readable, so someone should get some use out of them. It’s not like throwing away a toaster. Books are art, and the purpose of art is to affect the emotions, so it’s no surprise that people have fond feelings for them. They want someone else to enjoy them the way they did.

Another similar service has grown up recently – Little Free Pantries. They contain non-perishable food items where people can take what they need and leave what they don’t:

Same style of hutch, different product. It’s sad that this wealthy country needs these, but here we are. The organization promoting them is much more casual than the LFLs. They appear to be tracking about 2300 of them, and they’re often paired with LFLs.

Let me finish by pointing you to a recent New Yorker cartoon – The Tiny Library of Babel by Adam Douglas Thompson. As in Borges’ original story, books are a road into infinity! Even these small outposts can lead one endlessly on.

Posted in Uncategorized | Tagged | Leave a comment

From Creepy DARPA Concept to Useful Widget – A Long But Typical Engineering Journey

Posted on October 20, 2025 by jlredford

The standard engineering story goes like this – someone thinks of a good idea, they found a startup to develop it, it gets backed by venture capitalists, it becomes a hit, everyone gets rich, and the field moves forward.

Boy that doesn’t happen often! I recently came across a much more typical journey. It was a lot more tortuous, and it took a really long time, but it does have a happy ending. It’s the path of low-power mesh networking from concept to craze to finally useful product. I heard one of the principals in this whole project, Lance Doherty, speak about his almost 30-year journey to actually make this work.

So here’s how this story went:

1997, Proposal. Prof Kristofer Pister of UC Berkeley comes up with a cool idea – Smart Dust. He could see that chips were getting steadily smaller and cheaper, and foresaw a day when they could be scattered around like grains of rice. Each would have a sensor and a tiny radio with just enough power to send a signal to another dust mote nearby. That could pass it on to another mote and another until it reached some useful destination.

He gets funding from DARPA, the research arm of the Pentagon. For obscure historical reasons they’re the only federal agency that funds research into chips and computers. They have a creepy purpose in mind – they want to track every soldier and vehicle on a battlefield, the better to kill them.

2001, Demo – With the help of ace grad students like Doherty, they get a system working. It uses commercial parts and so is about 1″ on a side, but it can be deployed by a drone and dropped on the sides of roads to watch for vehicles:

It has a magnetometer that can detect the steel in a vehicle by the way it changes the Earth’s magnetic field, and can pick up cars 5 m away and trucks at 10 m.

2004, Startup – They spin it out into a startup, Dust Networks, with venture capitalist money. Joy Weiss comes in as CEO. They figure out the key to the whole concept – each node must be time-synchronized to all the others so that it only has to turn on its power-hungry radio receiver when another node is likely to be transmitting. Leaving the receiver on all the time will kill whatever power is available.

They ride a rising wave of a new Valley fad, the Internet of Things. Just as they did with computer networking and later autonomous driving, the Valley follows DARPA’s lead. They think that the Internet will now be absolutely everywhere. They start adding processors and WiFi to refrigerators and washers and stoves. The high point is when Google acquires Nest Labs and their smart thermostat in 2011 for $3.2 billion. They envision acquiring data about absolutely everything in people’s homes and lives, the better to sell to advertisers.

That fails. The thermostat market is tiny and easily served by much cheaper widgets. No one wants talking refrigerators that creepily track your grocery consumption. No one trusts Alexa voice recognition boxes in every room, and rightly so.

Dust Networks also goes down around this time. They had some success in wireless networks in factories, with an installed base of maybe 20,000 places, but that’s not actually all that many for a chip company.

2011, Acquisition. They get bought by a strong Valley analog chip company called Linear Technology. Linear had done extremely well by building radio chips, analog-to-digital converter chips and power management chips, ones that change one DC voltage efficiently into several others. These are needed everywhere. They explicitly avoid getting into price wars for commodity parts since that drops profits to zero. Instead they aim for steadily higher performance and better features, and that lets them do fun engineering.

They discover an ideal use for these low-power networks – electric vehicle battery packs. These contain dozens of battery cells and each one of them needs a control and monitor chip. The chips can tell how much charge a battery holds, what temperature it’s at, and most crucially, whether it’s going to catch fire. Lithium battery fires are horrible. They’re very hard to extinguish and emit toxic smoke. If the battery is mis-charged, it can heat itself up so much that it goes into thermal runaway, and you have very unhappy car owners, assuming they live.

Running wires to all these control chips is a nightmare for the wiring harness, for labor, and for reliability. So what if they could all be connected by radio instead?

Complicated wired scheme on left, simple wireless on right

You don’t need to send a lot of data back and forth, just enough to know what each battery is doing. The connection has to be extremely reliable, and the power draw has to be minimized since it’s running on the battery itself. Yet this lets you connect the chip to the cell when it’s first made and then follow it for its entire life. You can tell what happens to it in the factory, in the truck that delivers it, when it’s installed in a car, during the car’s lifetime, and then in whatever other use the battery is put to after the car is done. You can know the cell’s capacity exactly instead of having to build in some margin in case of mis-measurement. That alone lets you get another few percent out of each cell. Plus you save the weight of the wiring, and can reconfigure the packs easily for different models.

By 2016 they have installed these wireless Battery Management Systems (wBMS) in an electric BMW i3 and driven it all over the place. It works great! They show it at the 2016 Electronica conference and it’s a smash. They think “Ah, it’s working at last.”

2017, Productization – That’s when reality kicks in. Proving that the radio data will get through in every possible circumstance is really hard, but car companies won’t accept it without thorough testing. Linear doesn’t have the resources to do it. Yet their Massachusetts competitor Analog Devices does. They buy Linear in 2017 for $14.5 billion with about 25% in stock and the rest in cash.

They set up a testing facility at the ADI plant in Limerick, Ireland. It’s got a turntable that can hold a whole battery pack and orient it every which way. They surround it with radio antennas so they can blast it with every kind of interference.

It takes years to work out all the possible failure modes and needs engineering groups from all over the world. Yet then GM agrees to use it! They had some bad fires with the first Bolt model and were being extra careful for the next ones. They initially called it the Ultium battery system, but that was too hard to explain and they’ve dropped it. GM starts shipping it in 2021.

A few Ultium models

At the same time they’re approaching all the other car companies. Now that one major company has accepted it, it should be fine for everyone, right? Ha. This company wants the controllers to be on all the time in case the batteries overheat in a sunny parking lot. Yet it all has to work on the 12V lead-acid battery, since everything else in the car is designed around that. If it sits there for a month, the idle power will kill the 12V battery, so nuh-uh. That can be fixed in software, thankfully. This other company wants to get a lot more bandwidth out of each controller to monitor it more closely. Yet another wants the maximum level of safety for all the chips and the design process that goes into them.

Every new user has new demands. It’s wasn’t until they got to the seventh customer that they could actually use the system as is. That didn’t happen until this year. There are bound to be further glitches as the scheme spreads, but they’re finally off and running.

So from initial concept in 1997 to shipping in 2021 is 24 years. That’s a good chunk of a career! It’s been a long road. Yet it’s saving lives and easing the electrification of cars, and ultimately helping the planet. It didn’t work to find soldier’s footsteps (it’s not used in Ukraine even now), and it didn’t work to network toasters because no one cared, but it finally found a place where it’s really useful. Let’s hope they now find some more!

Posted in Uncategorized | Tagged , , , , | 1 Comment

The Odd Longevity of Pink Floyd

Posted on October 11, 2025 by jlredford

Our town has a theater, the Regent, where a lot of tribute bands come to play. They do covers of Elvis, Sinatra, and the Eagles, which are unsurprising. They were all massive stars with the usual songs about falling in and out of love.

Yet they also have Pink Floyd cover bands regularly, and I saw one last weekend – Floydian Trip:

Laser light shows and everything!

None of these musicians were even born when “Dark Side of the Moon” came out in 1973, yet they could perform it precisely. Their female singer (and band lead) Stephanie Jones could even do the extraordinary wordless wail from “Great Gig in the Sky”. I happened to be sitting next to her father, who had driven down from Maine to see them. She got a music degree from Yale, but now works in Manhattan doing HR at a finance firm. The band does dozens of these gigs a year. So she’s a typical Millennial who happens to be completely taken by Boomer prog-rock.

I am a Boomer, and I do remember hearing Dark Side when it was released. I was driving home late one night with my dad when a radio station played the whole second side. We got home in the middle of the side and I found the station just to hear the end. It caught my ear even over a car radio. A while later in college I had a housemate who would get stoned and listen to it every single night. Floyd and Springsteen were the soundtrack of my college years.

The early Floyd is unlistenable (at least to me!) and they disintegrated after “The Wall” (1979). Their main material was only four albums (“Dark Side”, “Wish You Were Here”, “Animals” and “The Wall”) over 6 years. That’s fewer albums and less time than even the short-lived Beatles.

So this is music from 50 years ago. There isn’t a single love song. It doesn’t even have the rage of punk or the violent impulses of metal. Yet Dark Side is still in the top-selling album lists. Small cover bands like this play it all the time, and large ones like Australian Pink Floyd fill big venues. What’s going on?

I think it resonates because it’s talking about cracks in the world. Their founding genius, Syd Barrett, fell into those cracks via schizophrenia. Something broke inside him, and they see similar breaks all over. In “Brain Damage” on Dark Side, they sing that The lunatic is in my head, but the lunatics are also in the hall, where The paper holds their folded faces to the floor. In “Comfortably Numb” on The Wall, they sing:

When I was a child I caught a fleeting glimpse
Out of the corner of my eye.
I turned to look but it was gone,
I cannot put my finger on it now,
The child is grown, the dream is gone.

That same theme of loss permeates Wish You Were Here, where success made them trade A green field for a cold steel rail, A smile for a veil. It’s in “Time” on Dark Side, where Every year is getting shorter, Never seem to find the time, even though the Voice of God in the form of David Gilmour’s guitar riffs is telling them to wake up and get on with it.

Then there’s the ending of Dark Side, “Eclipse”, the biggest climax in pop music, beating even “A Day in the Life” on Sgt Pepper. Roger Waters sings about how All that is now and All that is gone and All that’s to come and Everything under the Sun is in tune, But the Sun is eclipsed by the Moon. Everything seems to be going along fine, and then the shadow falls. Who hasn’t had such an experience? That’s almost as universal as falling in and out of love. No other band captured it in quite the same way. No wonder they’re still getting covered.

Posted in Uncategorized | Tagged | Leave a comment

MIT Under Attack

Posted on July 24, 2025 by jlredford

Last May 31st I attended Technology Day at MIT, where the school presents talks on all kinds of interesting research to its alumni. Commencement is always the day before, and the alumni reunion events are that evening. I’m a big fan of these talks and have written about previous years here: 2024, 2023, 2019, 2017, and 2011, Videos of the three sessions can be seen here:

Yet this year it had a much darker tone. The Trump administration is explicitly trying to destroy US universities and scientific organizations. They are canceling research contracts, defunding science agencies, denying visas to international students and even threatening to remove accreditation. Even the research that is kept on will have its institutional overhead (“indirect cost”) cut from a median of 56% to 15%, meaning the federal gov’t will pay almost nothing for the facilities of the university that make the research possible.

At Trump’s insistence, Congress just raised the tax rate on leading university endowments by six times, from 1.4% to 8%, significantly damaging their finances. Former MIT president Rafael Reif wrote about the effect of this tax in the Boston Globe: Big Beautiful Bill Will Raise the Cost of College. The tax is as large as MIT’s entire undergraduate financial aid package, which currently grants free tuition to any student with a family income of less than $200K. Incomes of less than $100K get free board as well.

So how did MIT respond on Technology Day? The first session was devoted to fascinating medical advances. The first speaker (11:03) is looking for ways to detect tiny ovarian cancer tumors using optics and AI. The second (19:39) is planning how to rework the US insurance system for greater equity. The third (28:36) has made extraordinary progress on robotic prosthetics, and lost his own legs in a climbing accident. The fourth (36:44) has new attacks on the ancient and deadly disease of tuberculosis, against which we only have one fading vaccine.

They all talked about how the funding cuts were ruining their plans. At the very end the moderator asked “What’s the one missing piece that would aid your research?” One said flexible funding to pursue strategy changes and another thought there should be more focus on therapeutics across a diverse community instead of research. Yet the last to speak, Prof Bryan Bryson, said:

What we need is you. We need you to go out and talk to everyone about the wonderful things you’ve heard here. Academia is not some ivory tower – it’s attacking problems that affect everyone. Someone in your life has had TB or cancer or a terrible accident. When people ask ‘What do they do over there?’, say ‘They help you.’ Go out and tell twenty people about the amazing things you see here.

So that’s what I’m doing!

Yet the leader of MIT, President Sally Kornbluth, was not as inspirational. In the third session above, she is interviewed by the head of the alumni association, Natalie Lorenz Anderson, ’84. At 25:22 she asks Kornbluth very gingerly about “How do you think MIT should be positioned to meet the challenges of the day?” She says that in the worst case scenario, they could see their budget cut a third (!). In that case, she would cancel some programs outright to favor maintaining others – “We will not spread the peanut butter thinly.” They have joined a group of 14 universities to sue to prevent the new indirect cost rate from going into effect and so far have gotten a restraining order to prevent the change.

What struck me, though, was how accepting she was of this situation. This country is in the midst of a right-wing, foreign-backed coup. A concentration camp has been built in Florida, and masked, armed police are kidnapping people off the streets, including students. The president is openly senile. Congress is openly giving away the federal government to their donors. Yet she still keeps going down to Washington, telling people that MIT is crucial to the medical and defense and technical innovation pipeline. If they cared about any of that, they never would have started down this road. If they cared about the general good of humanity or America, they would not have destroyed millions of tons of food aid for Africa, and millions of doses of vaccines. She cannot accept that these people consider the likes of MIT to be the enemy. She was attacked personally on Dec 5, 2023 when she was hauled in front of a Congressional committee headed by Rep Elise Stefanik. They called her an anti-semite for not crushing the student protests against Israel’s horrific treatment of the Gazans. The attack succeeded in ousting the presidents of Harvard and U Penn. Yet she still persists in talking to them. She cannot think of anything else to do.

She is bound by the constraints of her position, just like the helpless Democrats. They can only work within the system even as it is failing in front of them. What it will take is general public disgust and revolt. That’s what Prof Bryson was asking for. These people are stealing your present in the form of federal support for public goods, and stealing your future when they go after the likes of MIT.

Posted in Uncategorized | Tagged , , | Leave a comment

“Whoosh” – Autonomous Gondolas vs Robo-taxis

Posted on June 21, 2025 by jlredford

A New Zealand-based company, Whoosh Solutions, is building a new kind of urban transit system – a network of overhead fixed cables that are used by self-driving battery-powered gondolas to go from point to point. They can switch from one cable to another, unlike regular gondola systems, and so can take any route through a network to get around without having to change cars or stop. They let people on and off at stations every few blocks at ground level:

CGI street render from Swyft Cities

They’ve built a quarter-scale model in the resort town of Queensland NZ in the middle of the South Island, and are looking to build a full system there. They’re also looking to build a system in the Houston suburb of Sugarland. They were picked by the Google Transit team as a favored solution, which was then spun out as Swyft Cities. They’ll build the systems while Whoosh develops the tech.

The cars can hold up to 5 people, and so can take a family or someone with a lot of luggage or delivery packages. They’re a minimum of 9 meters above the ground, and so have lots of clearance over the streets. They run at up to 30 km/hr and go directly to a destination, so you can get there in a similar time to a car in a city. The cars are much safer than ground cars, which have to deal with drunken or distracted drivers, and kids running into the street, and the occasional animal.

So this has a lot of advantages over other mass-transit schemes:

  • The cable towers and stations take hardly any land and far cheaper to build than light rail or even bus lanes. Whoosh estimates it at $5M / km vs $100M / km for rail. Boston would love to be able to build for $100M/km! Multiply that by 10 for New York. The towers are typically 150 m apart, and can be more in sparse areas. The cables can easily cross rivers and gorges.
  • That also means they can be built quickly, like in years instead of decades. The system can grow to reach all kinds of areas now badly served by transit.
  • Unlike a gondola on a fixed cable, the number of cars can vary with demand. A lot can be deployed at rush hour, and they can be taken offline at night for maintenance and recharging.
  • They can come on call like a ride-share, instead of at fixed intervals like a bus or train, decreasing transit time and increasing utilization.
  • The car entrances are at the same level as the platform, so there’s no step up or down. This is great for accessibility. This is a constant problem with buses, which have always had trouble with wheelchairs, and for regular cars for that matter.
  • That also means that you can wheel hand trucks into them for deliveries.
  • They can run in all weather, even when snow blocks roads. Queenstown is a big ski resort town. High winds are bad though.
  • You get great views as you travel! Being in subway tunnels is always oppressive, and a bus is little better.
  • It’s even more energy-efficient than an electric vehicle (EV) – the cars run continuously without many stops and starts for traffic, and the rolling resistance on the cable is low.
  • Like gondolas, they can handle elevation changes easily, like mountains on a city’s edge. Trains can’t handle height changes at all, and even roads have problems. This is why many South American cities like Medellin, La Paz, Caracas, and Rio have already built big gondola systems.
  • It’s riding the global trend towards EVs. All the same battery, inverter, and motor tech can be applied. The wheels are just on a cable instead of asphalt. The cars should ultimately cost only tens of thousands, and can be upgraded easily over time.

Yet there are disadvantages too:

  • The capacity will be much lower than a train. A typical train can hold 200 people and arrive 10 times per hour for 2000 people per hour. Here there are only 5 people per car max, so at one minute per car that’s only 300 people per hour. That’s about like a bus. This was also the problem with the small tunnels of the Boring Company. Pods just can’t hold as many people as long cars.
  • Many people don’t like heights and won’t go on something that swings in the air.
  • This is new tech from a tiny company in a tiny country, and so lacks scale. Everybody builds buses and subway cars.

The biggest threat to it, though, is probably robo-taxis. Vast amounts are being invested in this by tech overlords who see automobiles as the next grand market to conquer, now that they’ve seized finance and advertising. A hundred million households in the US spending $500/month on their cars represents $600 billion. That’s why Waymo has put tens of billions into their autonomous system, and that’s what props up Tesla’s valuation at its ludicrous price-to-earnings ratio.

Yet Whoosh has some advantages here:

  • They’re much safer. They don’t have to deal with all the random events of the street. Even the most highly tuned LIDAR/radar/V2V system is going to kill people who wander around. We only accept this now because there’s little other way to get around easily.
  • They don’t need roads, which suck. They cut up neighborhoods, destroy green space, are noisy, and are hugely expensive to maintain. They’re constantly damaged by trucks and frost heaves and sewer/water work. Big fast roads like highways are a menace to humanity.
  • They’re green from the start, without the legacy fossil fuel systems that have poisoned landscapes and now the atmosphere.
  • They’re quiet! Even EVs make a lot of tire noise, while these have rubber wheels on cables high in the air.

So will this go anywhere? The Google interest is promising, but the hype around self-driving is overwhelming and the entrenched auto industry is hard to overcome. They killed street cars long ago, after all. This might be yet another tech that gets more traction (ha!) outside the US than in it.

Posted in Uncategorized | Tagged , , | 1 Comment