Tuesday, March 21, 2023

Godzilla vs. Rodan

You can't really root for either, but it's still fun to sit back and enjoy the spectacle.




2022 was a remarkably bad year for conventional wisdom pretty much across the board from the Red Wave to Russia's inevitable conquest of Ukraine to DeSantis eclipsing a rapidly fading Trump.

Marshall is being sarcastic here, since the rest of his timeline has been prominent Republicans and conservative influencers defending Trump and often attacking DeSantis. It's true that the indictments have turned up the heat, but the underlying trends we've been seeing for weeks haven't shifted. 

We all knew that Trump wasn't going to be a good sport and go away quietly. His relationship with the GOP has always been purely transactional. There was no reason for him to play nice for the good of the party and every reason for him to do everything in his power to get the nomination. Add to that his personal hold over a large segment (possibly a plurality) of the party and a vindictive nature, and you've given non-cult Republicans everything to fear.

DeSantis, on the other hand, has a personal hold on almost no one. His support comes all but entirely from the combination of far-right positions, perceived viability, and a protected spot in the political hothouse, and the last two of those are quite fragile. When the fighting began in earnest, it was obvious he was going to lose that right-out-of-the-box shine quickly.

Perhaps the best guide to the battle is Ron Filipkowski, who reads GOP/MAGA tweets so we won't have to.



A brief note to those making the "they all line up behind the nominee" argument. That was true in the past, and the rest of this post suggests it will be true if Trump wins, but are you really confident usually US political precedent to predict the behavior of this cult of personality?


Remember how the establishment press embraced Vance as an alternative to Trump?





 






One final interesting twist.
Posted by at No comments:

Monday, March 20, 2023

Revisiting the cutting edge world of Tesla

A couple of months ago, the NYT argued that Elon Musk's long history of cutting corners with respect to ethics and safety was simply the price to be paid for advancing potentially life-saving technology.

Some of Musk’s most questionable decisions, though, begin to make sense if seen as a result of a blunt utilitarian calculus. Last month, Reuters reported that Neuralink, Musk’s medical-device company, had caused the needless deaths of dozens of laboratory animals through rushed experiments. Internal messages from Musk made it clear that the urgency came from the top. “We are simply not moving fast enough,” he wrote. “It is driving me nuts!” The cost-benefit analysis must have seemed clear to him: Neuralink had the potential to cure paralysis, he believed, which would improve the lives of millions of future humans. The suffering of a smaller number of animals was worth it.

There was, as we pointed out at the time, a subtle flaw in that argument.

With the complicated exception of SpaceX, none of Musk's businesses are on the cutting edge of anything. In autonomous  driving, AI, solar cell development, brain-machine interfaces, tunneling machines, and countless other technologies where Musk has promised revolutionary disruptions, his companies are, at best, in the middle of the pack and, in some cases, not making any serious effort at all. (On a related note, despite attempts to muddy the waters with creative statistics, Tesla spends far less than any of its major competitors on R&D.)

Now Faiz Siddiqui, writing for the Washington Post, has done an excellent deep dive into how Tesla slipped to the back of the pack in self-driving.

Some Tesla engineers were aghast, said former employees with knowledge of his reaction, speaking on the condition of anonymity for fear of retribution. They contacted a trusted former executive for advice on how to talk Musk out of it, in previously unreported pushback. Without radar, Teslas would be susceptible to basic perception errors if the cameras were obscured by raindrops or even bright sunlight, problems that could lead to crashes.

Musk was unconvinced and overruled his engineers. In May 2021 Tesla announced it was eliminating radar on new cars. Soon after, the company began disabling radar in cars already on the road. The result, according to interviews with nearly a dozen former employees and test drivers, safety officials and other experts, was an uptick in crashes, near misses and other embarrassing mistakes by Tesla vehicles suddenly deprived of a critical sensor.

 ...

Even with radar, Teslas were less sophisticated than the lidar and radar-equipped cars of competitors.

“One of the key advantages of lidar is that it will never fail to see a train or truck, even if it doesn’t know what it is,” said Brad Templeton, a longtime self-driving car developer and consultant who worked on Google’s self-driving car. “It knows there is an object in front and the vehicle can stop without knowing more than that.”

...

After Tesla announced it was removing radar in May 2021, the problems were almost immediately noticeable, the former employees said. That period coincided with the expansion of the Full Self-Driving testing program from thousands to tens of thousands of drivers. Suddenly, cars were allegedly stopping for imaginary hazards, misinterpreting street signs, and failing to detect obstacles such as emergency vehicles, according to complaints filed with regulators.

...

The data showed reports of “phantom braking” rose to 107 complaints over three months, compared to only 34 in the preceding 22 months. After The Post highlighted the problem in a news report, NHTSA received about 250 complaints of the issue in a two-week period. The agency opened an investigation after, it said, it received 354 complaints of the problem spanning a period of nine months.

Months earlier, NHTSA had opened an investigation into Autopilot over roughly a dozen reports of Teslas crashing into parked emergency vehicles. The latest example came to light this month as the agency confirmed it was investigating a February fatal crash involving a Tesla and a firetruck. Experts say radar has served as a way to double check what the cameras, which are susceptible to being washed out by bright light, are seeing.

“It’s not the sole reason they’re having [trouble] but it’s big a part of it,” said Missy Cummings, a former senior safety adviser for NHTSA, who has criticized the company’s approach and recused herself on matters related to Tesla. “The radar helped detect objects in the forward field. [For] computer vision which is rife with errors, it serves as a sensor fusion way to check if there is a problem.”

 Longtime followers of this story will remember Dr. Cummings, fighter pilot and real engineer, This isn't a post on how Musk handles criticism, but if it were, hers would be the first case I'd mention.

 

Posted by at No comments:

Friday, March 17, 2023

Charging for a feature, then charging more to remove that feature... Where have I heard that before?

For those who think the Tesla CEO gets all of his ideas from an old Thunderbirds Are Go! DVD, this clearly comes from a more sophisticated source.






Posted by at No comments:

Thursday, March 16, 2023

Thursday Tweets -- "Residents hope the next sister city comes with a Google search"



Political analysts have been making the same mistake with DeSantis they did with Walker, assuming that the air of inevitability shown in the safe and protected phase of the campaign will hold up when the bullets start to fly. As we've been pointing out since at least August, DeSantis is a weak politician. The unprecedented build up he's gotten from the conservative (and often the mainstream) press failed to hide just how devoid of charisma and personality he is.  His likeliest path to the nomination was to stake out a position to the right of Trump (particularly with the anti-vaxxers), then hope that disease, death, or imprisonment would take out his rival, leaving him to fill the vacuum with his personal void.

Now that the battle is heating up, you can see the oracles start backing away from the next-big-thing narrative, perhaps thinking of how earlier predictions turned out.

 


Quick side note, Tom Bonier has become the political numbers guy to watch.

And, of course, Chait is always worth listening to.

 From a policy standpoint, DeSantis was to the right of Trump on every issue except Ukraine. We knew that wasn't going to last.



 

Putin's unpopularity makes his hold on Fox et al. all the more remarkable.


Of course, there are still those in establishment warrens like the Brookings Institution trying to paint DeSantis as a normal politician, but they aren't getting much traction.
There is no real place for normal in today's GOP.


 

Excellent points from Grossman and Sewer.

 

 The irony of Bannon playing the Hitler card

 

Forget all that stuff we said about the collapse of Silicon Valley Bank, the WSJ has the real scoop.

I also learned BLM = HBCUs
Wokeness is so much better an explanation than incompetence and bad business practices.

Since the WSJ has explained everything, I guess we can skip this article from Linette Lopez.


From crypto to AI

My favored measure is avoiding nature/nurture debates, but this is good too.



And we'll close out the topic with Dean Baker


 Great thread on military technology from our favorite historical blogger.



Dis-disinformation


Cool legal thread (I had no ideas the statute of limitations worked that way).

...


Those cute toddler videos just got a little more depressing.

Any Ag nerds in the audience?

The GOP and Social Security

Also the party of local control and parents' rights

Not often that we praise the Cato Institute.

About a year ago, Josh Marshall set up a couple of lists on the war in Ukraine. They were enormously useful, providing timely and insightful news and commentary from trustworthy sources. Based on that track record, this should definitely be worth your time.









A classic example of foregrounf/background staging of comedy.

Posted by at 1 comment:

Wednesday, March 15, 2023

"Everything you need to know about the wild world of heat pumps" is nowhere near everything you need to know about heat pumps

 



I had such high hopes for this MIT Technology Review article by Casey Crownhart . Finally, someone was talking about an important, ready-to-go technology for addressing climate change and the power grid. Unfortunately, other than a handy introductory explanation of how heat pumps work, we get the same old narratives, biased - - biased in favor of a technology I actually have been pushing, but biased nonetheless - - with all the same old bad habits, getting key details wrong, using questionable examples, making big logical leaps, omitting important problems, and ignoring the most promising aspects of the technology.

The claim that heat pumps don’t work well in really cold weather is often repeated by fossil-fuel companies, which have a competing product to sell.

There’s a kernel of truth here—heat pumps can be less efficient in extreme cold. As the temperature difference between inside and outside increases, a heat pump will have to work harder to gather heat from that outside air and disperse it into the room, so efficiencies drop.

...

There are heat pumps running everywhere from Alaska to Maine in the US. And about 60% of buildings in Norway are heated with heat pumps [Norway is an interesting case but a bad example. It has abundant electricity almost all of which comes from hydro which plays by a very different set of rules with respect to the power grid. -- MP], along with 40% in Sweden and Finland. [Sweden and Finland rely heavily on ground-source heat pumps, but we'll get back to that -- MP]

That first link points to a quite good Washington Post article by Anna Phillips on the increasing popularity of heat pumps in Maine, which is caused less by their advantages in winter (as the MIT piece implies) and more by the need for summer AC in New England due to climate change. Phillips also doesn't assume an argument can be ignored just because it comes from someone associated with the gas industry. 

Again from the WP:

The Maine Energy Marketers Association raised questions about heat pumps’ viability by suggesting they would tax the region’s electric grid. In 2021, ISO New England, the state’s power grid operator, warned of rolling blackouts because of supply chain issues affecting natural gas. Yet the trade group’s president blamed the situation on the state’s promotion of heat pumps.

“Our power grid is not equipped to handle the demand that is now being put on it,” Charles Summers said in a radio interview. Summers said he and his fellow industry group leaders in New England had sent letters to their governors “asking that states pushing so hard toward electrification, pushing complete conversion to heat pumps, just tap the brakes for a few minutes.”

 Strain on the power grid is the main devil in the details of electrification, whether you're talking about heat pumps or EVs, but it receives disturbingly little attention from journalists and commentators. Here is the only mention of the grid in the MIT piece.

 Heat pumps run using electricity from the grid. While fossil-fuel plants still help power grids around the world, renewables and low-carbon power sources also contribute. So with the current energy mix in all major markets, heat pumps are better for the climate than directly fossil-fuel-powered heating, Monschauer says.
[It's a bit a a digression from our primary point but ramping up immediate demand for electricity will almost inevitably shift the generation mix toward natural gas, particularly in situations where solar isn't much of an option such as winter in New England.]

 One of the main disadvantages of using electricity instead of something like natural gas or oil to heat or cool a house is that the power grid is far more affected by spikes. If everyone gets home at 5:45 On a winter's day and turns on their heat pump, or if everyone cranks up the AC on a hot Saturday afternoon, the results can be disastrous for the power grid, ranging from brownouts all the way to cascading power failures. By the same token, if everyone decides to conserve at the same time it really doesn't do a lot of good with electricity compared to something like heating oil.

Particularly with heat pumps, this problem can be a double whammy. A massive cold front/heat wave will cause a huge increase in people turning on the heat while at the same time making heat pumps far less efficient. We've already seen something like this in summer since the vast majority of air conditioners run on electricity. The existing infrastructure simply couldn't handle it if everyone currently heating with gas or oil suddenly went over solely to electricity.

Unless..

What if there was a way that you could get everyone to use their heat pump only when it was somewhere between 50 to 60° outside? 

From the Department of Energy:

Geothermal heat pumps (GHPs), sometimes referred to as GeoExchange, earth-coupled, ground-source, or water-source heat pumps, have been in use since the late 1940s. They use the relatively constant temperature of the earth as the exchange medium instead of the outside air temperature.

Although many parts of the country experience seasonal temperature extremes -- from scorching heat in the summer to sub-zero cold in the winter—a few feet below the earth's surface the ground remains at a relatively constant temperature. Depending on latitude, ground temperatures range from 45°F (7°C) to 75°F (21°C). Like a cave, this ground temperature is warmer than the air above it during the winter and cooler than the air in the summer. The GHP takes advantage of these more favorable temperatures to become high efficient by exchanging heat with the earth through a ground heat exchanger.

As with any heat pump, geothermal and water-source heat pumps are able to heat, cool, and, if so equipped, supply the house with hot water. Some models of geothermal systems are available with two-speed compressors and variable fans for more comfort and energy savings. Relative to air-source heat pumps, they are quieter, last longer, need little maintenance, and do not depend on the temperature of the outside air.

And from Wikipedia:

The US Environmental Protection Agency (EPA) has called ground source heat pumps the most energy-efficient, environmentally clean, and cost-effective space conditioning systems available. Heat pumps offer significant emission reductions potential, particularly where they are used for both heating and cooling and where the electricity is produced from renewable resources.

By comparison, here's the entire discussion of ground-source heat pumps from the MIT piece.

There’s already a wide range of heat pumps available today. About 85% of those installed are air-source heat pumps like the one I’ve described. These come in a wide range of shapes and sizes. But other models—so-called ground-source or geothermal heat pumps—gather heat from underground instead of collecting it from the air.
Just to summarize, the technology with the most promise for reducing the climate change impact of heating and cooling only gets one sentence while the biggest challenge we have to address doesn't even get mentioned. This is more or less what we've come to expect from coverage of global warming stories, which given the magnitude of the problem, isn't acceptable at all.


Posted by at 2 comments:

Tuesday, March 14, 2023

Housing starts and neighborhood character

This is Joseph.

Mike Moffatt is at it again. He looks at the growth in non-permanent residents in Canada from 2014 to 2022 and it has increased from 253,525 to 732,135. In eight years. It was pointed out that this excludes those without papers. It also excludes new permanent residents, intended to be at the 500K mark for the country as a whole (Ontario is 40% of the population). 

It is very hard to get concrete numbers, but the sort of numbers the government gives are small for this type of population growth. Between 1991 and 2021, Toronto grew from 3.9 M to 6.2M people. In 1991 a one bedroom in Toronto was $592 per month ($998 in 2021 dollars). Instead the cost grew to $1,439 for a one bedroom in Toronto in 2021 and $1,527 a year later. 

There are lots of absurd examples of preserving character (the 1940's outhouse is just too scenic) that exist but the central tension is that you cannot maintain neighborhood character everywhere and have the city expand by 60% in population in a generation and keep all of the neighborhoods the same. I mean that could be a goal, keeping things in statis, but if you are pursuing that goal there is an issue here. 

We have transitioned from 28 million Canadians in 1991 to 38 million in 2021 (36% growth), so the whole country has been robustly growing in population over this time period. But, seriously, student permits went from 139,890 to 411,985 over that 8 year period. This tells me schools are using international students to balance their books but that these measures are only going to increase housing scarcity. This answers the question of where the low end housing is going -- students compete for the same housing as low income Canadians. 

Canada is also unusual in that it is a settler nation. One issue that this really brings tot he forefront is the terrible treatment that first nations experience. From Statistics Canada:
In 2018, among Indigenous people responsible for housing decisions within their households, about 12% of off-reserve First Nations people, 10% of Inuit and 6% of MΓ©tis said that they had experienced unsheltered homelessness in the past. The corresponding proportion for non-Indigenous people was 2%.

 A shortage of housing disproportionately hurts the most vulnerable members of society and there is no question that off-reserve First Nations are extremely vulnerable. What use are land acknowledgements if we can't help these people with basic housing in a climate that drops to -30 C (a credible winter temperature in Winnipeg)?  

Compared to this is the question of neighborhood character? 

The real question is if Canada has the will to build housing to match the aggressive immigration targets that it sets. Many states in the United States, including California, have responded to the housing cost spiral and it has been a while since the least affordable housing was in the US. Back in 2017 Vancouver was less affordable than either Manhattan or San Francisco, and the situation has not improved (3 years later, Toronto crossed this threshold too). At this point, corrected for salary, not even rents are cheaper in Toronto. 

This is a remarkable state of affairs that one presumes must come to an end eventually, but hopefully with less human cost. 

Postscript: This twitter thread came out after I wrote this piece but highlights the issues rather well. The one piece I neglected was that the richest parts of Toronto are in population decline as the detached houses becomes less and less occupied. 
Posted by at No comments:

Monday, March 13, 2023

Your Silicon Valley Bank Primer

 As is often the case, the best thumbnail comes from Matt Levine.

And so if you were the Bank of Startups, just like if you were the Bank of Crypto, it turned out that you had made a huge concentrated bet on interest rates. Your customers were flush with cash, so they gave you all that cash, but they didn’t need loans so you invested all that cash in longer-dated fixed-income securities, which lost value when rates went up. But also, when rates went up, your customers all got smoked, because it turned out that they were creatures of low interest rates, and in a higher-interest-rate environment they didn’t have money anymore. So they withdrew their deposits, so you had to sell those securities at a loss to pay them back. Now you have lost money and look financially shaky, so customers get spooked and withdraw more money, so you sell more securities, so you book more losses, oops oops oops 

For those who prefer their analysis in tweet-sized bites, Krugman  has an excellent thread. Here are some excerpts.

...
...
...
...

"Cultivating relationships" is a bit of an understatement.

Josh Marshall has also been keeping an eye on the story. 

One thing that has come out of my exchanges with readers is unclarity about just what counts as a “bailout”. In one sense it is in the eyes of the beholder: a bailout is when someone else gets it. The more important thing is that term has no real or precise meaning. In general it’s any radical departure from the existing legal/contractual set of rules and obligations in response to a financial crisis. Websters defines it as a “rescue from financial distress.” Clearly the shareholders, i.e., the owners, of Silicon Valley Bank should be wiped out entirely or at least be last in line for any proceeds if the bank goes through a liquidation. They had a business and it failed.

The operative question is the bank’s depositors. What’s being discussed is whether the FDIC should step in and guarantee all the deposits rather than just the 2% or 3% which the FDIC guarantees up to $250,000. What complicates the question is that SVB did very, very little consumer banking. It mainly held deposits of venture capital funds and the start-ups those funds invested in. As noted earlier, the bank likely has enough or close to enough assets to cover all deposits. What it lacks is time and liquidity. So the best solution is for the bank to be purchased by another bank which has time and liquidity. Depositors are protected; risks to the broader economy are prevented; employees at the businesses which banked there have their paychecks covered. That is certainly what regulators are trying to accomplish right now. The question is what happens if they can’t. What basically all the high profile Silicon Valley notables are demanding now is that the FDIC back all the deposits. That’s a bailout by any definition. The point I tried to make above is that such a move could only be justified by grave risk to the broader economy. And that’s a factual question that regulators at the Fed, the FDIC and the Treasury will have to answer.

Still somehow it doesn't feel like a true Silicon Valley disaster until someone from the PayPal mafia makes an appearance.





And for those who are here mainly for the outrage...


 

And schadenfreude.



  And another member of the PayPal mafia pops his head in.


 

 Now, if we could just get three members of the PayPal mafia...






 



via GIPHY



And just as we were going to the presses...


Posted by at No comments:

Friday, March 10, 2023

Deferred Thursday Tweets -- Dems in Disarray and other favorites

We've been talking about consistently this here at the blog since 2015. All Trump has to say is "don't vote" and the GOP is screwed.







The Age of Discovery


200% if you count me.


How many times has the NYT failed to vet these people? Honestly, I've lost count.


  
Wrong ones based on the second condition.
Lots of people dunking on Hamid. Not enough but still lots.





"Hypocrisy is the hangover of an addiction to attention" -- A great line from an epic thread.

We haven't had a eugenics post for a while.

Still not as crazy as Neom.





New York Times Pitchbot has provided us with the perfect template for a major genre of think pieces.

 

Which is appropriate since the NYT has bought into every tech scam of the 21st century.







I don't quote Cory Doctorow often enough.


 The SAT (a topic we spent some time on a few years ago) is back in the news.

Putting me in the odd position of quoting Josh Barro



Remind me to come back to this one.

And now for something completely different...

The poster did some checking and this appears to be real.


No particular significance. Just a cool thread about bugs.


Apologies if I already posted this but I think this would be really funny.
Posted by at No comments:

Thursday, March 9, 2023

We need to talk about tests (in the "I got an A" sense, not the RCT sense)

We are hearing all sorts of hype about large language models passing various exams, but before we can have that conversation, we need to have one about how those tests are supposed to work, the assumptions them, and why a good test of human understanding can be meaningless if approached in a different way.

 Here's a post I wrote a few years ago about a type of test I encounter back when I was getting my BFA. As you read over the description, think about how a LLM would approach this task, and about what (if anything) its performance would tell us.

Friday, May 25, 2012

"Of course, Shakespeare was much newer at the time"

Back when I was an undergrad I took a class in Shakespeare. I'm mentioning this because a couple of aspects came back to me recently while thinking about education. [The second aspect was covered in this later post -- MP] The first was the format of the tests the teacher used. They consisted of a list of quotes from the four plays we had covered since the last test. Each quote had a pronoun underlined which came with a two part question: who was the speaker and who was the antecedent?

I've never seen that format used in another class (even by the same teacher) and I always thought it was an interesting approach. I wouldn't necessarily recommend using it widely but I'm glad I had it in at least one course. It was a method that encouraged attentive reading (particularly useful with Shakespeare).

Experiencing different styles of teaching and evaluation are part of a well-rounded education. I've seen a wide range approaches. Some were successful. Some were not. Some successful as one-shots but weren't models I'd suggest routinely following, like the number theory class I took that didn't allow mathematical notation (all proofs had to be written out in grammatical sentences without abbreviations or symbols -- more or less the way Fermat would have done it). That pedagogical diversity has been of immense value.

A book on quality control I read a few years ago said that quality in a QC sense was equivalent to a lack of variation; quality meant all parts came out the same. Sometimes I'm afraid that the some in the education reform movement are starting to think of uniformity as an end to itself.

 

Posted by at No comments:
Subscribe to: Posts (Atom)